2011 Speeding Survey Final Report

dot_1960_DS1.pdf

National Survey of Speeding Attitudes and Behaviors

2011 Speeding Survey Final Report

OMB: 2127-0613

Document [pdf]

Download: pdf | pdf

2011 National Survey
Of Speeding Attitudes
And Behaviors

DISCLAIMER
This publication is distributed by the U.S. Department of Transportation, National Highway
Traffic Safety Administration, in the interest of information exchange. The opinions,
findings, and conclusions expressed in this publication are those of the authors and not
necessarily those of the Department of Transportation or the National Highway Traffic
Safety Administration. The United States Government assumes no liability for its content or
use thereof. If trade or manufacturers’ names or products are mentioned, it is because they
are considered essential to the object of the publication and should not be construed as an
endorsement. The United States Government does not endorse products or manufacturers.
Suggested APA Format Citation:
Schroeder, P., Kostyniuk, L., & Mack, M. (2013, December). 2011 National Survey of
Speeding Attitudes and Behaviors. (Report No. DOT HS 811 865). Washington, DC:
National Highway Traffic Safety Administration.

Technical Report Documentation Page
1. Report No.

2. Government Accession No.

3. Recipient’s Catalog No.

DOT HS 811 865
4. Title and Subtitle

5. Report Date

2011 National Survey of Speeding Attitudes and Behaviors

December 2013
6. Performing Organization Code

7. Author(s)

8. Performing Organization Report No.

Paul Schroeder, Lidia Kostyniuk, Mary Mack
9. Performing Organization Name and Address

10. Work Unit No. (TRAIS)

Abt SRBI, Inc.
8403 Colesville Road, Suite 820
Silver Spring, Maryland 20910

11. Contract or Grant No.

DTNH22-08-F-00129
12. Sponsoring Agency Name and Address

13. Type of Report and Period Covered

National Highway Traffic Safety Administration
Office of Behavioral Safety Research
1200 New Jersey Avenue, SE
Washington, DC 20590

FINAL REPORT
14. Sponsoring Agency Code

15. Supplementary Notes

Eunyoung Lim (initial) and Randolph Atkins (final) were the Contracting Officer’s Technical
Representatives.
16. Abstract

The 2011 National Survey of Speeding Attitudes and Behavior (NSSAB) is the third in a series of
surveys on speeding that have provided data to help further the understanding of driving behavior and
to contribute to the development of countermeasures and interventions to reduce speeding. Like the
previous studies, this survey yields national estimates of behavior and attitudes toward speeding in the
United States. The present study differs from the earlier studies in that it developed and used a driver
typology based on the pattern of responses across six speeding behavior questions. Cluster analysis
identified three distinct groups of drivers with similar overall behavioral tendencies and, among those
categorized, 30% are nonspeeders, 40% are sometime speeders, and 30% are speeders. Driver type
is a powerful predictor of norms and attitudes toward speeding behavior, speeding countermeasures,
experience with sanctions and crash experience. This report details the findings from the 2011 NSSAB,
examining the data using the above mentioned driver typology as well as standard demographics. In
the final chapter, results from the current study are compared to those of the 2002 NSSAB and the
1997 NSSAB. Using data from over the last 14 years allows us to identify trends in speeding and
driving behavior, especially as new technologies such as cell phones become more pervasive in the
driving community.
17. Key Words

18. Distribution Statement

Speeding, Speeding Countermeasures, Unsafe Driving, Cell phone

Available to the public from the
National Technical Information
Service www.ntis.gov

19 Security Classif. (of this report)

20. Security Classif. (of this page)

Unclassified

21 No. of Pages

22. Price

182

Form DOT F 1700.7 (8/72)

Reproduction of completed page authorized

TABLE OF CONTENTS
Executive Summary ...................................................................................................................... 1
Chapter 1: Introduction ............................................................................................................... 8
Chapter 2: Driver Characteristics ............................................................................................. 10
Chapter 3: Driving Behavior on Different Types of Roads .................................................... 22
Chapter 4: Norms and Attitudes About Speeding ................................................................... 34
Chapter 5: Attitudes Toward Enforcement and Speeding Countermeasures ...................... 47
Chapter 6: Automated Photo Enforcement Devices ................................................................ 68
Chapter 7: Crash Experience .................................................................................................... 76
Chapter 8: Personal Sanctions................................................................................................... 83
Chapter 9: Other Risky Behavior ............................................................................................. 91
Chapter 10: Trend Analysis ..................................................................................................... 100
Conclusion ................................................................................................................................. 105
References .................................................................................................................................. 109

APPENDICES
Appendix A: Survey Instrument ............................................................................................. A-1
Appendix B: Survey Methodology .......................................................................................... B-1
Appendix C: Raking Output .................................................................................................... C-1

EXECUTIVE SUMMARY
The 2011 National Survey of Speeding Attitudes and Behavior (NSSAB) is the third in a series
of surveys conducted by the National Highway Traffic Safety Administration that focus on
speeding and yield national estimates of driver behavior and attitudes toward speeding in the
United States. The previous surveys were conducted in 1997 and 2002. For the 2011 survey, data
were collected via telephone interviews with 6,144 U.S. households. To account for the trend
toward cell phones as a replacement for landline telephones, the survey employed an overlapping
dual-frame sample design and contacted people living in households with landline telephones
and households that relied only or mostly on cell phones. Because young drivers are a high-risk
group of particular interest, the survey included an oversample of respondents 16 to 34 years old
so that national estimates could be obtained. Interviews were conducted from March 31, 2011, to
September 4, 2011.
This report presents the survey findings. Data are weighted to yield national estimates. Readers
are cautioned that some subgroup analyses are based on a smaller number of cases. The survey
questionnaire and a full description of the survey methodology are provided in the appendices to
this report.
Driver Characteristics
Driver Type. A driver typology based on the pattern of responses across six speeding behavior
questions was developed using cluster analysis. Three distinct groups of drivers with similar
overall behavioral tendencies were identified. Because of the nature of these behavioral
tendencies, the driver types are referred to as nonspeeders, sometime speeders, and speeders in
this report. Of those respondents categorized, 30% are nonspeeders, 40% are sometime speeders,
and 30% are speeders.
Drivers classified as speeders tended to be younger when compared to nonspeeders. One-half of
the drivers 16 to 20 years old were classified as speeders, as compared to 15% of drivers 65 or
older. Speeders were also more likely to have higher household incomes; 42% of drivers with
annual household incomes exceeding $100,000 were classified as speeders, while only 25% of
drivers with annual household incomes of $30,000 or less were in this driver type category.
Driving Frequency. More than 4 out of 5 drivers (82%) drive every day or almost every day.
Thirteen percent report driving several days a week, while 5% say they drive once a week or less
often.
Vehicle Type. The majority of drivers (57%) report they drive passenger cars most often. Close
to a fifth (18%) of drivers drive SUVs and 13% report driving pickup trucks most often. Almost
a tenth (9%) report driving vans or minivans most often.

Driving Behavior
Road Type. The majority of drivers drive on all types of roads. Nearly 4 out of 5 drivers (79%)
report driving on neighborhood or residential streets frequently. Frequent use of two-lane
highways and multi-lane divided highways is reported by 61% and 52% of drivers respectively.
Driving Speed. Overall, people drive at approximately the speed that they perceive to be safe for
the type of road that they are on. There are no substantive differences in the driving speed and
the perceived safe speed for all road types considered. Average driving speed on multi-lane
divided highways is 63.6 miles per hour (mph), approximately the same as the average perceived
safe speed limit (64 mph). The difference between these two measures on two-lane highways and
residential roads is even smaller.
Drivers who had been stopped by police within the past year and received a warning rather than
a speeding ticket, on average, report that one can travel 10.6 mph over the limit on multi-lane
divided highways and 11.4 mph over the limit on two lane highways before receiving a speeding
ticket. This is a larger margin than the average perceived “allowable” speed over the speed limit
reported by drivers who had been ticketed for speeding.
Norms and Attitudes about Speeding
Normative Attitudes. An overwhelming majority (91%) of drivers agreed with the statement
that “Everyone should obey the speed limits because it’s the law.” Two-thirds (67%) agreed
strongly with this statement. There was also agreement with the statement, “It is unacceptable to
exceed the speed limits by more than 20 mph.” More than 17 out of 20 drivers agreed and 76%
strongly agreed with this statement. Drivers also agreed that, “People should keep up with the
flow of traffic,” with 82% agreeing with this statement. Approximately one-half of drivers
agreed that speeding tickets have more to do with raising money than they do with reducing
speeding (51%) as well as with the statement, “There is no excuse to exceed the speed limits”
(48%). Less than a fifth of drivers agreed with the statements, “If it is your time to die, you’ll
die, so it doesn’t matter whether you speed,” (17%) and “Driving over the speed limit is not
dangerous for skilled drivers” (16%).
When we examine normative attitudes by driver type, we find that less than one-half (48%) of
the drivers classified as speeders strongly agree that “Everyone should obey the speed limits
because it’s the law.” However, more than 4 out of 5 drivers classified as nonspeeders (81%)
strongly agree with this statement. Almost two-thirds (64%) of speeders strongly agree that
“People should keep up with the flow of traffic,” but only 42% of the nonspeeders strongly agree
with this statement. While two-fifths (41%) of nonspeeders strongly agree that “There is no
excuse to exceed the speed limit,” only 1 in 6 speeders (16%) strongly agree with this statement.
Speeders are twice as likely (11%) as sometime speeders (5%) or nonspeeders (5%) to agree that,
“Driving over the speed limit is not dangerous for skilled drivers.”
Personal Attitudes. Three in five (60%) drivers agreed that they often get impatient with slower
drivers. Close to one-half of all drivers (47%) agreed with the statement “I worry a lot about

having a crash.” There was considerably less agreement with the statements, “Speeding is
something I do without thinking” (27%), “I enjoy the feeling of driving fast” (27%), and “I try to
get where I am going as fast as I can” (20%). Only 9% of respondents agreed with the statement,
“I consider myself a risk taker while driving.”
When we examine personal attitudes by driver type, drivers classified as speeders were almost
three times as likely as sometime speeders to strongly agree with the statements, “I often get
impatient with slower drivers” (45% versus 18%), “I enjoy the feeling of driving fast” (19%
versus 6%), and “I try to get where I am going as fast as I can” (11% versus 3%).
Driving the Speed Limit. Over 4 out of 5 (82%) drivers agreed with the statement, “Driving at
or near the speed limit makes it easier to avoid dangerous situations.” There was also agreement
with the statement, “Driving at or near the speed limit reduces my chances of an accident” (79%)
and the statement, “Driving at or near the speed limit uses less fuel” (73%). More than 2 out of 5
respondents (42%) agreed that driving at or near the speed limit makes it difficult to keep up
with traffic, and less than a fifth (17%) agreed that driving at or near the speed limit makes them
feel annoyed.
Attitudes Toward Speeding Countermeasures
Importance of Reducing Speeding. Close to half of the respondents (48%) said that it was very
important that something be done to reduce speeding on the nation’s roadways. Almost 2 out of
5 (39%) said that it is somewhat important, while 8% of drivers said that it was not too important
and 3% said that it was not at all important.
When we examine attitudes toward speeding countermeasures by driver type, among drivers
classified as speeders, 30% state that reducing speeding is very important, while 49% of those
classified as sometime speeders and 61% of drivers classified as nonspeeders believe that it is
very important.
Enforcement of Speed Limits. Close to half of all respondents (48%) said that the speed limit
should be enforced all of the time. Almost a third (30%) said it should be enforced often and
18% said it should be enforced sometimes. One in seven drivers (13%) said they see motor
vehicles pulled over on the roadway all the time. Three in ten said they see vehicles pulled over
often and 40% said sometimes. Interestingly, 16% said they see vehicles pulled over rarely,
which is higher than those who said they see vehicles pulled over all the time.
Use of Countermeasures in the Community. The two speeding countermeasures with the
highest approval rating were electronic signs by the road that warn drivers that they are speeding
and should slow down (89%), and increasing public awareness of the risks of speeding (88%). It
should be noted that both of these items do not include any specific penalties to drivers. Four out
of five drivers (80%) thought that increased use of speed cameras in dangerous or high-crash
locations was a good idea, and two-thirds (66%) thought that more frequent ticketing for
speeding was a good idea. The least popular idea was issuing higher fines for speeding tickets,
which 2 out of 5 (41%) respondents thought was a good idea.

When we examine this by driver type, among drivers classified as speeders, just over one-half
(54%) think that increasing ticketing for speeding is a good idea. Among drivers classified as
sometime speeders and nonspeeders this percentage is 65% and 78%, respectively. Only 32% of
speeders compared to 39% of sometime speeders and 51% of nonspeeders think that higher fines
for speeding tickets is a good idea.
In-Vehicle Countermeasures. A device in the motor vehicle that notifies you if you are
speeding was endorsed by 61% of drivers. A device that records speed data and reports it to the
insurance company to lower premiums was endorsed by 62% of drivers. A device that slows
down the vehicle when it senses another car or object is too close to the vehicle was endorsed by
60% of drivers. Roughly the same percentage who thought each item was a good idea also
reported that the countermeasure would prevent them from speeding.
Female drivers were more likely than male drivers to agree that in-vehicle countermeasures were
a good idea, and also to indicate that the countermeasures would prevent them from speeding.
Increasing levels of formal education and household income were negatively associated with
agreement that countermeasures were a good idea and that they would prevent speeding.
When we examine this by driver type, about 4 in 10 drivers classified as speeders (43%) reported
that a speeding notification inside the car would prevent them from speeding, in contrast to 62%
and 69% of drivers classified as sometime speeders and nonspeeders. Among speeders, 54%
indicated that a device that records speeding information and reports it to the insurance company
would prevent them from speeding. Among the other driver types, this percentage was 65% for
sometime speeders, and 73% for nonspeeders. Only 45% of speeders say a device in their vehicle
that slows the vehicle down if an object gets too close would prevent them from speeding,
compared to 56% of sometime speeders and 60% of nonspeeders.
Use of Automated Photo Enforcement Devices
Heard of Speed Cameras. The overwhelming majority of drivers (85%) reported that they have
heard of speed cameras being used to ticket drivers who speed.
Location of Speed Cameras. The majority of drivers thought that speed cameras would be
useful in school zones (86%), places where there have been many accidents (84%), construction
zones (74%), areas where it would be hazardous for a police officer to stop a driver (70%), and
areas where stopping a vehicle could cause traffic congestion (63%). A little over one-third
(35%) of drivers thought speed cameras would be useful on all roads.

Experience with Speed Cameras. More than one-third (37%) of drivers reported that there are
speed cameras in use along the routes they usually drive. Interestingly, 10% of drivers did not
know whether speed cameras were being used along the routes they normally drive. Less than 1
in 10 drivers (8%) has received a speeding ticket in the mail from a speed camera.
Purpose of Speed Cameras. Drivers are more likely to agree with the statement that “Speed
cameras are used to generate revenue,” (70%) than they are to agree that “Speed cameras are
used to prevent accidents” (55%). This pattern holds true among those who strongly agree with
each statement as well, (38% versus 29%, respectively).
Crash Experience
Speeding-Related Crash in Past 5 Years. The majority of respondents (96%) had not been in
any speeding-related crashes in the past 5 years. Only 3% had been in one speeding-related crash
in the past 5 years and even fewer (1%) had been in two or more speeding-related crashes in the
past 5 years.
A greater percentage (11%) of drivers 16 to 20 had at least one speeding-related crash in the past
5 years than any other age group, even though drivers in this age group may not have been
driving for all of the past 5 years. Of drivers 21 to 24, 9% had a speeding-related crash in the past
5 years, while only 1% of drivers 55 and older had at least one speeding-related crash in the past
5 years.
Injuries Resulting from Crash. Of the respondents that reported being in a speed-related crash,
most (68%) reported they were not injured in their most recent speeding-related crash, while
nearly 1 in 3 (29%) reported being injured. The remaining 3% of respondents in speed-related
crashes either did not know if they were injured or refused to answer.
Personal Sanctions
Stopped for Speeding. Less than 1 driver in 10 (9%) reported being stopped for speeding in the
past 12 months. The majority (84%) of drivers who were stopped for speeding were stopped only
once in the past 12 months. One in seven (15%) drivers were stopped 2 to 4 times in the past 12
months, and 1% were stopped 5 times or more. Younger drivers and male drivers were more
likely than older drivers and female drivers to have been stopped. Among the drivers classified
as speeders, 1 in 5 (20%) were stopped for speeding in the past 12 months. Only 4% and 5% of
drivers classified as nonspeeders and sometime speeders, respectively, were stopped.
Sanctions for Speeding. Most (68%) respondents received a ticket if they were stopped for
speeding. More than a quarter (27%) received a warning, and 1 in 20 (5%) did not receive a
ticket or a warning. The proportion of drivers receiving a ticket for speeding was highest (75%)
for drivers 25 to 34 and lowest (61%) for drivers 16 to 20. More than one-third (34%) of drivers
35 to 44 or 65 and older were given a warning instead of a ticket, while 22% of drivers 25 to 34
received a warning.

Risky Behavior and Cell Phone Behavior
Seat Belt Usage. About 9 in 10 (89%) drivers report that they wear their seat belts all of the time
while driving their primary vehicle. While there are 11% who still do not wear their seat belts all
the time, only 1% of drivers report that they never wear their seat belts while driving.
Alcohol Use. Only a small proportion of drivers overall (2%) reported driving a vehicle after
they thought they had drank too much alcohol to drive safely. The highest percentage of drivers
who admit to this is among drivers 25 to 34 years old, where 4% reported that they have driven
after they had consumed too much alcohol to drive safely. Among drivers 16 to 20 and 21 to 24,
this percentage was 3%.
Use of Cell Phone While Driving. The majority of drivers (89%) drive with cell phones in their
vehicle. Only about 1 in 10 (11%) reported not having a cell phone in their vehicles when
driving. Talking on the phone while driving was reported more often than reading or sending text
messages while driving. A total of 22% of drivers stated they talk on their phones while driving
during half or more of their trips. Approximately 5% reported they text while driving during half
or more of their trips. A small portion of drivers reported they talk (3%) or text (1%) while
driving during all of their trips.
Among drivers 25 to 34, 16% report talking on the cell phone, and 6% report sending or reading
text messages on most trips. Only 3% of drivers 65 or older talk on the phone on most trips and
none report reading or sending text messages while driving. The youngest drivers are most likely
to read and send text messages while driving While 4% of drivers 16 to 20 report talking on the
cell phone on all or most of their trips, 8% of this age group state that they read or send text
messages on all or most trips.
When we examine cell phone use by driver type, speeders (16%) are more likely than sometime
speeders (8%), and nonspeeders (7%) to talk on their cell phones while driving.
Use of Hands-Free Devices in Vehicles. One-third (32%) of drivers hold the phones in their
hands when they talk on them while driving. About 1 in 5 (19%) use the speakerphone feature
built into their cellular devices. Very few (2%) drivers squeeze the phones between their ears and
shoulders to talk on it while driving.
Trends from Previous Speeding Surveys
Respondents reported driving less frequently in 2011 than in either 2002 or 1997 and some
attitudes toward speeding have changed. In the current survey, 81% of drivers report driving
every day or almost every day, as compared to 83% in 2002 and 88% in 1997. Enjoyment of
driving fast, driving as fast as possible, and the belief that speed increases driver alertness appear
to have decreased over this time period. In 1997, 40% of drivers reported enjoying the feeling of
driving fast, while this percentage dropped to 34% in 2002 and to 27% in 2011. In both 1997 and
2002, about 30% of drivers agreed with the statement that they try to go as fast as they can to get
somewhere. In 2011, only 21% of drivers agreed with this statement. Similarly, agreement with

the statement “the faster I drive the more alert I feel” decreased from about 30% in both 1997
and 2002 to 15% in 2011.
On the other hand, some attitudes and behaviors have not changed since 1997. The percentage of
drivers who reported worrying about having crashes in 2011 (48%) is about the same as it was in
1997 (47%) and 2002 (46%). The percentage of drivers who reported being impatient with
slower drivers was 60% in 1997. This dropped to 53% in 2002 and was back up to 61% in 2011.
The portion of drivers stopped by police for speeding, and the rate of receiving speeding tickets
if stopped have remained relatively stable over the past 14 years. In 1997, 9% of drivers reported
having been stopped by police for speeding in the past 12 months. This rose to 11% in 2002 and
was back down to 9% in 2011. Of those stopped for speeding, 65% reported being ticketed in
1997, 70% in 2002, and 68% in 2011.

CHAPTER 1: INTRODUCTION
Background
The National Highway Traffic Safety Administration directs highway safety and vehicle
consumer programs in the United States that promote the safety of automobiles, their occupants,
and other road users. NHTSA’s mission includes the reduction of traffic crashes, fatalities and
injuries, together with their economic repercussions.
Speeding, defined as exceeding the speed limit or driving too fast for conditions, is one of the
most common factors in traffic crashes. In the decade from 2002 through 2011, speeding was a
contributing factor in nearly one-third of all fatal crashes, claiming a total of 123,804 lives and
resulting in an annual economic cost to society of approximately 40 billion dollars per year
(NHTSA, 2013). In 2011 alone, 9,944 lives were lost in speeding-related crashes.
Research has demonstrated a strong correlation between drivers’ attitudes toward speeding and
other driving behaviors and actual traffic outcomes (Elliot, Armitage, & Baughan, 2003; (De
Pelsmacker & Janssens, 2007). Because attitudes toward speeding, other driving habits, and the
interaction of driving habits with intentions have been found to be of particular importance in
traffic outcomes, these self-reported driving measures of intention and attitude have often served
as approximations of behavior in models used to design interventions to reduce speeding and
other hazardous driving behaviors (Parker & Manstead, 1996; Conner, Lawton et al., 2005).
Because knowledge of drivers’ attitudes toward speeding is important to its mission of
improving traffic safety, NHTSA periodically conducts a National Survey of Speeding Attitudes
and Behaviors to collect this information.
To date, NHTSA has conducted three rounds of the National Survey of Speeding Attitudes and
Behavior, in 1997, in 2002, and in 2011. This report presents findings from the 2011 National
Survey on Speeding Attitudes and Behavior. Specifically, this self-report survey looks at
speeding behavior of drivers, their norms and attitudes about speeding, their attitudes toward
various speeding countermeasures, and their experience with speed-related crashes, as well
personal sanctions for speeding.
Methodology
The 2011 National Survey of Speeding Attitudes and Behavior was conducted from March 31,
2011, until September 4, 2011. A total of 6,144 telephone interviews were conducted among a
nationally representative sample of people 16 or older who drive motor vehicles. To account for
the current trend toward cell phone use, 1,137 interviews were conducted with people from
households that relied only or mostly on cell phones and 4,507 interviews were conducted with
people from households with landline phones. In addition, 500 interviews were completed with
an oversample of drivers who were 16 to 34 years old, an age group that is over represented in
crashes and of particular interest for traffic safety. The samples were combined and weighted to

produce national estimates of the target population within specified limits of expected sampling
variability, from which valid generalizations can be made to the general public.
For a complete description of the methodology and sample disposition, including the
computation of weights, please refer to Appendix B and Appendix C.
The percentages presented in this report are weighted to accurately reflect the national
population of those 16 or older. Unweighted sample sizes (Ns) are included to show the exact
number of respondents answering a given question, and to allow interested readers to estimate
sampling precision.
Percentages for some items may not add to 100% because of rounding, or because the question
allowed for more than one response. In addition, the number of cases involved in some subgroup
analyses may not sum to the grand total of those who responded to the primary question being
analyzed. Reasons for this include nonresponse on the grouping variable (e.g., some respondents
answer “Don’t know” or refused to answer a question), or use of only selected subgroups in the
analysis.
Please note that when categories of responses that appeared in a table or figure are combined for
discussion in the text (e.g., combining “very likely” and “somewhat likely”), the total shown is
based on the sum of the numbers in the tables or figures, and not on the results of an additional
analysis that combined the two response categories. For rounding purposes, all variables are
rounded based on two decimal places. Any value that had a decimal of .50 or greater was
rounded up and any value that had a decimal below .50 was rounded down.

CHAPTER 2
DRIVER CHARACTERISTICS
This chapter describes the demographics of the sample of respondents, presents the typology of
drivers developed from the patterns of responses to a set of speeding-behavior questions, and
shows the distribution of these driver types by demographics.
Because of the shift to exclusive (or almost exclusive) cell phone use by an increasing portion of
the U.S. population, some groups of people are often not accessible by landline telephones. To
capture a sample of respondents representative of drivers 16 and older in the United States, this
survey used both a cell phone sample and a landline based sample (see Table 2-1). Of the total
6,144 respondents, 1,137 were sampled via cell phone (18.51%) and 5,007 (81.49%) were
sampled through landline phones. The respondents in the cell phone sample are younger than
those in the landline sample. Drivers 16 to 20 make up 9% of the cell phone sample, and more
than 20% of the cell phone sample is under 25, while less than 7% of the landline sample is
under 25. By contrast, drivers in the sample 65 and older make up more than 25% of respondents
in the landline sample, but only 5.9% of the cell sample. The cell phone sample also allowed
better representation of lower household income, education level, and racial minority groups.
While the cell phone sample was more urban, the overall sample was fairly evenly split between
rural and urban respondents. There was little difference between the cell and landline sample in
frequency of driving and type of vehicle.
Table 2-1: Demographics by Sample Type - Unweighted
Cell Phone Sample Landline Sample
Total Sample
(N=1,137)
(N=5,007)
(N=6,144)
Gender***
Female
44.4%
58.8%
56.1%
Male
55.6%
41.2%
43.9%
Age***
Mean
38.43
52.18
49.61
16 to 20
9.4%
3.8%
4.9%
21 to 24
12.4%
2.9%
4.6%
25 to 34
26.2%
13.0%
15.5%
35 to 44
16.8%
13.1%
13.7%
45 to 54
17.8%
19.9%
19.5%
55 to 64
11.6%
21.9%
19.9%
65 or older
5.9%
25.5%
21.9%
2010 Household Income***
Less than $30,000
31.5%
22.4%
24.2%
$30,000 to $49,999
19.6%
19.3%
19.3%
$50,000 to $74,999
20.4%
21.0%
20.9%
$75,000 to $99,999
12.3%
14.9%
14.4%
$100,000 or More
16.2%
22.4%
21.2%
10

Table 2-1: Demographics by Sample Type - Unweighted (Continued)
Cell Phone
Landline Sample
Total Sample
Sample
(N=5,007)
(N=6,144)
(N=1,137)
Education***
No HS Degree
11.1%
6.8%
7.6%
HS Graduate
58.4%
53.5%
54.4%
College Degree
20.0%
23.4%
22.8%
Graduate Degree
10.5%
16.3%
15.3%
Race/Ethnicity
White
73.6%
82.9%
81.2%
Black
13.5%
7.6%
8.7%
Hispanic
9.1%
3.9%
4.9%
Asian
3.5%
2.3%
2.5%
Other
5.4%
3.5%
3.9%
Metro Status
Urban
55.7%
50.0%
51.1%
Nonurban
44.3%
50.0%
48.9%
Frequency of Driving***
Everyday or Almost
85.1%
81.0%
81.8%
Everyday
Several Days a Week
9.2%
13.8%
12.9%
Once a Week or Less
4.1%
4.3%
4.3%
Only Certain Times of
1.5%
0.9%
1.0%
Year
Primary Type of Vehicle
Car
55.6%
57.4%
57.1%
Van/Mini-Van
7.5%
9.5%
9.2%
SUV
17.0%
18.7%
18.4%
Pickup Truck
16.4%
12.6%
13.3%
Other Truck
2.3%
1.0%
1.2%
Motorcycle
0.7%
0.4%
0.4%
Other/Don’t Know
0.5%
0.3%
0.4%
*** p<.001

Driver Types
In examining drivers’ attitudes and speeding behaviors, it is useful to group drivers by their
driving tendencies. Rather than rely on any single indicator of general driving tendency or prior
assumptions about appropriate categories of drivers, this study developed a typology of drivers
using cluster analysis of responses to six questions about driving and speeding tendencies.
Cluster analysis allowed the identification of discrete types of drivers based on the overall
pattern of responses across all six speeding behavior questions.
Table 2-2 shows the response distributions to each of the six driving and speeding questions used
in the cluster analysis. Two questions addressed general driving tendencies. Respondents were
asked whether they tended to pass other cars or be passed by other cars more frequently and
whether they tended to stay with slower moving traffic or keep up with faster traffic. Over onehalf (59%) of drivers report a tendency for other cars to pass them more often than they pass
other cars; however, nearly one-half (45%) say they tend to keep up with faster traffic rather than
staying with slower moving traffic. Speeding appears less common when respondents are asked
about their behavior in specific contexts. Three questions addressed speeding behaviors under
particular driving conditions: respondents were asked how frequently they drive 15 mph over the
speed limit on multi-lane, divided highways, how frequently they drive 15 mph over the speed
limit on two-lane highways, and how frequently they drive 10 mph over the speed limit on
neighborhood or residential streets. Even when asked about driving on multi-lane divided
highways, nearly one-half (48%) reported never exceeding the speed limit by 15 mph. A final
question asked respondents how many times in the previous 12 months they had been stopped by
police for speeding. Only 9% of respondents had been stopped in the last year. Of those who
were stopped, the vast majority (84%) had been pulled over only once.
Table 2-2. Questions Employed in Cluster Analysis (Weighted)
Q3. Which best describes your driving…
I tend to pass other cars more often than other cars pass me.
Other cars tend to pass me more often than I pass them
Both about equally
Unweighted N

26.6%
59.3%
14.1%
5,995

Q4. When driving I tend to…
Stay with slower moving traffic
Keep up with the faster traffic
Both about equally
Unweighted N

34.9%
44.8%
20.3%
5,952

Table 2-2. Questions Employed in Cluster Analysis (Weighted)
N
Often
Sometimes
How often would you say you…
Q5e. Drive 15 miles an hour over
5,878
4.9%
14.2%
the speed limit on Multi-Lane,
divided Highways?
Q6e. Drive 15 miles an hour over
5,865
2.3%
7.5%
the speed limit on Two-Lane
Highways?
Q7e. Drive 10 miles an hour over
6,032
2.8%
7.9%
the speed limit on Neighborhood
or Residential streets?
How many times have you been stopped for speeding in
the past twelve months?
MEAN:
None
Once
Twice
3 or more times
Unweighted N

Rarely
33.2%

Never
47.7%

25.9%

64.2%

24.8%

64.4%

0.1174
90.8%
7.7%
1.0%
0.5%
6,144

Three distinct clusters of drivers with similar overall behavioral tendencies were identified and
86% of respondents were classified as one of three distinct types of driver. There were 845
respondents (14% of the sample) who could not be classified by driver type because some did
not answer all six speeding behavior questions or because their responses to these questions did
not fit well with any of the three clusters. These respondents are excluded from any analyses
which use driver type; however, they are included in analyses within this report where driver
type is not employed. The core characteristic of each of the three groups identified in the cluster
analysis was determined by examining how each group scored on each speeding behavior
variable. As can be seen in Figure 2-1 and 2-2, one group was composed of drivers who
consistently reported speeding, one group was composed of drivers who rarely reported
speeding, and a third group contains those drivers who sometimes speed. For the purposes of this
report, these groups were named: nonspeeders, speeders, and sometime speeders, respectively.
Of those respondents classified by this typology, 30% are nonspeeders, 40% are sometime
speeders, and 30% are speeders.
Figure 2-1: Speeding Behavior on Various Road Types by Driver Type
Nonspeeders (N=1,579)
Sometime Speeders (N=2,148)
Speeders (N=1,572)

100%

75%

25%

91%

29%

79%

Never Drive 15mph
over speed limit on
Multi-lane Highways

38%

30%
62%

17%
25%

22%
0%

45%

46%

62%

50%

21%

Often Drive 15mph
over the speed limit
on Multi-lane
Highways

Never Drive 15mph
over speed limit on
Two-lane Highways

25%
Often Drive 15mph
over the speed limit
on Two-lane
Highways

Never Drive 10mph
over speed limit on
Residential Streets

Often Drive 10mph
over the speed limit
on Residential
Streets

Q5e/Q6e. How often would you say you drive 15 miles an hour over the speed limit on Multi-Lane, Divided Highways/Two-lane
Highways, one lane in each direction?
Q7e. How often would you say you drive 10 miles an hour over the speed limit on Neighborhood or Residential Streets?
Base: All Respondents Assigned a Driver Type
Unweighted N=See Chart

Two key questions that helped define driver type dealt with driver tendencies on the road. The
first, whether the respondent tends to pass or be passed by other cars on the road, clearly
separates speeders from sometime speeders and nonspeeders, with 100% of speeders saying they
tend to pass other cars. The second question, whether the driver stays with slower traffic or keeps
up with faster traffic, further helps to define the sometime speeder category, with 47% of
sometime speeders saying they tend to stay with the slower traffic and 28% of sometime
speeders keeping up with the faster traffic (see Figure 2-2).
Figure 2-2: Driver Type by Driving Tendency***
Nonspeeders (N=1,579)
Sometime Speeders (N=2,148)
Speeders (N=1,572)

100%
75%

15%
43%

50%
25%
0%

43%

28%

100%
47%

54%

Other cars tend to
pass me

57%

10%
I tend to pass
other cars

Stay with slower
traffic

Keep up with
faster traffic

Q3. Which of the following statements best describes your driving? I tend to pass other cars more often than other cars pass me
OR Other cars tend to pass me more often then I pass them?
Q4. When driving I tend to stay with slower moving traffic OR keep up with the faster traffic.
Base: All Respondents Assigned a Driver Type
Unweighted N=See Chart *** p < .001

Figure 2-3 shows the distribution of driver types among male and female drivers. While the
percentage of sometime speeders in both genders is about the same (39% and 40%), about onethird of women drivers (32%) are classified as nonspeeders, while only one-quarter of men
(25%) fall into this category. Conversely, 36% of men, but only 28% of women are speeders.
Figure 2-3: Driver Type by Respondent Gender***

100%

25%

32%

75%

50%

39%

Nonspeeders (N=1,579)

40%

25%

36%

28%

0%
Male

Female

SA3. Record Respondent’s Gender by Observation.
Base: All Respondents Assigned a Driver Type
Unweighted N=See Chart *** p < .001

Sometime Speeders (N=2,148)
Speeders (N=1,572)

Figure 2-4 indicates that speeders tend to be younger when compared to nonspeeders and
sometime speeders. Half of the drivers 16 to 20 are speeders as compared to only 15% of those
65 or older. The opposite relationship by age is seen among nonspeeders, with 38% of those
older than 55 but only 17% of those 20 or younger in this category.
Figure 2-4: Driver Type by Respondent Age***
Nonspeeders (N=1,579)
Sometime Speeders (N=2,148)
Speeders (N=1,572)

100%

17%

24%

20%

29%

38%

75%

34%

30%

35%
36%

50%

42%
43%

25%

50%

47%

45%

34%

29%

20%

47%

15%

0%
16 to 20

21 to 24

25 to 34

35 to 44

D1. How old are you?
Base: All Respondents Assigned a Driver Type
Unweighted N=See Chart *** p < .001

45 to 54

55 to 64

65+

Figure 2-5 indicates that speeders tend to live in households with higher household incomes,
compared to nonspeeders. In the highest household income group ($100,000+), 42% are
categorized as speeders, while only 20% are categorized as nonspeeders. In contrast, of those in
the lowest household income group (<$30,000), 37% are nonspeeders, while only 25% were
categorized as speeders. While income tends to increase with age, very young respondents with
high household incomes and older respondents with low household incomes are an exception to
this trend. Respondents 20 and younger (11.4%) were more likely to report household incomes
over $100,000 a year than respondents 21 to 24 (5.5%) and 25 to 34 (11.4%), probably because
they still live at home with parents who are older and thus have higher incomes. In addition,
respondents 65 and older (6.5%) were less likely to report household incomes over $100,000 a
year than respondents in their thirties, forties and fifties. As a result, many respondents who live
in households which have incomes of more than $100,000 a year may be younger than those who
report household incomes less than $100,000, which could account for the increase in speeding
drivers in this demographic group; and many of the older drivers (65+) have incomes under
$50,000 per year (61%), which could account for the greater percentage of nonspeeders in the
income groups below $50,000/year.
Figure 2-5: Driver Type by Household Income***
Nonspeeders (N=1,579)
Sometime Speeders (N=2,148)
Speeders (N=1,572)

100%
37%

25%

27%

40%

35%

36%

38%

$50,000-$74,999

$75,000-$99,999

30%

20%

75%

50%

38%

43%

25%

37%

25%

28%

Less than $30,000

$30,000-$49,999

42%

$100,000+

D9. Which of the following categories best describes your total household income before taxes in 2010? Your best estimate is
fine.
Base: All Respondents Assigned a Driver Type
Unweighted N=See Chart *** p < .001

Figures 2-6 through 2-8 look at each driver type by geographic region; specifically, the 10
NHTSA Regions. The difference between regions is no more than 10 percentage points;
however, speeders are more prevalent in the western part of the US, particularly Regions 8 and 9,
while Regions 4, 5, 7, and 10 are the least likely to contain speeders.
Figure 2-6: Percentage of Speeders by NHTSA Region***

Base: All Respondents Assigned a Driver Type
*** p < .001

Sometime speeders are most likely to be found in the northeastern states, Regions 1 and 2. The
Midwest and Pacific Northwest, Regions 7, 8 and 10, are the least likely to have sometime
speeders.
Figure 2-7: Percentage of Sometime Speeders by NHTSA Region***

Base: All Respondents Assigned a Driver Type
*** p < .001

Figure 2-8 shows the percentage of nonspeeders by NHTSA Region. The Pacific Northwest and
the central Midwest have the highest proportion of nonspeeders. The western states and New
England have the lowest proportion of nonspeeders in the United States. The range between the
regions is more than 20 percentage points, the largest range for the three driver types.
Figure 2-8: Percentage of Nonspeeders by NHTSA Region***

Base: All Respondents Assigned a Driver Type
*** p < .001

CHAPTER 3
DRIVING BEHAVIOR ON DIFFERENT TYPES OF ROADS
The network of roads and highways in the United States is made up of a variety of road types
ranging from local residential and neighborhood streets to multi-lane divided highways, such as
those in the Interstate system of roads. The road types differ in the level of access they provide to
the surrounding land, by their geometric characteristics, as well as their design and speed limits.
Most drivers frequently travel on different types of roads. Figure 3-1 shows drivers’ frequency of
use of different types of roads. Neighborhood or residential streets are the most regularly used
road types, with nearly 80% of drivers (79%) reporting frequent use of these streets. Drivers also
frequently travel on multi-lane, divided highways and two-lane highways. About 80% of drivers
at least sometimes use multi-lane and two-lane highways.
Figure 3-1: How Often Drive on Various Road Types
Multi-lane Divided Highways

Two-lane Highways

Residential Streets

100%

79%
75%

61%
52%
50%

28%
25%

20%

15% 14%
5%

12%

0%
Never

Rarely

Sometimes

Frequently

Q5a/Q6a/Q7a. How often do you drive on (Multi-lane Divided Highways/Two Lane Highways, one lane in each
direction/Neighborhood or Residential streets)?
Base: All Respondents
Unweighted N=6,144

Comparing average driving speed and average perceived safe speed limits on various road types
shows that, overall, people drive at approximately the same speed that they perceive to be safe.
Figure 3-2 compares the average perceived safe limit to the average travel speed in mph at
which people drive on multi-lane divided highways, two-lane highways and residential streets.
There are no substantive differences between the average driving speed and average perceived
safe speed limits on all three road types. Average driving speed on multi-lane divided highways
is 63.6 mph, roughly the same as the average perceived safe limit (64 mph). The differences on
two-lane highways and residential roads are similar in scale.
Figure 3-2: Reported Driving Speed and Perceived Safe Driving Speed by Road Type
Driving Speed

Safe Driving Speed

63.6

64.0

48.9

49.0

26.9

27.2

0
Multi-lane Divided
Highways (N=5,895)

Two-lane Highways
(N=5,881)

Residential Streets
(N=6,048)

Q5c/Q6c/Q7c. What do you consider to be a safe speed limit for (most) (ROAD TYPE)s in good weather during the day?
Q5d/Q6d/Q7d. When driving on (ROAD TYPE)s in good weather during the day, how fast do you normally drive?
Base: Respondents Who Drive on Road Type
Unweighted N=See Chart

Average driving speed was compared to average perceived safe speed by driver type and across
different demographic groups. Figures 3-3 through 3-5 show average driving speed and average
perceived safe speed limit across age groups for the three different road types. On all road types
the general trend is that young drivers’ average driving speed is slightly faster than the speed
they consider as safe, but as drivers age, their average driving speed drops below the speed they
recognize as a safe limit.
Figure 3-3: Reported Driving Speed and Perceived Safe Driving Speed
on Multi-Lane Divided Highways by Age***
Driving Speed

Safe Driving Speed

66
65.1

65.1 65.1

64.7

63.4

64.4
63.7

63.1

64.1
63.6

63.6
62.9

62.8
62.0

60
16 to 20
(N=279)

21 to 24
(N=275)

25 to 34
(N=914)

35 to 44
(N=815)

65+
45 to 54
55 to 64
(N=1,157) (N=1,175) (N=1,214)

Q5c. What do you consider to be a safe speed limit for (most) Multi-Lane, Divided Highways in good weather during the day?
Q5d. When driving on Multi-Lane, Divided Highways in good weather during the day, how fast do you normally drive?
Base: Respondents Who Drive on Multi-Lane, Divided Highways
Unweighted N=See Chart *** p < .001

Figure 3-4: Reported Driving Speed and Perceived Safe Driving Speed
on Two-Lane Highways by Age***
Driving Speed

Safe Driving Speed

50
49.7

49.6
49.3

49.2

49.3
49.0

48.8

49.0
48.8

49.1

48.9

48.6

48.1
48

47
16 to 20
(N=275)

21 to 24
(N=255)

25 to 34
(N=898)

35 to 44
(N=799)

45 to 54
55 to 64
65+
(N=1,144) (N=1,169) (N=1,280)

Q6c. What do you consider to be a safe speed limit for (most) Two-Lane Highways, one lane in each direction, in good weather
during the day?
Q6d. When driving on Two-Lane Highways, one lane in each direction, in good weather during the day, how fast do you
normally drive?
Base: Respondents Who Drive on Two-Lane Highways
Unweighted N=See Chart *** p < .001

Figure 3-5: Reported Driving Speed and Perceived Safe Driving Speed
on Residential Streets by Age***
Driving Speed

Safe Driving Speed

30
28.8

29
27.9

28
27

26.9

27.3 27.2
26.8

27.0

26.7 26.7
25.8

27.2

27.4

26.1

25
24
23
16 to 20
(N=292)

21 to 24
(N=275)

25 to 34
(N=924)

35 to 44
(N=824)

45 to 54
55 to 64
65+
(N=1,167) (N=1,199) (N=1,300)

Q7c. What do you consider to be a safe speed limit for (most) Neighborhood or Residential streets, one lane in each direction, in
good weather during the day?
Q7d. When driving on Neighborhood or Residential streets in good weather during the day, how fast do you normally drive?
Base: Respondents Who Drive on Neighborhood or Residential streets
Unweighted N=See Chart *** p < .001

Low household income drivers believe the safe driving speed on highways is significantly lower
than drivers with higher household incomes and they also report slower driving speeds than what
they perceive to be safe (See Figures 3-6 through 3-8). On average, drivers with annual
household incomes of $75,000 or more report 65.8 mph as a safe speed limit on multi-lane
highways. The average estimated safe speed limit reported by drivers with annual household
incomes less than $30,000 for multi-lane highways was 62.2 mph. Those in the lowest household
income group, on average, drive 0.9 mph slower on multi-lane divided highways.
Figure 3-6: Reported Driving Speed and Perceived Safe Driving Speed
on Multi-Lane Divided Highways by Household Income***
Safe Driving Speed

Driving Speed
70

65.5
64.6

63.3

65.8

65.6

65.8

64.7

63.8

62.2
61.3

60
Less than
$30,000
(N=1,163)

$30,000 to
$49,999
(N=988)

$50,000 to
$74,999
(N=1,081)

$75,000 to
$99,999
(N=752)

$100,000 or
more
(N=1,097)

In Figure 3-7, a similar pattern emerges for two-lane highways, although the middle household
income group ($50,000 to $74,999) reports the highest average safe driving speed on these roads
(50.1 mph). The upper household income groups still report driving slightly faster than the speed
they would consider safe, while the lower household income groups drive slightly slower than
the speed they consider safe on two-lane highways.
Figure 3-7: Reported Driving Speed and Perceived Safe Driving Speed
on Two-Lane Highways by Household Income***
Driving Speed

Safe Driving Speed

51
50.3
50

50.1

49.9
49.5

49.2

49.3

49.2

48.8

49
48.2
48

47.8

$30,000 to
Less than $30,000
(N=1,197)
$49,999 (N=980)

$50,000 to
$75,000 to
$74,999 (N=1,058) $99,999 (N=740)

$100,000 or more
(N=1,080)

Drivers are much more likely to report driving slower than what they consider to be a safe
driving speed, regardless of household income, when it comes to residential streets. Four of the
five household income categories reported driving slower than the safe driving speed on these
streets, and the lowest household income category report driving an average of only 0.1 mph
higher than the safe driving speed on residential streets.
Figure 3-8: Reported Driving Speed and Perceived Safe Driving Speed
on Residential Streets by Household Income***
Driving Speed

Safe Driving Speed

28
27.6
27.2
27

26.8

26.7

27.4

27.3

27.4

26.7

Less than $30,000
$30,000 to
(N=1,244)
$49,999 (N=993)

$75,000 to
$50,000 to
$74,999 (N=1,092) $99,999 (N=755)

$100,000 or more
(N=1,113)

Driving speed and perceived safe limits on the three road types were further explored by the
consequences of speeding-related episodes experienced by drivers (See Figures 3-9 through 311). Drivers who have not been stopped for speeding by police in the past year drive at speeds
approximately equal to the perceived safe speed limit on multi-lane highways, two-lane
highways, and residential streets. Those who have been stopped for speeding within the past year
report normally traveling faster than their perceived safe speed limit would allow. Those who
have been stopped and received a ticket for speeding drive an average of 66.3 mph on multi-lane
highways, though, on average, they perceive the safe speed limit to be 65.1 mph on these roads.
This relationship reverses on residential or neighborhood streets, where those who had received a
ticket within the past year, on average, perceive the safe speed limit to be 28.2 mph, yet report
traveling at an average speed of 27.2 mph on the streets.
Figure 3-9: Reported Driving Speed and Perceived Safe Driving Speed
on Multi-Lane Divided Highways by Consequences***
Driving Speed

Safe Driving Speed

66.2
66

66.0

66.3 66.5
65.4

65.1
64.5

63.8

63.6

63.3

64.0

63.9

Not Stopped
(N=5,434)

Stopped by
Police (N=451)

Ticket (N=304) Warning (N=121)

No Accidents
(N=5,692)

One or More
Accidents
(N=196)

Figure 3-10: Reported Driving Speed and Perceived Safe Driving Speed
on Two-Lane Highways by Consequences
Driving Speed

Safe Driving Speed

52
51.0

51.3
50.8

51.0

50.9 51.0

50
49

48.7

48.9 49.1

48.8

47.9

48.2

47
46

Not Stopped
(N=5,421)

Stopped by Ticket (N=299)
Police (N=450)

Warning
(N=121)

No Accidents
(N=5,671)

One or More
Accidents
(N=203)

Figure 3-11: Reported Driving Speed and Perceived Safe Driving Speed
on Residential Streets by Consequences***
Driving Speed

Safe Driving Speed

29
28.2
27.9

28
27.2
27

27.2

27.4

27.2 27.3

27.2
26.9

26.9

27.3
26.7

Not Stopped
(N=5,576)

Stopped by Ticket (N=309)
Police (N=461)

Warning
(N=124)

No Accidents
(N=5,836)

One or More
Accidents
(N=203)

The perceived risk of a speeding ticket is considered a possible deterrent to speeding behavior.
Respondents were asked how many mph over the speed limit one could drive before receiving a
ticket. Figure 3-12 shows the mean mph drivers believe they can speed over the speed limit
without being pulled over for speeding by the police and broken out by the consequences of their
previous speed-related episodes (stops and crashes). Those who have been stopped by the police
within the past year and received a warning rather than a speeding ticket, on average, believe that
one can travel 10.6 mph over the limit on multi-lane highways and 11.4 mph over the limit on
two lane highways before receiving a speeding ticket (See Figure 3-12). This is a larger margin
than the average “allowable” speed over the speed limit reported by any other group.
Figure 3-12: Mean mph Over Limit Without Receiving a Ticket by Consequences***
Multi-lane Divided Highways

Two-lane Highways

Residential Streets

14
12
10
8

9.4

9.3

8.7
6.9

9.9
8.6
7.0

11.4
10.6
9.4

9.2

8.7
6.9

6.5

9.8

9.2
7.6

8.4
6.9

6.6

6.3

6
4
2
0

Total
(N=5,607)

No Stops
(N=5,162)

Stopped
(N=436)

No
Accidents
(N=5,411)

Accident
(N=491)

Ticket
(N=293)

Warning
(N=118)

Q5f/Q6f/Q7f. How far above the speed limit do you think the average driver can go on (ROAD TYPE) before he or she will
receive a ticket?
Base: Respondents Who Drive on ALL Road Types
Unweighted N=See Chart *** p < .001

CHAPTER 4

NORMS AND ATTITUDES ABOUT SPEEDING
Respondents were asked a series of questions pertaining to their attitudes toward speeding from
both normative and personal perspectives (See Figure 4-1). When asked whether everyone
should obey the speed limits because it’s the law, an overwhelming majority of drivers (91%)
agree either strongly or somewhat, with two thirds (67%) agreeing strongly with this statement.
There was also strong agreement that “It is unacceptable to exceed speed limits by more than 20
mph.” More than 17 out of 20 (87%) drivers agreed with this statement, with 76% agreeing
strongly. Drivers agreed that people should keep up with the flow of traffic with 52% of drivers
strongly agreeing and 30% somewhat agreeing with this statement. Approximately one-half of
drivers agree that speeding tickets have more to do with raising money than they do with
reducing speeding (51%) and that there is no excuse to exceed the speed limits (48%). Less than
one-fifth of drivers agreed with the statements, “If it is your time to die, you’ll die, so it doesn’t
matter whether you speed” (17%), and “Driving over the speed limit is not dangerous for skilled
drivers” (16%).
Figure 4-1: Normative Attitudes Regarding Speeding
100%

Somewhat Agree
Strongly Agree

80%

24%

11%
30%

60%
40%
20%
0%

67%

76%

22%

18%

29%

30%

52%
6%
11%

8f. It is
8b. People
8a. Everyone
should obey unacceptable should keep
to exceed
pace with the
the speed
limits because speed limits by flow of traffic.
more than
it’s the law.
20mph

8c. Speeding 8e. There is no
excuse to
tickets have
more to do
exceed the
with raising
speed limits
money than
they do with
reducing
speeding

9%
7%

8g. If it is your 8d. Driving
time to die, over the speed
you’ll die, so it limit is not
doesn’t matter dangerous for
whether you skilled drivers
speed

Q8. Now I’m going to read a few statements about driving and speed limits. After I read each one, please tell me whether you
agree, disagree or neither. (READ ITEM). Would you say you strongly (AGREE/DISAGREE) or somewhat
(AGREE/DISAGREE)?
Base: All Respondents
Unweighted N= 6,144

Table 4-1 shows the average rating of the normative attitude statements by driver age group,
gender, education level, household income and metro status. The ratings for each statement range
from 1 to 5 with 5=strongly agree and 1=strongly disagree, so that the higher the average value,
the more agreement there is in the group of drivers with that particular statement. There are no
large differences when these items are examined by demographics, although some nuances
become apparent. For example, older drivers are more likely than younger drivers to agree with
the statement “Everyone should obey the speed limits because it’s the law,” and less likely to
agree that “People should keep pace with the flow of traffic.” As household income level
increases, agreement with the statement that “Everyone should obey the speed limits because it’s
the law,” becomes less likely.
Table 4-1: Normative Attitudes Regarding Speeding by Demographics

Age
16-20
21-24
25-34
35-44
45-54
55-64
65 or older
Gender
Male
Female
Education
Less than HS
HS diploma
College degree
Graduate degree
2010 Household Income
< $30K
$30K - $50K
$50K - $75K
$75K - $100K
$100K or more
Metro Status
Urban
Non-urban

8b People should
keep pace with the
flow of traffic

8c Speeding
tickets have more
to do with raising
money than they
do with reducing
speeding

8d Driving over
the speed limit is
not dangerous for
skilled drivers

4.31
4.41
4.45
4.53
4.51
4.61
4.66

4.36
4.23
4.26
4.25
4.06
4.12
4.07

3.14
3.14
3.36
3.22
3.33
3.20
3.01

1.88
1.81
1.83
1.91
1.86
1.82
1.68

2,696
3,448

4.37
4.66

4.25
4.11

3.36
3.09

2.07
1.59

464
3,327
1,392
933

4.61
4.57
4.39
4.27

4.10
4.16
4.26
4.26

3.32
3.20
3.18
3.22

1.92
1.75
1.90
2.04

1,275
1,019
1,102
761
1,119

4.69
4.57
4.48
4.41
4.21

4.13
4.16
4.16
4.27
4.29

3.23
3.20
3.25
3.12
3.27

1.79
1.71
1.85
1.90
2.04

3,030
2,903

4.46
4.52

4.25
4.07

3.25
3.04

1.85
1.74

8a Everyone
should obey the
speed limits
because it’s the
law

295
281
939
835
1,185
1,211
1,328

Table 4-1: Normative Attitudes Regarding Speeding by Demographics (Continued)

8e There is no excuse
to exceed the speed
limits

8f It is unacceptable to
exceed speed limits by
more than 20mph

8g If it is your time to
die, you’ll die, so it
doesn’t matter whether
you speed

295
281
939
835
1,185
1,211
1,328

3.13
2.95
3.08
3.14
3.07
3.28
3.44

4.39
4.41
4.46
4.42
4.53
4.49
4.46

1.78
1.82
1.81
1.87
1.79
1.78
1.73

2,696
3,448

3.01
3.33

4.41
4.52

1.96
1.64

464
3,327
1,392
933

3.56
3.21
2.84
2.76

4.26
4.49
4.52
4.56

2.19
1.78
1.59
1.53

1,275
1,019
1,102
761
1,119

3.56
3.32
3.00
2.89
2.67

4.28
4.55
4.54
4.57
4.52

2.02
1.77
1.76
1.70
1.60

3,030
2,903

3.07
3.05

4.50
4.53

1.65
1.67

A different pattern emerges when normative attitudes are examined by driver type. While there is
strong agreement across demographic groups, as a whole, on the three highest rated normative
statements, there are significant differences between driver types regarding these same normative
statements. To clearly show the pattern of response by driver type, the percentages in Figure 4-2
are limited to the percentages who strongly agree with each statement. Less than one-half (48%)
of the people classified as speeders strongly agree that everyone should obey the speed limits
because it’s the law. Conversely, more than 4 out of 5 drivers classified as nonspeeders (81%)
strongly agree with this statement. Almost two-thirds (64%) of speeders strongly agree that
people should keep up with the flow of traffic; however, only 42% of the nonspeeders strongly
agree with this statement. One in six speeders (16%) strongly agrees that there is no excuse to
exceed the speed limit; however, more than two-fifths (41%) of nonspeeders strongly agree with
this statement. Although the number of respondents who strongly agree is smaller, speeders are
twice as likely (11%) as sometime speeders (5%) or nonspeeders (5%) to agree that driving over
the speed limit is not dangerous for skilled drivers.
Figure 4-2: Normative Attitudes Regarding Speeding by Driver Type
% Strongly Agree
100%

Speeders (N=1,572)

81%

80%

68%

60%
48%

40%

Sometime Speeders (N=2,148)

77%
70%
64%
50%
42%

41%
35%
27%
23%

20%
0%

Nonspeeders (N=1,579)

84%

29%
16%

8b. People
8c. Speeding
8a. Everyone
8f. It is
should obey unacceptable should keep tickets have
the speed
to exceed pace w ith the m ore to do
flow of
w ith raising
lim its
speed lim its
because it’s by m ore than
traffic***
m oney than
the law ***
20m ph***
they do w ith
reducing
speeding***

13%
9%10% 11%
5%5%

8g. If it is
8d. Driving
8e. There is
over the
no excuse to your tim e to
exceed the die, you’ll die, speed lim it is
so it doesn’t
not
speed
dangerous
lim its***
m atter
w hether you
for skilled
drivers***
speed***

When the normative statements are compared across age groups, the differences are not nearly as
pronounced as they were when compared across driver type. The percentages in Figure 4-3 are
again limited to the percentages who strongly agree with each statement. The youngest group (16
to 34) is less likely to strongly agree with the statement that everyone should obey the speed
limits because it’s the law (60%) when compared to the 35 to 54 age group (67%) or the 55 and
older age group (75%). The majority of those in the 16 to 34 age group (57%) strongly agree that
people should keep pace with the flow of traffic, while a slightly lower proportion (50%) of
those in the other two age groups strongly agree with this statement. Drivers older than 55 years
are most likely to strongly agree with the statement, there is no excuse to exceed the speed limits,
with more than one-third (36%) strongly agreeing. Among younger drivers, 28% of those 16 to
34 and 27% of those 35 to 54 strongly agree with this statement.
Figure 4-3: Normative Attitudes Regarding Speeding by Age Group
%Strongly Agree
100%

16-34 (N=1,515)

35-54 (N=2,020)

55 or older (N=2,540)

79%
77%
75% 73%

80%

67%

60%

57%
50%50%

40%

36%
30%31%
28%27%
26%

20%
0%

13%
11%
10%

8f. It is
8b. People
8a. Everyone
should obey unacceptable should keep
to exceed
pace with the
the speed
flow of
limits because speed limits by
more than
traffic***
it’s the law***
20mph***

8c. Speeding 8e. There is no
tickets have
excuse to
more to do
exceed the
with raising speed limits***
money than
they do with
reducing
speeding***

6% 7% 6%

8g. If it is your 8d. Driving
time to die, over the speed
you’ll die, so it limit is not
doesn’t matter dangerous for
skilled
whether you
drivers***
speed***

The respondents were also asked a series of questions that measured their personal feelings
toward speeding and speeding behaviors. As shown in Figure 4-4, 3 in 5 (60%) drivers agree that
they often get impatient with slower drivers. Close to half of all respondents (47%) agreed with
the statement “I worry a lot about having a crash.” There was considerably less agreement with
the statements “Speeding is something I do without thinking” (27%), “I enjoy the feeling of
driving fast” (27%), and “I try to get where I am going as fast as I can” (20%). Only 9% of
drivers agree with the statement, “I consider myself a risk taker while driving.”
Figure 4-4: Personal Attitudes Regarding Speeding

100%

Somewhat Agree
Strongly Agree

80%
60%
40%
20%
0%

36%

24%

21%
26%

20%

18%

c. I often get e. I worry a lot g. Speeding is a. I enjoy the
impatient with about having a something I do
feeling of
slower drivers
crash
without
driving fast
thinking

15%
5%

7%
8%

4%
5%

d. I try to get b. The faster I f. I consider
where I am drive, the more myself a risk
going as fast
alert I am
taker while
as I can
driving

Q9. Now I’m going to read a few statements. After I read each one, please tell me whether you agree, disagree or neither. (READ
ITEM). Would you say you strongly (AGREE/DISAGREE) or somewhat (AGREE/DISAGREE)?
Base: All Respondents
Unweighted N=6,144

Table 4-2 shows the mean rating of the personal attitude items by driver age group, gender,
education level, household income, and metro status. The ratings for each item range from 1 to 5
with 5=strongly agree and 1=strongly disagree, so that the higher the mean value the more
agreement there is with that particular statement. There were no large differences in these items
when examined by driver demographics, although some nuances become apparent. For example,
older drivers are much more likely than younger drivers and women are much more likely than
men to disagree with the statement, “I enjoy the feeling of driving fast.” Drivers with higher
household incomes are less likely to disagree with this statement than those with lower
household incomes.
Table 4-2: Personal Attitudes Regarding Speeding by Demographics

Age
16-20
21-24
25-34
35-44
45-54
55-64
65 or older
Gender
Male
Female
Education
Less than HS
HS diploma
College degree
Graduate degree
2010 Household
Income
< $30K
$30K - $50K
$50K - $75K
$75K - $100K
$100K or more
Metro Status
Urban
Non-urban

9a I enjoy the
feeling of driving
fast

9b The faster I
drive the more
alert I am

9c I often get
impatient with
slower drivers

9d I try to get
where I am going
as fast as I can

295
281
939
835
1,185
1,211
1,328

2.98
2.73
2.41
2.25
2.16
1.99
1.66

2.20
2.11
1.82
1.79
1.62
1.61
1.67

3.64
3.60
3.53
3.33
3.28
3.13
2.97

2.47
2.43
2.41
2.07
1.97
1.82
1.61

2,696
3,448

2.51
1.94

2.00
1.56

3.41
3.22

2.23
1.89

464
3,327
1,392
933

2.15
2.17
2.33
2.44

1.91
1.73
1.78
1.80

3.13
3.33
3.36
3.45

1.97
2.00
2.19
2.35

1,275
1,019
1,102
761
1,119

1.99
2.20
2.27
2.41
2.49

1.74
1.73
1.80
1.82
1.82

3.07
3.31
3.46
3.47
3.53

1.93
1.95
2.08
2.25
2.32

3,030
2,903

2.23
2.11

1.74
1.69

3.31
3.28

2.10
1.96

Table 4-2: Personal Attitudes Regarding Speeding by Demographics (Continued)

Age
16-20
21-24
25-34
35-44
45-54
55-64
65 or older
Gender
Male
Female
Education
Less than HS
HS diploma
College degree
Graduate degree
2010 Household
Income
< $30K
$30K - $50K
$50K - $75K
$75K - $100K
$100K or more
Metro Status
Urban
Non-urban

9e I worry a lot
about having a
crash

9f I consider
myself a risk taker
while driving

9g Speeding is
something I do
without thinking

295
281
939
835
1,185
1,211
1,328

3.40
3.40
3.13
3.25
2.89
2.72
2.64

1.85
1.81
1.59
1.71
1.41
1.34
1.32

2.60
2.46
2.44
2.21
2.08
1.85
1.74

2,696
3,448

2.86
3.13

1.65
1.42

2.27
2.03

464
3,327
1,392
933

3.36
3.00
2.80
2.71

1.98
1.44
1.44
1.45

2.32
2.08
2.16
2.20

1,275
1,019
1,102
761
1,119

3.26
3.00
2.85
2.89
2.80

1.73
1.51
1.45
1.41
1.46

2.05
2.06
2.27
2.18
2.26

3,030
2,903

2.87
2.88

1.44
1.40

2.06
2.15

Figure 4-5 shows the personal attitudes toward speeding by driver type. The percentages in
Figure 4-5 are limited to the percent who strongly agree with each statement. Drivers classified
as speeders are almost three times as likely to strongly agree with the statement, “I often get
impatient with slower drivers,” when compared to drivers classified as sometime speeders (45%
versus 18%, respectively). The same pattern is apparent for other items as well. Although the
percentages are lower, speeders are three times more likely to strongly agree with the statement,
“I enjoy the feeling of driving fast” (19%), compared to sometime speeders (6%). Speeders are
more than three times as likely to strongly agree with the statement “I try to get where I am
going as fast as I can,” compared to sometime speeders (11% and 3%, respectively). Conversely,
speeders are less concerned about crashes.
Figure 4-5: Personal Attitudes Regarding Speeding by Driver Type
% Strongly Agree
100%

Speeders (N=1,572)

Sometime Speeders (N=2,148)

Nonspeeders (N=1,579)

80%
60%
45%

40%
27%28%

20%

20%
18%
14%

19%
13%
6%

11%
4%

c. I often get e. I worry a lot g. Speeding is a. I enjoy the
impatient with about having a something I do
feeling of
slower
crash***
without
driving fast***
drivers***
thinking***

3% 2%

10%
7%

5% 5% 4% 4%

d. I try to get b. The faster I
f. I consider
where I am drive, the more myself a risk
going as fast
alert I am***
taker while
as I can***
driving***

Respondents were asked, on those occasions when you speed, what do you think are the main
reasons you drive faster than the speed limit. This was an open-ended question, i.e., choices were
not offered to the respondent, and each respondent gave his/her own reasons. More than one
reason could be given by the respondent. As seen in Figure 4-6, the most frequent reasons are
“I’m late” (35%) and emergency or illness (31%). One-tenth of drivers indicate that they were
not paying attention to how fast they were driving, 7% said they were in a hurry (but did not
elaborate further as to why), while another 7% said they were going with the flow of traffic and
8% of drivers indicated that they never drove faster than the speed limit.
Figure 4-6: Reasons for Speeding

Never Speed (VOL)

Traffic flow

In a hurry

Not paying attention

10%
31%

Emergency/Illness

35%

I'm Late

10%

20%

30%

40%

50%

Q10. People sometimes go faster than the speed limit for different reasons. On those occasions when you do, what do you think
are the main reasons you drive faster than the speed limit? Anything else? MULTIPLE RECORD. DO NOT READ.
Base: All Respondents
Unweighted N=6,144

Respondents were asked whether they agree or disagree with a series of statements regarding
their attitudes about driving at or near the speed limit. Some items were negative and suggested
problems with driving near the speed limit while other items were positive and suggested
benefits for driving near the speed limit. Over 4 out of 5 (82%) drivers agree with the statement,
“Driving at or near the speed limit makes it easier to avoid dangerous situations.” There was also
strong agreement with the statements, “Driving at or near the speed limit reduces my chances of
an accident” (79%) and “Driving at or near the speed limit uses less fuel” (73%). A little over 2
out of 5 drivers (42%) agree that driving at or near the speed limit makes it difficult to keep up
with traffic, and less than a fifth (17%) agree that driving at or near the speed limit makes them
feel annoyed (See Figure 4-7).
Figure 4-7: Attitudes Toward Driving at or Near the Speed Limit
100%

Somewhat Agree
Strongly Agree

80%
25%

23%

57%

56%

60%

22%

40%
20%
0%

51%

28%
14%

e. Makes it easier a. Reduces my
to avoid
chances of an
dangerous
accident
situations

f. Uses less fuel

11%
6%

b. Makes it
c. Makes me feel
difficult to keep
annoyed
up with traffic

Q11. Now I’m going to read a few statements. After I read each one, please tell me whether you agree, disagree, or neither.
(READ ITEM). Would you say you strongly (AGREE/DISAGREE) or somewhat (AGREE/DISAGREE)? Driving at or near the
speed limit . . .
Base: All Respondents
Unweighted N=6,144

Table 4-3 shows attitudes toward driving at or near the speed limit by driver age group, gender,
education level, household income and metro status. For each item, a score of 5=strongly agree
and a score of 1=strongly disagree. The higher the mean value, the more agreement there is with
that particular statement. The differences in the scores for these items across demographic
subgroups are not large, although some nuances are apparent. Older drivers are less likely to feel
annoyed about driving at or below the speed limit, and older drivers are more likely to believe
driving at the speed limit uses less fuel. There is not a lot of difference across age groups in that
most drivers feel that driving at or near the speed limit reduces their chance for an accident. The
same holds true across gender, education, and household income level.
Table 4-3: Attitudes Regarding Driving at or Near the Speed Limit by Demographics
Driving at or near the
speed limit . . .

Age
16-20
21-24
25-34
35-44
45-54
55-64
65 or older
Gender
Male
Female
Education
Less than HS
HS diploma
College degree
Graduate degree
2010 Household
Income
< $30K
$30K - $50K
$50K - $75K
$75K - $100K
$100K or more
Metro Status
Urban
Non-urban

11a Reduces
my chances of
an accident

11b Makes it
difficult to
keep up with
traffic

11c Makes me
feel annoyed

11e Makes it
easier to avoid
dangerous
situations

11f Uses less
fuel

295
281
939
835
1,185
1,211
1,328

4.14
4.14
4.11
4.06
4.09
4.14
4.12

2.69
2.77
2.67
2.86
2.93
2.87
2.78

2.40
2.24
2.04
2.00
1.80
1.71
1.66

4.21
4.17
4.18
4.16
4.17
4.21
4.03

3.68
3.68
3.87
4.00
4.13
4.27
4.05

2,696
3,448

4.04
4.17

2.97
2.66

2.05
1.80

4.02
4.28

3.98
4.03

464
3,327
1,392
933

4.10
4.11
4.14
4.09

2.78
2.78
2.91
2.90

2.05
1.86
1.95
2.04

4.20
4.18
4.09
4.05

3.84
4.02
4.09
4.09

1,275
1,019
1,102
761
1,119

4.15
4.15
4.19
4.09
3.99

2.61
2.75
2.84
2.99
3.09

1.87
1.84
2.00
1.99
2.04

4.20
4.17
4.23
4.14
4.04

3.87
4.04
4.08
4.12
4.10

3,030
2,903

4.09
4.15

2.90
2.79

1.92
1.83

4.13
4.17

4.04
4.15

A clear pattern emerges when the personal statements regarding driving at or near the speed limit
are examined by driver type. The speeders are less likely to agree with the sometime speeders or
nonspeeders on the positive aspects of driving at or near the speed limit and more likely to agree
with the negative aspects (See Figure 4-8). Slightly less than one-half (47%) of the speeders
strongly agree that driving at or near the speed limit makes it easier to avoid a dangerous
situation while the majority of sometime speeders (57%) and nonspeeders (67%) strongly agree
with this statement. Conversely, speeders are more likely to feel annoyed (10%) about driving at
or near the speed limit when compared to sometime speeders (3%) or nonspeeders (3%).
Figure 4-8: Attitudes Toward Driving at or Near the Speed Limit by Driver Type
% Strongly Agree
100%

Speeders (N=1,572)

80%
60%

67%
57%
47%

Sometime Speeders (N=2,148)

63%
57%
47%

Nonspeeders (N=1,579)

57%
52%
45%

40%
16%13%
13% 10%

20%
0%

e. Makes it
easier to avoid
dangerous
situations***

a. Reduces my
chances of an
accident***

f. Uses less
fuel***

3% 3%

b. Makes it
c. Makes me
difficult to keep feel annoyed***
up with traffic***

CHAPTER 5
ATTITUDES TOWARD ENFORCEMENT AND SPEEDING
COUNTERMEASURES
Attitudes toward speeding enforcement and various countermeasures designed to discourage
speeding are examined in this chapter. These countermeasures include such items as more
frequent ticketing, photo enforcement (discussed in more detail in Chapter 6), new in-vehicle
technologies that alert the driver when he or she is speeding as well as speed governors, invehicle devices that limit the speed at which a vehicle can travel. Respondents were first asked
about the importance that something be done to reduce speeding by drivers. The results shown in
Figure 5-1 indicate that close to one-half (48%) of drivers believe that this was very important,
and 39% believe that it is somewhat important that something be done to reduce speeding. Only
8% of drivers say that it was not too important and 3% state that it was not at all important that
something be done to reduce speeding.
Figure 5-1: Importance Something Be Done to Reduce Speeding
100%
80%
60%

48%
39%

40%
20%
0%

8%
Very important

Somewhat
important

Not too
Important

Not at all
Important

Not sure

Q12. How important is it that something be done to reduce speeding by drivers? Is it . . .
Base: All Respondents
Unweighted N=6,144

Table 5-1 shows the distribution of the level of importance placed on something being done to
reduce speeding by drivers’ age group, gender, education level, and household income. There
were no large differences when these items are examined by demographics, although some
nuances do become apparent. For example, a majority of older drivers (55 and above) and
women indicate that it is very important that something be done to reduce speeding, while the
importance placed on reducing speeding decreases as formal education and household income
increase.
Table 5-1: Importance That Something Be Done to Reduce Speeding by Demographics
12. How important is it that
something be done to reduce
speeding by drivers?
Age***
16-20
21-24
25-34
35-44
45-54
55-64
65 or older
Gender***
Male
Female
Education***
Less than HS
HS diploma
College degree
Graduate degree
2010 Household Income***
< $30K
$30K - $50K
$50K - $75K
$75K - $100K
$100K or more
Metro Status
Urban
Non-urban

Very important

Somewhat
important

Not too
important

Not at all
important

295
281
939
835
1,185
1,211
1,328

41.7%
44.3%
44.2%
45.8%
45.7%
55.9%
58.0%

49.2%
43.4%
41.0%
41.5%
40.3%
34.2%
32.8%

5.4%
10.7%
11.1%
9.1%
9.9%
6.4%
5.2%

3.6%
1.5%
3.2%
2.5%
3.2%
2.4%
2.4%

2,696
3,448

40.3%
56.1%

42.5%
36.6%

11.7%
5.4%

4.3%
1.3%

464
3,327
1,392
933

61.3%
50.0%
36.3%
36.1%

29.5%
40.1%
45.7%
44.2%

4.3%
7.3%
13.4%
13.3%

4.0%
1.7%
3.4%
5.7%

1,275
1,019
1,102
761
1,119

62.8%
50.1%
42.9%
38.7%
34.6%

30.8%
40.3%
43.3%
45.5%
45.4%

4.5%
5.9%
9.7%
11.9%
14.3%

1.5%
2.3%
3.5%
3.4%
4.8%

3,030
2,903

48.2%
48.6%

39.6%
40.0%

8.4%
8.4%

3.1%
2.2%

*** p<.001

Figure 5-2 compares the importance placed on reducing speeding by driver age group. Older
drivers (57%) are more likely than younger drivers to say that it is very important that something
be done to reduce speeding. Only 44% of drivers in the 16 to 34 age group and 46% of those in
the 35 to 54 age group indicated that reducing speeding is very important. Drivers in the 16 to 34
age group (44%) and the 35 to 54 age group (41%) are more likely to think reducing speeding is
somewhat important when compared to the 55 and older group (33%).
Figure 5-2: Importance Something Be Done to Reduce Speeding by Age Group***
100%

16-34 (N=1,515)

35-54 (N=2,020)

55 or older (N=2,540)

80%
60%
40%

57%
44%46%

44%
41%
33%

20%
0%

10%10%

Very important

Somewhat
important

Not too
Important

3% 3% 2%

0% 1% 1%

Not at all
Important

Not sure

Q12. How important is it that something be done to reduce speeding by drivers? Is it . . .
Base: All Respondents
Unweighted N=6,144 ***p<.001

The importance placed on whether something should be done to reduce speeding varies by driver
type. Figure 5-3 shows that less than one-third of people classified as speeders (30%) think that
reducing speeding is very important, while 49% of those classified as sometime speeders and
61% of drivers classified as nonspeeders believe that it is very important. Speeders are more
likely to think that reducing speeding is somewhat important (48%) compared to sometime
speeders (43%) or nonspeeders (32%). Speeders are also more likely to say that reducing
speeding is not at all important (6%) compared to sometime speeders (1%) or nonspeeders (2%).
Figure 5-3: Importance Something Be Done to Reduce Speeding by Driver Type***
100%

Speeders (N=1,572)

Sometime Speeders (N=2,148)

Nonspeeders (N=1,579)

80%
60%
40%

61%
49%
30%

48%
43%
32%

20%

16%
7%

Very important

Somewhat
important

Not too
Important

1% 2%

Not at all
Important

Q12. How important is it that something be done to reduce speeding by drivers? Is it . . .
Base: Respondents Assigned a Driver Type
Unweighted N=See Chart ***p<.001

0% 1% 1%
Not sure

Respondents were asked how often they thought police officers should enforce the speed limit.
Figure 5-4 shows that close to one-half (48%) of drivers think that the speed limit should be
enforced all of the time. Almost one-third (30%) say it should be enforced often and 18% say it
should be enforced sometimes. Respondents were also asked how often they see motor vehicles
pulled over by police on the streets and roads they normally drive. One in seven drivers (13%)
indicate that they see motor vehicles pulled over all the time. Three in ten (30%) see vehicles
pulled over often and 40% report that they sometimes see vehicles pulled over. Interestingly,
16% of drivers report that they rarely see vehicles pulled over, which is a higher percentage of
drivers than those that say that they see vehicles pulled over all the time.
Figure 5-4: Preferred and Perceived Enforcement of Speed Limits

Rarely 2% Never 1% Not Sure 1%

Never 1%

Rarely
16%

Sometimes
18%
All the time
48%

All the time
13%

Often
30%
Sometimes
40%

Often
30%

Q13. How often police should
enforce speed limit

Q14. How often see vehicles
pulled over on road

Q13. How often do you think police should enforce the speed limit? Should they enforce it . . .
Q14. How often do you see motor vehicles that have been pulled over by police on the streets and roads you normally drive? Do
you see motor vehicles pulled over . . .
Base: All Respondents
Unweighted N=6,144

Table 5-2 shows the distribution of responses to the question of how often police should enforce
the speed limit by demographics. In general, an attitude in support of a higher frequency of
enforcement was associated with older age, female gender, decreased formal education, and
lower household income.
Table 5-2: Frequency That Speed Limit Should Be Enforced by Demographics
13. How often do you think

police should enforce the speed
limit? Should they enforce it . . .

Age***
16-20
21-24
25-34
35-44
45-54
55-64
65 or older
Gender***
Male
Female
Education***
Less than HS
HS diploma
College degree
Graduate degree
2010 Household Income***
< $30K
$30K - $50K
$50K - $75K
$75K - $100K
$100K or more
Metro Status
Urban
Non-urban

*** p<.001

All the time

Often

Sometimes

Rarely

Never

295
281
939
835
1,185
1,211
1,328

33.5%
42.4%
48.3%
46.1%
46.7%
52.2%
56.1%

39.3%
31.8%
28.0%
30.5%
29.5%
29.6%
28.1%

20.0%
20.3%
19.3%
19.1%
19.5%
15.0%
13.4%

4.4%
3.2%
3.0%
3.3%
2.1%
1.7%
1.0%

2.4%
1.2%
0.7%
0.9%
0.3%
0.3%
0.1%

2,696
3,448

41.9%
53.1%

30.4%
30.1%

22.3%
13.8%

3.5%
1.6%

1.1%
0.3%

464
3,327
1,392
933

54.8%
49.9%
38.4%
37.2%

26.3%
29.5%
36.3%
32.2%

14.5%
16.6%
21.5%
25.6%

2.7%
2.4%
2.4%
3.1%

1.4%
0.4%
0.5%
1.0%

1,275
1,019
1,102
761
1,119

57.3%
53.8%
42.3%
40.5%
36.9%

25.0%
28.8%
35.2%
36.2%
30.5%

14.1%
14.0%
18.5%
19.0%
26.7%

2.7%
2.0%
1.9%
3.1%
3.6%

0.3%
0.6%
1.2%
0.6%
0.8%

3,030
2,903

46.1%
49.6%

30.4%
30.6%

18.6%
16.6%

3.4%
1.6%

0.6%
0.7%

Table 5-3 shows the distribution by demographics of responses to the question of how often
respondents see vehicles pulled over by police on the side of the road. No large differences in the
frequency of seeing vehicles pulled over are apparent across the various demographic categories
of drivers.
Table 5-3: Frequency of Seeing Vehicles Pulled Over on Side of Road by Demographics

14. How often do you see motor
vehicles that have been pulled
over by police on the streets and
roads you normally drive?

*** p < .001

All the time

Often

Sometimes

295
281
939
835
1,185
1,211
1,329

11.9%
14.5%
17.8%
16.4%
14.2%
7.9%
5.0%

33.1%
33.3%
31.5%
32.8%
29.0%
30.1%
25.5%

37.9%
39.2%
36.7%
35.8%
38.6%
44.1%
45.0%

2,696
3,448

14.5%
10.6%

31.9%
28.9%

37.5%
41.7%

15.2%
17.0%

0.7%
1.5%

464
3,327
1,392
933

12.7%
13.3%
11.8%
9.5%

32.1%
30.1%
30.0%
28.8%

37.4%
40.2%
38.9%
41.7%

15.4%
15.2%
18.3%
19.3%

2.4%
1.0%
0.7%
0.6%

1,275
1,019
1,102
761
1,119

14.1%
10.6%
13.6%
12.9%
12.0%

29.7%
34.3%
27.4%
29.8%
31.8%

38.9%
39.2%
40.1%
41.1%
39.6%

15.0%
15.0%
18.1%
15.4%
15.4%

2.0%
0.5%
0.9%
0.7%
0.9%

3,030
2,903

14.0%
10.8%

30.7%
30.7%

39.0%
40.2%

15.0%
17.0%

1.2%
1.1%

Rarely
14.5%
12.9%
13.0%
14.4%
16.7%
16.8%
22.3%

Never
2.6%
0.1%
1.0%
0.6%
1.2%
0.7%
1.8%

Drivers’ attitudes toward countermeasures intended to reduce speeding were explored next.
Examples of speed reduction countermeasures were read to respondents, who were then asked
whether implementing each countermeasure in their community was either a good idea or bad
idea. Figure 5-5 shows the percentage of drivers who indicated that implementing a particular
countermeasure in their community was a good idea. Of the countermeasures offered, the two
with the highest rating were electronic signs by the road that warn drivers that they are speeding
and should slow down (89%) and increasing public awareness of the risks of speeding (88%). It
should be noted that both of these items did not include any specific penalties to drivers. Four out
of five drivers (80%) think that increased use of speed cameras in dangerous or high-crash
locations is a good idea, and two-thirds (66%) think that more frequent ticketing for speeding is a
good idea. Two out of five (41%) respondents thought issuing higher fines for speeding tickets is
a good idea.
Figure 5-5: Attitudes Toward Using Countermeasures in Community (% Good Idea)
100%
80%

89%

88%
80%
66%

60%

62%
41%

40%
20%
0%

20e.
20c.
20f.
Electronic
Increasing
Increased
signs that tell awareness of use of speed
motorists
speeding
cameras in
they are
risks
dangerous
speeding
locations

20a. More
frequent
ticketing for
speeding

20d. Road
design
changes

20b. Issuing
higher fines
for speeding

Q20. How would you feel about using the following measures in your community to reduce speeding?
Base: All Respondents
Unweighted N=6,144

Table 5-4 shows the percentage of drivers in each demographic category who think that
implementing a specific countermeasure in their community is a good idea. In general, a larger
proportion of female drivers agree that countermeasures are a good idea when compared with
male drivers. Education and household income tend to be negatively associated with the
percentage of drivers who think that countermeasures are a good idea. No clear pattern emerges
by age group, except that older drivers are more likely than younger drivers to indicate that more
frequent ticketing for speeding is a good idea.
Table 5-4: Percentage of Drivers Indicating Countermeasures Are Good Idea by
Demographics
Please tell me whether you think
each of the following is a good idea
or a bad idea.

*** p < .001

20a More
frequent ticketing
for speeding***

20b Issuing
higher fines for
speeding***

20c Increasing
public awareness
of speeding
risks***

295
281
939
835
1,185
1,211
1,328

54.4%
57.8%
62.5%
67.5%
66.7%
71.1%
71.4%

32.4%
37.7%
40.2%
46.0%
41.2%
42.0%
44.9%

94.3%
92.0%
88.5%
85.3%
87.2%
90.2%
87.6%

2,696
3,448

62.8%
68.4%

39.5%
43.2%

86.5%
90.3%

464
3,327
1,392
933

71.3%
66.0%
62.3%
60.3%

51.4%
40.9%
36.9%
35.0%

87.9%
89.8%
85.9%
86.6%

1,275
1,019
1,102
761
1,119

71.2%
65.7%
66.6%
61.2%
60.7%

49.9%
39.4%
38.4%
37.1%
38.8%

89.4%
89.3%
89.3%
87.2%
86.0%

3,030
2,903

64.3%
67.2%

42.9%
39.7%

88.7%
89.2%

Table 5-4: Percentage of Respondents Who Indicated Countermeasures Are a Good Idea
by Demographics (Continued)
Please tell me whether you think
each of the following is a good idea
or a bad idea.

*** p < .001

20d Road design
changes***

20e Electronic signs
that tell motorists
they are speeding***

20f Increased use of
speed cameras in
dangerous
locations***

295
281
939
835
1,185
1,211
1,328

57.7%
61.7%
63.6%
67.5%
59.7%
61.9%
58.9%

86.0%
86.1%
85.1%
89.9%
89.7%
92.6%
93.7%

85.6%
78.9%
79.8%
78.7%
78.3%
78.7%
82.8%

2,696
3,448

57.3%
66.1%

87.9%
90.8%

73.5%
85.8%

464
3,327
1,392
933

62.7%
61.8%
62.0%
61.2%

90.1%
90.1%
87.5%
87.2%

87.0%
80.5%
75.1%
72.8%

1,275
1,019
1,102
761
1,119

67.9%
64.8%
57.9%
57.0%
58.0%

92.7%
90.3%
90.0%
86.7%
87.3%

88.2%
80.9%
79.7%
76.7%
70.7%

3,030
2,903

62.6%
60.6%

89.6%
89.1%

79.8%
80.3%

There was high agreement among the three driver types that implementing electronic signs that
warn drivers they are speeding and increasing public awareness of the risks of speeding are both
good ideas. However, differences between the three driver types become apparent when the
countermeasure includes increased penalties for speeding. Figure 5-6 presents the percentage of
drivers in each driver type who indicated that a particular countermeasure is a good idea.
Although over one-half of the drivers classified as speeders (54%) think it’s a good idea to
increase the frequency of ticketing for speeding, they are not as inclined as the drivers classified
as sometime speeders (65%) or those classified as nonspeeders (78%) to indicate that this is a
good idea. Similarly, speeders (32%) were less likely to think that higher fines for speeding
tickets are a good idea compared to sometime speeders (39%) and nonspeeders (51%).
Figure 5-6: Attitudes Toward Using Countermeasures in Community
(% Good Idea) by Driver Type

100%
80%

Speeders (N=1,572)

90%91%
90%90%
87%
86%

Sometime Speeders (N=2,148)

84%
81%
74%

78%
65%

60%

Nonspeeders (N=1,579)

54%

64%61%
59%
51%
39%
32%

40%
20%
0%

20e. Electronic 20c. Increasing 20f. Increased
signs that tell awareness of use of speed
speeding
cameras in
motorists they
dangerous
are
risks***
locations***
speeding***

20a. More
frequent
ticketing for
speeding***

20d. Road
design
changes***

20b. Issuing
higher fines for
speeding***

Q20. How would you feel about using the following measures in your community to reduce speeding?
Base: Respondents Assigned a Driver Type
Unweighted N=See Chart *** p < .001

Respondents were asked whether they thought the use of speed governors was a good or bad
idea. The responses varied based on the type of driver that would be required to use this type of
device. As shown in Figure 5-7, 3 out of 5 drivers (60%) think that mandating use of speed
governors by truck drivers is a good idea. An even higher proportion of drivers support
mandatory use of speed governors for drivers under 18 (77%) and drivers with multiple speeding
tickets (82%). However, less than a quarter (24%) of drivers think that mandatory speed
governors for all drivers is a good idea.
Figure 5-7: Use of Speed Governor (% Good Idea)
100%

60%

82%

77%

80%
60%

40%
24%

20%
0%

21a. Truck Drivers

21b. Drivers 18 yrs or
younger

21c. Drivers with
multiple speeding
tickets

21d. All drivers

Q21. A speed governor is a device which does not allow the vehicle to go above a certain speed. Do you think the mandatory use
of a speed governor is a good idea or a bad idea for . . . ?
Base: All Respondents
Unweighted N=6,144

Table 5-5 presents the percentage of respondents in each demographic group who think that
speed governors are a good idea. Overall, women were more likely than men to agree that speed
governors are a good idea. Agreement with the statement that speed governors are a good idea
decreased with household income and formal education. Support for speed governors varied
across the age groups, with a larger percentage of older drivers indicating that speed governors
are a good idea for young drivers. However, older drivers are less likely than younger drivers to
agree that speed governors are a good idea for truck drivers.
Table 5-5: Percentage of Drivers Indicating Speed Governors Are a Good Idea by
Demographics
Do you think the mandatory use of
a speed governor is a good idea or a
bad idea for…

*** p < .001

21a Truck
drivers***

21b Drivers 18
years or
younger***

21c Drivers
with multiple
speeding
tickets***

21d All
drivers***

295
281
939
835
1,185
1,211
1,328

66.6%
69.6%
68.3%
63.8%
54.7%
54.2%
51.7%

68.0%
78.9%
78.6%
79.4%
79.4%
77.2%
73.6%

82.6%
81.5%
80.9%
81.9%
82.4%
83.7%
82.7%

29.8%
30.2%
25.8%
21.5%
19.7%
23.6%
27.4%

2,696
3,448

53.1%
66.3%

70.4%
82.9%

77.1%
86.7%

19.0%
29.4%

464
3,327
1,392
933

68.2%
60.8%
53.1%
52.0%

77.2%
78.8%
73.3%
72.1%

87.0%
83.4%
77.3%
74.6%

38.6%
24.4%
16.1%
12.6%

1,275
1,019
1,102
761
1,119

72.2%
62.2%
54.8%
50.8%
53.0%

82.1%
79.1%
76.3%
76.9%
70.1%

86.1%
85.2%
84.1%
78.7%
74.6%

38.4%
23.5%
17.1%
16.6%
15.1%

3,030
2,903

62.1%
58.0%

77.1%
76.8%

81.0%
83.5%

24.8%
23.4%

There were only slight differences across driver types regarding the mandatory use of speed
governors for specific populations. Figure 5-8 shows that speeders were slightly less likely than
sometime speeders or nonspeeders to think this was a good idea. However, when the use of
speed governors is mandatory for all drivers, the differences become more apparent with 15% of
speeders saying this is a good idea, compared to 23% of sometime speeders and 29% of
nonspeeders.
Figure 5-8: Use of Speed Governor by Driver Type (% Good Idea)

100%

Speeders (N=1,572)

80%
60%

61%
56%59%

Sometime Speeders (N=2,148)

79%80% 76%
71%

Nonspeeders (N=1,579)

84%86%

40%
23%
15%

20%
0%

21a. Truck
Drivers***

21b. Drivers 18 21c. Drivers with
yrs or
multiple
speeding
younger***
tickets***

29%

21d. All
drivers***

Respondents were presented with three in-vehicle speeding countermeasures and asked whether
placing each in their vehicle was a good or bad idea. Regardless of their response, they were then
asked whether the countermeasure would prevent them from speeding. Figure 5-9 displays the
percentage of drivers who think that a particular countermeasure is a good idea and whether it
would prevent them from speeding. Approximately 3 out of 5 drivers think that each of the
countermeasures is a good idea. A device in the motor vehicle that notifies you if you are
speeding was endorsed by 61% of drivers, a device that records the speed data and reports it to
the insurance company to lower premiums was endorsed by 62% of drivers, and a device that
slows down the vehicle when it senses another car or object is too close to the vehicle was
endorsed by 60% of drivers. For each in-vehicle countermeasure, approximately the same
percentage who thought it was a good idea also stated that it would prevent them from speeding.
The largest difference between the percentage of drivers who thought it was a good idea (60%)
and those that indicated it would keep them from speeding (54%) was indicated for a device that
slows down the vehicle if it senses an object is too close would prevent them from speeding.
Figure 5-9: Use of In-Vehicle Speeding Countermeasures

100%

Good Idea

80%
60%

61%

59%

62%

65%

Prevent from Speeding

60%

54%

40%
20%
0%

22a. Notification in
vehicle if you exceed
the speed limit

22b. Records speed
data and lets you
provide info to
insurance company

22c. Slows vehicle
down when senses
another car or object is
too close

Q22. Please tell me whether you think each of the following is a good idea or a bad idea to help reduce speeding. A device in
your motor vehicle that . . .
Q22a. Would it prevent you from speeding?
Base: All Respondents
Unweighted N=6,144

Table 5-6 presents the percentage of drivers in each demographic group who think that each invehicle speeding countermeasure is a good idea and also the percentage of drivers who say that
this device would prevent them from speeding. Female drivers were more likely than male
drivers to agree that countermeasures were a good idea, and also to indicate that the
countermeasures would prevent them from speeding. Increasing education and household
income were negatively associated with agreement that the countermeasures were a good idea
and that they would prevent speeding. No clear patterns emerged across age groups.
Table 5-6: Percentage of Drivers Indicating Speeding Countermeasures Are a Good Idea
And Would Prevent Them From Speeding by Demographics
Please tell me whether you think
each of the following is a good idea
or bad idea to help reduce
speeding…

*** p < .001

22a
Notification
in vehicle if
you exceed
the speed
limit***

22a Prevent
from
speeding***

22b Records
speed data
and lets you
provide info
to insurance
company***

22b Prevent
from
speeding***

295
281
939
835
1,185
1,211
1,328

66.5%
62.6%
56.3%
55.7%
57.7%
66.4%
70.1%

62.5%
62.5%
52.1%
54.6%
56.9%
64.4%
67.1%

74.1%
72.4%
66.3%
58.9%
57.1%
57.3%
61.4%

77.6%
78.7%
67.0%
62.1%
59.8%
61.4%
63.0%

2,696
3,448

56.2%
66.3%

52.6%
65.4%

57.7%
65.8%

58.9%
70.2%

464
3,327
1,392
933

76.2%
61.7%
50.1%
52.1%

71.1%
60.1%
49.2%
49.6%

74.8%
62.6%
53.0%
50.5%

78.3%
65.7%
52.2%
56.0%

1,275
1,019
1,102
761
1,119

72.8%
64.2%
57.8%
52.0%
49.7%

70.2%
60.9%
55.4%
48.8%
47.8%

75.6%
63.1%
60.7%
54.5%
50.3%

77.1%
64.4%
64.0%
56.8%
55.6%

3,030
2,903

62.1%
61.1%

59.3%
59.0%

61.3%
63.3%

64.6%
65.7%

Table 5-6: Percentage of Respondents Who Indicated Speeding Countermeasures Are a
Good Idea and Whether They Would Prevent Them From Speeding by Demographics
(Continued)
Please tell me whether you think
each of the following is a good idea
or bad idea to help reduce
speeding…

*** p < .001

22c Slows vehicle
down when senses
another car or
object too close***

22c Prevent from
speeding***

295
281
939
835
1,185
1,211
1,328

68.0%
54.8%
57.3%
56.3%
56.1%
58.7%
69.0%

65.9%
58.5%
50.2%
54.9%
49.8%
53.8%
58.2%

2,696
3,448

58.9%
60.4%

49.0%
59.6%

464
3,327
1,392
933

70.0%
58.6%
54.6%
55.8%

67.8%
55.0%
45.2%
44.0%

1,275
1,019
1,102
761
1,119

68.0%
61.3%
53.9%
51.6%
55.3%

67.2%
56.5%
51.1%
43.4%
44.5%

3,030
2,903

60.5%
58.0%

54.1%
54.7%

There was a distinct pattern of opinions by driver type about whether the countermeasures would
prevent speeding. As shown in Figure 5-10, less than one-half of drivers classified as speeders
(43%) reported that a speeding notification inside the car would prevent them from speeding, in
contrast to 62% and 69% of drivers who were classified as sometime speeders and nonspeeders,
respectively. The largest effect on speeding prevention came from the device that records
speeding information and reports it to the insurance company to lower the premiums if speed
limits are obeyed. Among speeders, 54% indicated that it would prevent them from speeding.
Among the other driver types, this percentage was 65% for sometime speeders, and 73% for
nonspeeders. Finally, only 45% of speeders say a device in their vehicle that slows the vehicle
down if an object gets too close would prevent them from speeding, compared to 56% of
sometime speeders and 60% of nonspeeders.
Figure 5-10: Percentage of Drivers Indicating Countermeasures Would Prevent Speeding
by Driver Type

100%

Speeders (N=1,572)

80%
62%

60%
40%

Sometime Speeders (N=2,148)

69%

65%

Nonspeeders (N=1,579)

73%
56%

54%

60%

45%

43%

20%
0%

22a. Notification in
vehicle if you exceed
the speed limit***

22b. Records speed
22c. Slows vehicle
data and lets you
down when sense
provide info to
another car or object
insurance company***
is too close***

Q22a. Would it prevent you from speeding?
Base: Respondents Assigned a Driver Type
Unweighted N=See Chart *** p < .001

Respondents were asked about the likelihood that they would use specific speeding
countermeasure devices in their own vehicle. As shown in Figure 5-11, close to one-half (48%)
of drivers state that they would be very likely or somewhat likely to use a device that limited the
speed of the vehicle to 10 mph over the posted speed limit. The majority of drivers (56%) say
that they would be very or somewhat likely to use a device that can be turned on or off, and
prevents the driver from driving faster than the speed limit. More than 4 out of 5 drivers (81%)
indicate that they would be very or somewhat likely to use an in-vehicle device that allows
parents to limit the maximum speed of the vehicle when a teenager drives the motor vehicle.
Figure 5-11: Likelihood of Using Countermeasure in Own Vehicle
100%
Somewhat likely
Very likely

80%
19%

60%
40%

29%

24%

62%

20%

27%

24%

23a. Does not allow you to 23b. You can turn on or off 23c. Allows parents to limit
drive 10 MPH over the speed which prevents you from
the maximum speed of
limit
driving over the speed limit vehicle when teenager is
driving

Q23. Now I’m going to read a few statements. After I read each one, please tell me whether you would be likely, unlikely, or
neither to use the following devices in your own vehicle(s). (READ ITEM). Would you say you would be very
(LIKELY/UNLIKELY) or somewhat (LIKELY/UNLIKELY) to use this device?
Base: All Respondents
Unweighted N=6,144

Respondents were asked whether they thought that the use of signs that change the speed limit on
a section of road based on traffic or weather conditions was a good idea or a bad idea. Figure 512 shows the percentage of drivers who believe that these signs are a good idea by situation.
Overwhelmingly, drivers indicate that these signs are a good idea when used for construction
zones (95%), school zones (96%), bad weather (93%), and congested roadways (89%).
Figure 5-12: Support Use of Digital Signs to Adjust Speed Limit (% Good Idea)
100%

95%

96%

93%

89%

80%
60%
40%
20%
0%

24a. Construction
Zones

24b. School zones

24c. Bad weather

24d. Congested
Roadways

Q24. Some roadways use digital signs to change the speed limit on a section of road based on traffic or weather conditions. Do
you think it is a good idea or a bad idea to use these signs in the following situations?
Base: All Respondents
Unweighted N=6,144

Table 5-7 presents the percentage of drivers in each demographic group who indicate that digital
variable speed signs are a good idea in various conditions. In general, there is high agreement
that these signs are a good idea. Women are more likely than men to indicate that signs are a
good idea in various situations. No clear pattern emerges across age groups, except that older
drivers are more likely to indicate that signs are a good idea on congested highways. The
percentage of drivers indicating that signs are a good idea for bad weather and congested
highways seems to be negatively correlated with education and household income.
Table 5-7: Percentage of Drivers Indicating Digital Variable Speed Signs Are a Good Idea
In Various Situations by Demographics
Do you think it is a good idea or a
bad idea to use these signs in the
following situations?

*** p < .001

24a
Construction
zones***

24b School
zones***

24c Bad
weather***

295
281
939
835
1,185
1,211
1,328

95.3%
95.7%
95.2%
93.4%
95.3%
96.8%
95.8%

94.6%
94.8%
97.3%
94.4%
96.3%
97.7%
97.0%

90.9%
93.3%
92.7%
90.1%
93.0%
96.0%
94.2%

82.3%
87.4%
86.7%
87.3%
89.8%
92.5%
92.3%

2,696
3,448

94.8%
95.7%

95.5%
96.7%

90.6%
95.1%

86.1%
91.1%

464
3,327
1,392
933

93.9%
95.2%
96.5%
96.1%

96.1%
96.0%
96.5%
97.3%

93.7%
93.3%
92.4%
90.6%

90.9%
89.2%
86.2%
87.1%

1,275
1,019
1,102
761
1,119

94.7%
96.7%
96.7%
95.8%
94.1%

97.3%
97.5%
95.2%
96.4%
95.8%

94.9%
94.8%
93.5%
90.4%
90.8%

92.8%
90.6%
89.4%
85.3%
84.4%

3,030
2,903

95.7%
95.0%

96.7%
95.6%

93.0%
93.0%

89.1%
88.5%

24d Congested
roadways***

CHAPTER 6
AUTOMATED PHOTO ENFORCEMENT DEVICES
Drivers’ awareness, beliefs, and perceptions of the usefulness of automated speed enforcement
cameras as well as their experiences with these devices are presented in this chapter.
Respondents were first asked if they have ever heard of speed cameras being used to ticket
drivers who exceed the speed limit. Figure 6-1 shows that the overwhelming majority of drivers
(85%) have heard of the use of speed cameras.
Figure 6-1: Ever Heard of Speed Cameras Being Used to Ticket Drivers

No, 14%

Yes, 85%

Q15. Before today, have you ever heard of speed cameras being used to ticket drivers that speed?
Base: All Respondents
Unweighted N=6,144

Respondents were asked if specific locations would be acceptable for speed camera
implementation. Figure 6-2 shows that the majority of drivers think that speed cameras would be
useful in school zones (86%), places where there have been many accidents (84%), construction
zones (74%), areas where it would be hazardous for a police officer to stop a driver (70%), and
areas where stopping a vehicle could cause traffic congestion (63%). A little over one-third
(35%) of drivers think that speed cameras would be useful on all roads.
Figure 6-2: Locations Where Speed Cameras May Be Useful
100%
86%

84%

80%

74%

70%

63%

60%
35%

40%
20%
0%

16d. School
zone

16c. Where
there have
been many
accidents

16e.
Construction
zone

16a.
Hazardous for
a police officer
to stop a
vehicle

16b. Where
stopping a
vehicle could
cause
congestion

16f. All roads

Q16. Thinking about locations where speed cameras might be useful, would you find it acceptable to use them . . .
Base: All Respondents
Unweighted N=6,144

Table 6-1 shows the percentage of drivers in each demographic category who think the use of
speed cameras in specific locations is acceptable. There were no large differences when these
items are examined by demographics, although the acceptance of speed camera use appears to
decrease with household income and education.
Table 6-1: Location of Speed Cameras by Demographics
Thinking about locations where
speed cameras might be useful,
would you find it acceptable to use
them . . . .

*** p < .001

16a Where it could
be hazardous for a
police officer to
stop a driver***

16b Where
stopping a vehicle
could cause traffic
congestion***

16c Where there
have been many
accidents***

295
281
939
835
1,185
1,211
1,328

72.6%
76.9%
74.4%
66.1%
67.8%
69.1%
67.8%

70.1%
68.7%
65.4%
62.8%
60.9%
61.9%
57.6%

87.9%
83.5%
84.0%
80.4%
82.1%
83.0%
88.1%

2,696
3,448

66.6%
72.5%

59.1%
66.1%

77.7%
89.3%

464
3,327
1,392
933

73.3%
70.2%
67.8%
64.8%

70.9%
63.2%
56.2%
57.5%

89.6%
84.9%
78.0%
77.3%

1,275
1,019
1,102
761
1,119

75.7%
73.8%
67.4%
67.1%
64.7%

72.4%
61.8%
60.1%
59.8%
57.4%

90.4%
86.0%
83.3%
78.3%
77.1%

3,030
2,903

69.8%
70.4%

63.2%
62.9%

83.6%
84.5%

Table 6-1: Location of Speed Cameras by Demographics (Continued)
Thinking about locations where
speed cameras might be useful,
would you find it acceptable to use
them . . . .

*** p < .001

16d In a school
zone***

16e In a
construction
zone***

16f On all roads***

295
281
939
835
1,185
1,211
1,328

85.1%
85.6%
87.4%
84.9%
83.9%
85.1%
89.0%

78.3%
75.0%
73.9%
71.8%
70.4%
74.4%
77.3%

32.6%
39.2%
36.3%
33.9%
32.3%
34.9%
41.0%

2,696
3,448

81.9%
89.5%

70.8%
76.6%

31.2%
39.5%

464
3,327
1,392
933

90.3%
87.4%
79.8%
79.9%

80.6%
74.3%
69.5%
67.5%

49.3%
35.4%
26.4%
26.9%

1,275
1,019
1,102
761
1,119

90.9%
88.3%
85.4%
83.4%
78.8%

79.1%
75.9%
73.3%
70.3%
68.3%

47.7%
37.7%
29.3%
29.1%
26.1%

3,030
2,903

85.0%
87.2%

72.1%
75.9%

34.9%
36.2%

Figure 6-3 displays the percentage of drivers who report speed cameras along the routes they
usually drive and also the percentage of drivers who have received a ticket for a speed violation
identified by a speed camera. Slightly more than a third of drivers report that there are speed
cameras along the routes they usually drive. Interestingly, 10% of drivers did not know whether
speed cameras are used along the routes they normally drive. Less than 1 in 10 drivers (8%)
report having received a speeding ticket in the mail from a speed camera.
Figure 6-3: Speed Cameras on Normal Routes and Received Ticket From Speed Camera
Q17 Along the routes you
normally drive, are there
speed cameras in use?

Q18 Have you ever received a
ticket in the mail for a speed
violation, identified by a speed
camera?

Not sure,
10%

Yes, 8%

Yes, 37%

No, 92%

No, 53%

Base: Heard of speed cameras previously
Unweighted N=5,271

Respondents were asked if they agree or disagree with the statements that speed cameras are
used to prevent accidents and/or generate revenue. Figure 6-4 shows the percentage of drivers
who agree (strongly and somewhat) with each of the two statements. Drivers are more likely to
agree with the statement that speed cameras are used to generate revenue (70%) than with the
statement that speed cameras are used to prevent accidents (55%). This pattern holds true among
those who strongly agree with each statement as well, (38% versus 29%, respectively).
Figure 6-4: Attitudes Regarding the Purpose of Speed Cameras
100%

Somewhat agree
Strongly agree

80%

60%

40%

20%

32%
26%

38%

29%

19a. Prevent accidents

19b. Generate revenue

Q19. Now I’m going to read a few statements. After I read each one, please tell me whether you agree, disagree, or neither.
(READ ITEM). Would you say you strongly (AGREE/DISAGREE) or somewhat (AGREE/DISAGREE)? Speed cameras are
used to . . .
Base: All Respondents
Unweighted N=6,144

The amount of agreement with the statements about the purpose of speed cameras was calculated
on a scale of 1 to 5 for each respondent, where 5=strongly agree and a 1=strongly disagree. Table
6-2 shows the average agreement scores for both statements for each age group, gender,
education level, and household income. The higher the mean value, the more agreement there is
with a particular statement. Drivers 65 and older expressed the most agreement with the
statement that speed cameras are used to prevent accidents, while respondents 25 to 34 expressed
the least amount of agreement. Drivers with the lowest household incomes (<$30,000) expressed
the most agreement with the statement that speed cameras are used to prevent accidents, while
drivers in the highest household income range ($100,000 or more) expressed the least amount of
agreement. The differences in agreement with the statement that speed cameras are used to
generate revenues were not large, but agreement with this statement increased with education
and household income.
Table 6-2: Attitudes Regarding Purpose of Speed Cameras by Demographics
Q19. Now I’m going to read a few
statements. After I read each one,
please tell me whether you agree,
disagree, or neither.

*** p < .001

19a Speed cameras
are used to prevent
accidents***

19b Speed cameras
are used to generate
revenue***

295
281
939
835
1,185
1,211
1,328

3.28
3.50
2.99
3.24
3.14
3.25
3.67

3.57
3.62
3.88
3.83
3.99
3.82
3.55

2,696
3,448

3.04
3.47

3.89
3.69

464
3,327
1,392
933

3.60
3.23
3.09
3.19

3.67
3.77
3.86
3.99

1,275
1,019
1,102
761
1,119

3.57
3.34
3.15
3.12
2.92

3.62
3.76
3.85
3.91
4.01

3,030
2,903

3.25
3.28

3.85
3.70

Figure 6-5 compares agreement with the statements about the purpose of speed cameras by
driver type, and is limited to the percentages who strongly agree with each statement. Drivers
who are classified as speeders are more than twice as likely to strongly agree with the statement
that speed cameras are used to generate revenue (44%) than with the statement that speed
cameras are used to prevent accidents (20%). The same proportion of drivers classified as
nonspeeders strongly agree with each of these two statements (34%).
Figure 6-5: Attitude Toward Purpose of Speed Cameras by Driver Type
% Strongly Agree

100%

Speeders (N=1,572)

Sometime Speeders (N=2,148)

Nonspeeders (N=1,579)

80%
60%
44%

40%
20%
0%

28%

34%

37%

34%

20%

19a. Prevent accidents***

19b. Generate revenue***

CHAPTER 7
CRASH EXPERIENCE
Drivers’ involvement in speeding-related crashes and experience with crash injuries requiring
hospitalization are examined in this chapter. Respondents were first asked how many times they
have been in speeding-related crashes in the past 5 years. Figure 7-1 shows that the majority of
drivers (96%) have not experienced any speeding-related crashes in the past 5 years. Only 3%
were involved in one speeding-related crash and even fewer (1%) had been involved in two or
more speeding-related crashes in that time period.
Figure 7-1 Speeding-Related Crashes in the Past Five Years
Two or more,
1%

One, 3%

None, 96%

Q25. How many times have you been in a speeding related accident in the past five years?
Base: All Respondents
Unweighted N=6,144

Table 7-1 breaks down information on speeding-related crash involvement by demographics.
The mean number of speeding-related crashes per driver, the percentage of drivers with at least
one speeding-related crash, and the average number of these crashes given at least one crash are
shown. The mean number of crashes per driver and the percentage of crashes decrease with
driver age, are similar across gender, and decrease with higher education and with greater
household income. The relationship between age group and speeding-related crashes is further
examined in the following section.
Table 7-1: Speeding-Related Crash Involvement in the Past Five Years by Demographics
How many times have you been in
a speeding-related accident in the
past five years?

*** p < .001

Mean number of
accidents for
respondents
reporting at least
one accident

Mean number of
accidents***

Percent reporting at
least one
accident***

295
281
939
835
1,185
1,211
1,328

0.16
0.12
0.41
0.46
0.05
0.10
0.01

11.0%
9.3%
6.7%
3.7%
2.5%
1.5%
1.2%

1.46 (n=28)
1.30 (n=30)
1.54 (n=53)
1.32 (n=31)
1.11 (n=31)
1.06 (n=17)
1.05 (n=19)

2,696
3,448

0.28
0.26

3.5%
4.9%

1.33 (n=75)
1.37 (n=134)

464
3,327
1,392
933

0.47
0.17
0.08
0.03

4.4%
4.6%
4.2%
1.5%

1.65 (n=21)
1.28 (n=132)
1.19 (n=42)
1.88 (n=14)

1,275
1,019
1,102
761
1,119

0.53
0.12
0.08
0.05
0.04

5.7%
3.8%
2.7%
3.7%
3.3%

1.37 (n=64)
1.32 (n=34)
1.80 (n=24)
1.22 (n=22)
1.23 (n=34)

3,030
2,903

.06
.06

4.1%
4.4%

1.39 (n=94)
1.33 (n=109)

Figure 7-2 shows the percentage of each age group of drivers that had been involved in at least
one speeding-related crash in the past 5 years. Among drivers 16 to 20, 11% had at least one
speeding-related crash in the past 5 years. This is the highest percentage among all the age
groups, even though these young drivers had not been driving for each of the past 5 years. Of
drivers 21 to 24, 9% had a speeding-related crash in the past 5 years, while only 1% of drivers
65 and older had at least one speeding-related crash in the past 5 years.
Figure 7-2 Percentage of Drivers Reporting at Least One Speeding-Related Crash in Past
Five Years by Age Group***

20%

11%

10%

9%
7%
4%
2%

16-20
(N=295)

21-24
(N=281)

25-34
(N=939)

35-44
(N=835)

45-54
(N=1185)

Q25. How many times have you been in a speeding related accident in the past five years?
Base: All Respondents
Unweighted N=See Chart *** p < .001

55-64
(N=1211)

65 or older
(N=1329)

Overall, the average number of speeding-related crashes in the last 5 years also decreases with
age. Figure 7-3 shows the average per driver number of speeding-related crashes in the past 5
years ranges from 0.01 accidents for drivers 65 or older, up to 0.16 accidents for drivers 16 to 20.
When only drivers who had at least one speeding-related crash in the past 5 years are considered,
the average number of speeding-related crashes in the past 5 years peaks at a value of 1.54 for
the 25 to 34 year old age group. Again, the oldest age group (65 and older) was the lowest value
on this measure, with an average of just over one speeding-related crash in the past 5 years.
Figure 7-3 Mean Number of Speeding-Related Crashes in
Past Five Years by Age Group

1.5

1.54

1.46

1.32

1.30

1.11

0.5

0.41
0.16

16-20

25-34

1.05

0.46

0.12
21-24

1.06

35-44

Drivers who have been in crashes*** (N=209)

0.05

0.02

0.01

45-54

55-64

65 or older

All Drivers*** (N=6,071)

Q25. How many times have you been in a speeding related accident in the past five years?
Base: Respondents who have been in a speeding related accident in the past five years
Base: All Respondents
Unweighted N=See Chart *** p < .001

Approximately two-thirds (68%) of drivers were not injured in their most recent speeding-related
crash, but nearly 1 in three (29%) crash-involved drivers reported being injured. An additional
3% of crash-involved drivers did not know or refused to say if they had been injured in their
most recent speeding-related crash.
Figure 7-4 Percentage of Drivers Reporting Injuries
In Most Recent Crash

No, 68%

Yes, 29%

Don't know/
Refused, 3%

Q27. Did you receive any injuries as a result of the most recent speeding related accident?
Base: Respondents who were in a speeding related accident
Unweighted N=219

Figure 7-5 shows that there is a correlation between driver type and the number of speedingrelated crashes for those drivers who had been in at least one speeding-related crash in the past 5
years. As driver type classification goes from nonspeeder to sometime speeder to speeder, the
percentage of drivers who had multiple speeding-related crashes in the past 5 years increases.
Among nonspeeders with more than one crash in the past 5 years, 1 in 6 (17%) had two or more
such crashes in the past 5 years. One-fifth of sometime speeders (20%), and 3 in 10 speeders
(30%) with more than one speeding-related crash had two or more such crashes in the past 5
years.
Figure 7-5 Percentage of Drivers with More Than One Speeding-Related Crash
In Past Five Years by Driver Type***
100%

80%

60%

83%

80%

70%

One

40%

Two to four
Five or more

20%
16%

Nonspeeders (N=53)

17%

28%

Sometime Speeders
(N=73)

Speeders (N=63)

Q25. How many times have you been in a speeding related accident in the past five years?
Base: Respondents Assigned a Driver Type who have been in at least one speeding related accident in the past five years
Unweighted N=See Chart *** p < .001

The results presented in Figure 7-6 suggest that drivers with patterns of speeding behavior are
more likely to suffer injuries in speeding-related crashes. Of all drivers reporting injuries
resulting from speeding-related crashes (n=192), 45% are speeders, 31% are sometime speeders,
and 24% are nonspeeders.
Figure 7-6 Drivers Reporting Injuries From Most Recent
Speeding-Related Crash by Driver Type***

Sometime Speeders
(N=74), 31%

Speeders (N=64), 45%
Nonspeeders (N=54),
24%

Q27. Did you receive any injuries as a result of the most recent speeding related accident?
Base: Respondents Assigned a Driver Type who were in a speeding related accident
Unweighted N=See Chart *** p < .001

CHAPTER 8
PERSONAL SANCTIONS
Sanctions experienced by people who were stopped by police for speeding are examined in this
chapter. Figure 8-1 shows that most drivers have not been stopped for speeding by the police in
the past 12 months, with less than 1 in 10 (9%) having been stopped for speeding.
Figure 8-1 Percentage of Drivers Stopped for Speeding in the Past 12 Months

Yes, 9%

No, 91%

Q30. In the past TWELVE MONTHS have you been STOPPED for speeding by the police?
Base: All Respondents
Unweighted N=6,144

Table 8-1 shows the percentage of drivers in each demographic group that had been stopped for
speeding in the past 12 months. Younger drivers and male drivers were more likely than older
drivers and female drivers to have been stopped the past 12 months. The driver’s level of
education or household income does not appear to be associated with the likelihood of being
stopped for speeding.
Table 8-1: Drivers Stopped for Speeding in the Past 12 Months by Demographics
In the past TWELVE MONTHS have
you been STOPPED for speeding by
the police?

*** p < .001

Percent stopped
for speeding in
past 12 months

295
281
939
835
1,185
1,211
1,328

17.5%
15.6%
17.8%
8.1%
5.8%
5.6%
2.5%

2,696
3,448

11.2%
7.5%

464
3,327
1,392
933

9.7%
9.1%
8.9%
10.6%

1,275
1,019
1,102
761
1,119

10.3%
9.4%
10.1%
9.6%
8.7%

3,030
2,903

8.2%
10.8%

Figure 8-2 shows the type of sanctions experienced by drivers who had been stopped for
speeding in the past 12 months. Most of these drivers (68%) were issued a ticket. More than a
quarter (27%) received a warning, and 1 in 20 (5%) did not receive a ticket or a warning.
Figure 8-2 Sanctions Experienced by Drivers Stopped for Speeding

Neither, 5%

Warning, 27%

Ticket, 68%

Q32a. Did you receive a ticket during the last time you were stopped for speeding?
Q32b. Did you receive a warning the last time you were stopped for speeding?
Base: Respondent who were stopped for speeding in the past twelve months
Unweighted N=465

Table 8-2 presents the distribution of the type of sanctions experienced by drivers who had been
stopped by police for speeding in the past 12 months by demographics. More formal education
appears to be positively associated with the likelihood of receiving a ticket rather than a warning.
No large differences by gender, age or other demographic categories were noted.
Table 8-2: Distribution of Sanctions by Demographics
Q32a. Did you receive a ticket
during the last time you were
stopped for speeding?
Q32b. Did you receive a warning
the last time you were stopped
for speeding?

*** p < .001

Ticket

Warning

Neither

46
36
137
78
71
62
33

60.9%
68.0%
75.2%
64.1%
70.2%
64.4%
62.5%

31.8%
29.7%
21.9%
34.2%
23.1%
24.1%
34.3%

7.3%
2.2%
2.9%
1.7%
6.7%
11.5%
3.2%

255
210

68.1%
69.2%

28.0%
25.5%

4.0%
5.3%

41
229
115
80

64.1%
67.7%
71.5%
75.8%

31.9%
27.6%
26.0%
17.2%

4.0%
4.7%
2.5%
7.0%

94
68
95
59
99

68.5%
74.0%
61.2%
68.2%
68.2%

29.1%
21.3%
36.2%
22,4%
25.0%

2.4%
4.7%
2.6%
9.5%
6.8%

204
249

75.1%
63.3%

19.8%
32.6%

5.0%
4.1%

Figure 8-3 shows the distribution of the frequency of speeding stops for drivers who had been
stopped at least once by police for speeding in the past 12 months. The majority (84%) of these
drivers had been stopped for speeding only once. One in seven (15%) had been stopped two to
four times in the past 12 months, and 1% were stopped five times or more.
Figure 8-3 Number of Times Stopped for Speeding in the Past 12 Months

Five to seven, 1%

Two to four, 15%

Mean: 1.3
Median: 1.0
One, 84%

Q31. How many times have you been stopped for speeding in the past twelve months?
Base: Respondents who were stopped for speeding in the past twelve months
Unweighted N=465

Figure 8-4 shows a clear pattern of who was stopped for speeding by police in the past 12
months by driver type. While 9% of all drivers reported being stopped for speeding in the
previous 12 months (see Figure 8-1), only 4% and 5% of drivers classified as nonspeeders and
sometime speeders, respectively, were stopped. Among the drivers classified as speeders, 1 in 5
(20%) was stopped for speeding by police in the past 12 months.
Figure 8-4 Stopped for Speeding in the Past 12 Months by Driver Type

30%

20%

10%
5%

Nonspeeder (N=1,579)

Sometime speeder
(N=2,148)

Speeder (N=1,572)

Q30. In the past TWELVE MONTHS have you been STOPPED for speeding by the police?
Base: Respondents Assigned a Driver Type
Unweighted N=See Chart *** p < .001

As noted earlier and shown in Figure 8-5, the majority of drivers stopped for speeding regardless
of driver type category received a ticket. However, the likelihood of receiving a ticket increased
with driver classification from nonspeeders, to sometime speeders, to speeders with 62% of
nonspeeders, 64% of sometime speeders, and 69% of speeders stopped for speeding receiving a
ticket.
Figure 8-5 Frequency of Receiving Ticket or Warning by Driver Type***

Nonspeeders
(N=46)

Sometime Speeders
(N=91)

Neither, 5%

Neither, 3%

Neither, 5%

Warning, 26%

Warning,
33%

Warning, 34%

Speeders
(N=283)

Ticket, 64%

Ticket, 62%

Ticket, 69%

Q32a. Did you receive a ticket during the last time you were stopped for speeding?
Base: Respondents Assigned a Driver Type who were stopped for speeding in the past twelve months
Unweighted N=See Chart *** p < .001

Drivers classified as speeders were the most likely drivers to get pulled over and ticketed and
were also the least likely to change their driving behavior as a result of their ticket or warning.
Figure 8-6 shows the percentage of drivers by driver type who had experienced a speed-related
stop and indicated that they changed their driving behavior because of that stop. Among
nonspeeders, 86% reported that they changed their driving behavior as a result of their ticket or
warning. About 4 in 5 (79%) of sometime speeders stated they changed their driving behavior.
Among speeders, the percentage reporting a changed driving behavior was 71%.
Figure 8-6 Percentage of Drivers Reporting Changing Their Driving Behavior by Driver
Type***

100%
86%
79%

80%

71%

60%
40%
20%
0%

Nonspeeders (N=44)

Sometime Speeders
(N=87)

Speeders (N=265)

Q33. Did you change your driving behavior as a result of receiving the (TICKET/WARNING) for speeding?
Base: Respondents Assigned a Driver Type who received a ticket or warning for speeding in the past twelve months
Unweighted N=See Chart *** p < .001

CHAPTER 9
OTHER RISKY BEHAVIOR
Incidence of risky behaviors such as not wearing seat belts while driving, driving after drinking
alcohol, and talking and texting while driving are examined in this chapter. Approximately 9 in
10 (89%) drivers reported that they wear their seat belts all of the time while driving their
primary vehicle (see Figure 9-1). One in 10 (10%) drivers stated that they do not wear their seat
belts all the time, and 1% reported that they never wear their seat belts.
Figure 9-1 Seat Belt Usage
Rarely, 1%
Never, 1%

Some of the time,
3%
Most of the time,
6%

All of the time, 89%

Q34. When driving your primary vehicle how often do you wear your seat belt?
Base: All Respondents
Unweighted N = 6,144

Examining seat belt use by age group shows that drivers under 35 are more likely than older
drivers to indicate that they wear their seat belts only some of the time, rarely or never. One in
twelve (8%) drivers 21 to 24 years old, and 7% of drivers 16 to 20 and 25 to 34, report that they
wear their seat belts less than most of the time while driving their primary vehicles (see Figure 92).
Figure 9-2 Seat Belt Use by Age Group***

10%

Never

Rarely

Some of the time

1%
2%

16-20
(N=295)

21-24
(N=281)

25-34
(N=939)

35-44
(N=835)

45-54
(N=1,185)

55-64
(N=1,211)

65 or older
(N=1,329)

Q34. When driving your primary vehicle how often do you wear your seat belt?
Base: All Respondents
Unweighted N=See Chart *** p < .001

Drivers classified as speeders are less likely than other drivers to wear their seat belts most of the
time. Figure 9-3 shows the percentage of drivers by driver type who wear their seat belts some of
the time, rarely and never. One in twelve (8%) speeders wear their seat belts only some of the
time or less while driving. Among the drivers classified as sometime speeders and nonspeeders,
only 5% and 3%, respectively, report wearing their seat belts only some of the time or less
frequently.
Figure 9-3 Seat Belt Use by Driver Type***

10%

Never

Rarely

Some of the time

8%
4%

5%
3%

2%
<1%

Nonspeeders (N=1,579)

Sometime Speeders
(N=2,148)

Q34. When driving your primary vehicle how often do you wear your seat belt?
Base: Respondents Assigned a Driver Type
Unweighted N=See Chart ***p<.001

2%
Speeders (N=1,572)

When asked about drinking and driving in the past 30 days, only a small proportion of drivers
(2%) reported driving a vehicle after they thought they had drank too much alcohol to drive
safely (see Figure 9-4).
Figure 9-4 Alcohol Consumption and Driving

Yes, 2%

No, 98%

Q35. In the past 30 days, have you driven a vehicle when you thought you might have consumed too much alcohol to drive
safely?
Base: All respondents
Unweighted N=6,144

Most (98%) drivers stated that they have not driven a vehicle when they thought they had too
much to drink in the past 30 days. The highest percentage (4%) of drivers who admit that they
have driven after they had consumed too much to drive safely is among drivers 25 to 34. Of
drivers 16 to 20 and 21 to 24, 3% reported driving after having too much to drink (see figure 95).
Figure 9-5 Alcohol Impaired Driving by Age Group***

5%
4%
3%

3%
2%

2%
1%

1%
0%<1%

16-20
(N=295)

21-24
(N=281)

25-34
(N=939)

35-44
(N=835)

45-54
(N=1,185)

55-64
65 or older
(N=1,211) (N=1,329)

Q35. In the past 30 days, have you driven a vehicle when you thought you might have consumed too much alcohol to drive
safely?
Base: All respondents
Unweighted N=See Chart *** p < .001

As shown in Figure 9-6, Speeders are more likely to drive their car when not wearing their seat
belts, although the vast majority (83%) still claim to buckle up all of the time. They are also
more likely to drive after drinking too much alcohol compared to nonspeeders (3% versus 1%,
respectively).
Figure 9-6 Seat Belt Usage and Alcohol Impaired Driving by Driver Type***

100%

Speeders (N=1,572)

83%

89%

Sometime Speeders (N=2,148)

Nonspeeders (N=1,579)

93%

80%
60%
40%
20%
3%

Q34. Seat Belt Usage (% All the time)

Q35. Drink before Driving (% Yes)

Q34. When driving your primary vehicle how often do you wear your seat belt?
Q35. In the past 30 days, have you driven a vehicle when you thought you might have consumed too much alcohol to drive
safely?
Base: All respondents
Unweighted N=See Chart *** p < .001

Respondents were asked a series of questions about their use of cell phones while driving. The
majority of drivers have a cell phone in their vehicle when they drive; only about 1 in 10 (11%)
report not having cell phones in their vehicles (see Figure 9-7).
Figure 9-7 Cell Phone in Vehicle While Driving

No, 11%

Yes, 89%

Q36. When you drive a motor vehicle, do you usually have a cell phone or wireless phone of some type in the vehicle with you?
Base: All Respondents
Unweighted N=6,144

Examining cell phone use while driving by age shows a relationship that peaks at 25 to 34, with
16% of drivers stating that they talk on the cell phone on all or most trips, and 6% reporting that
they send or read text messages while driving on all or most trips (see Figure 9-8). After age 35,
the proportion of drivers who talk and send or read text messages while driving on all or most of
their trips decreases as age increases. Among drivers 65 or older, only 3% report talking on the
phone while driving on all or most trips, and none report reading or sending text messages at any
time while driving. The youngest drivers are most likely to read and send text messages.
Figure 9-8 Percentage of Drivers Who Use Cellular Phone
While Driving on All or Most Trips

25%
20%
16%

15%

13%

10%
8%

5%
0%

12%

9%
7%

6%
4%

16-20

21-24

2%
25-34

35-44

Text while driving

3%
1%
45-54

55-64

65 or older

Talk while driving

Q37. How often do you talk on the phone while you are driving? Would you say you talk on the phone while driving during…?
Q39. How often do you read OR send text messages while you are driving and the vehicle is moving? Would you say you read
OR send text messages while driving during…?
Base: Respondents who have a cell phone in their vehicle while driving,
Unweighted N=5,340

As shown in Figure 9-9, speeders are more likely to engage in distracted driving behavior while
behind the wheel, when compared to nonspeeders and sometime speeders. Close to 1 in 6
speeders (16%) say they talk on the phone while driving during all or most of their trips,
compared to 8% of sometime speeders and 7% of nonspeeders. Similarly more speeders text
while driving (6%) when compared to sometime speeders (2%) and nonspeeders (<1%).
Figure 9-9 Talk and Text While Driving by Driver Type***

100%

Speeders (N=1,572)

Sometime Speeders (N=2,148)

Nonspeeders (N=1,579)

80%
60%
40%
20%
0%

16%
8%

Q37. Talk while driving (% All or Most
Trips)

Q39. Text while driving (% All or Most
Trips)

Q37. How often do you talk on the phone while you are driving? Would you say you talk on the phone while driving during…?
Q39. How often do you read OR send text messages while you are driving and the vehicle is moving? Would you say your read
OR send text messages while driving during…?
Base: Respondents Assigned a Driver Type
Unweighted N=See Chart

CHAPTER 10
TREND ANALYSIS
The 2011 National Survey of Speeding Attitudes and Behavior is the third in a series of surveys
on speeding conducted by NHTSA. The previous speeding surveys were conducted in 1997 and
2002. Questions that appeared in all three surveys and some questions that are similar are
compared across the three surveys. These comparisons offer insight into how driving habits and
behaviors have changed, or not changed, in the past 14 years.
At the beginning of each study, respondents were asked how often they drive. In the present
study, respondents report driving less than they did in 1997 or 2002. In 1997, 88% of drivers
stated that they drive every day or almost every day; in 2002, 83% reported that they drive every
day or almost every day; and in 2011, 81% reported driving daily or almost every day (See
Figure 10-1). In all three studies, 1% of drivers report driving only a few times a year or only at
certain times of the year.
Figure 10-1: How Frequently Do You Drive by Year
1997 (N=5,997)

2002 (N=2,004)

2011 (N=6,144)

100%

88%

83% 81%

75%

50%

25%

9%
0%

Every day or almost
every day/ Almost
every day

13% 13%
5%
2% 3%

1% 1% 1%

Several days a week/ A Once a week or less/ A Only certain times a
few days a week
few days a month
year/ A few times a
year

1997 – Q1. How often do you usually drive a car or other motor vehicle? Would you say that you usually drive…?
2002 – Q1. How often do you usually drive a car or other motor vehicle? Would you say that you usually drive…?
2011 – Q1. How often do you usually drive a car or other motor vehicle? Would you say that you usually drive…?
Base: All Respondents
Unweighted N=See Chart

100

Respondents in each study were asked if they tended to pass other cars more often than other
cars passed them. There has not been substantial change in these behaviors across the three
studies. Around 3 in 10 drivers state that they tend to pass other cars (31% in 1997, 30% in 2002,
27% in 2011). Nearly 3 in 5 report that other cars tended to pass them (59% in 1997, 58% in
2002, and 59% in 2011). In 2011, 1 in 7 (14%) drivers selected reported that the number of cars
that pass them and the number of cars that they pass are about equal, compared to about 1 in 10
in 1997 (10%) and 2002 (11%) (See Figure 10-2).
Figure 10-2: Passing Behavior by Year
1997 (N=2,956)

2002 (N=2,004)

2011 (N=6,144)

100%

75%

59%

58%

59%

50%

31%

30%

27%

25%

10%

11%

14%

0%
I tend to pass other cars

Other cars pass me

Both about equal

1997 – Q8a. Which of the following statements best describes your driving? READ STATEMENTS
Base: Respondents in version A of the 1997 survey
2002 – Q4a. Which of the following statements best describes your driving?
2011 – Q3. Which of the following statements best describes your driving?
Unweighted N=See Chart

101

A series of questions about attitudes and beliefs associated with driving were comparable across
all three surveys. Enjoyment of driving fast appears to have decreased over time, as did
agreement with the statement, “the faster I drive, the more alert I am.” In 1997, two-fifths (40%)
of drivers strongly agreed or somewhat agreed that they enjoyed driving fast. About one-third
(34%) of drivers agreed with this statement in 2002, and about a one-quarter (27%) of drivers
agreed with it in 2011. The percentage of drivers who strongly agreed or somewhat agreed that
the faster they drive, the more alert they feel did not change much from 1997 (29%) to 2002
(30%), but dropped by one-half to 15% in 2011. In 1997 and 2002, approximately 3 in 10 (30%
in 1997, and 31% in 2002) drivers, strongly agreed or somewhat agreed that they go as fast as
possible so they can get to their destination quicker. However, in 2011, only 1 in 5 (21%)
strongly or somewhat agreed with this statement.
The feelings of impatience with slow drivers and worrying about having a crash remained
relatively constant over the time of the three surveys. More than one-half of drivers in all three
surveys (60% in 1997, 53% in 2002, 60% in 2011) strongly or somewhat agreed that they often
get impatient with slower drivers. In each year, nearly one-half of drivers (47% in 1997, 46% in
2002, and 48% in 2011) strongly or somewhat agree that they worry a lot about having a crash
(See Figure 10-3).
Figure 10-3 Driver Attitude Trends – Strongly or Somewhat Agree
1997 (N=3,044)

100%

2002 (N=2,004)

2011 (N=6,144)

75%
61%

60%

50%
25%
0%

53%
40%

34%

30%31%

27% 29%30%

21%

15%

The faster I
I enjoy the
feeling of driving drive, the more
alert I am
fast/Enjoy
feeling of speed

47%46%48%

I often get
impatient with
slower drivers

I try to get where
I am going as
fast as I can/I go
as fast as I can
go to get
somewhere

I worry a lot
about having a
crash

1997 – Q10. People have different feelings about driving. I’d like you to tell me whether you agree or disagree with the following
statement about driving.
2002 – Q5. People have different feelings about driving. I’d like you to tell me whether you agree or disagree with the following
statements about driving. For each of the statements, please tell me whether you strongly agree, somewhat agree, somewhat
disagree or strongly disagree. (Read and rotate A-E)
2011 – Q9. Now I’m going to read a few statements. After I read each one, please tell me whether you agree, disagree, or neither.
(READ ITEM). Would you say you strongly (AGREE/DISAGREE) or somewhat (AGREE/DISAGREE)?
Base: All Respondents
Unweighted N=See Chart

102

There seems to be almost no change in the percentage of drivers being pulled over for speeding
by the police. It should be noted that, in the previous studies, respondents were asked if they had
been pulled over in the prior 12 months for any reason. Respondents were then asked for what
reason or reasons they were pulled over. In the current survey, respondents were asked
specifically about being pulled over for speeding. To enable comparison of responses from the
previous studies to the current study, a new variable was created for both the 1997 data and the
2002 data that combined the first question about being stopped with the follow-up question about
the reason of the stop, which identified respondents who were stopped only for speeding. As
shown in Figure 10-4, approximately 1 in 10 drivers were stopped for speeding across all three
studies. In 1997, 9% of drivers reported they were stopped for speeding in the prior 12 months
by a police officer. In 2002, this percentage was 11% and in 2011, it was 9%.
Figure 10-4 Stopped for Speeding in the Past 12 Months
15%

12%

11%
9%

1997 (N=2,953)

2002 (N=2,004)

2011 (N=6,144)

1997 – Q97a. In the past twelve months, have you been STOPPED by the police for any traffic-related reason?
Q97c. What were you stopped for?
2002 – Q79. In the past twelve months have you been STOPPED by the police for any traffic related reason?
Q81. What type of traffic related violation have you been stopped for?
2011 – Q30. In the past TWELVE MONTHS have you been STOPPED for speeding by the police?
Base: All Respondents
Unweighted N=See Chart

103

The proportion of drivers who received tickets when stopped by police for speeding appears to
be relatively constant over the three surveys, with approximately two-thirds of those stopped,
receiving tickets. In the 1997 and the 2002 surveys, respondents were asked if they had received
a ticket, a warning or both during their traffic stop. In the 2011 survey respondents were first
asked if they’d received a ticket; if they reported that they had not received a ticket they were
asked if they received a warning. Respondents who reported receiving a ticket for speeding were
not asked if they also received a warning. The percentages of drivers stopped for speeding who
reported receiving tickets are: 65% in 1997, 70% in 2002, and 68% in 2011. (See Figure 10-5).
In 1997, 4% of drivers reported receiving both a warning and a ticket for speeding, while, in
2002 8% reported receiving both a warning and a ticket for speeding.
Figure 10-5 Tickets and Warnings by Year

1997 (N=250)
Both, 4%

2002 (N=206)

Neither,
3%

Both, 8%
Warning,
10%

Neither,
5%
Neither,
19%

Warning,
33%
Ticket,
61%

2011 (N=465)

Warning,
27%

Ticket,
62%

1997 – Q97d. Did you receive a ticket or warning (on any of those occasions)?
2002 – Q82. Did you receive (A. a ticket/B. a written warning) on any of these occasions?
2011 – Q32a. Did you receive a ticket during the last time you were stopped for speeding?
Q32b. Did you receive a warning the last time you were stopped for speeding?
Base: Respondents pulled over (for speeding) in the past twelve months
Unweighted N=See Chart

104

Ticket,
68%

Conclusion
For over a decade, NSSAB studies have provided data that have helped further the understanding
of driving behavior and contributed to the development of countermeasures and interventions to
reduce speeding. The present study is the third in this series, and, like the previous studies, yields
national estimates of behavior and attitudes toward speeding in the United States. The present
study differs from the earlier studies in that it developed and used a driver typology based on the
pattern of responses across six speeding behavior questions. Cluster analysis identified three
distinct groups of drivers with similar overall behavioral tendencies and accounted for 86% of
respondents. Because of the nature of these behavioral tendencies, the driver types are referred to
as nonspeeders, sometime speeders, and speeders in this report. Among those categorized, 30%
are nonspeeders, 40% are sometime speeders, and 30% are speeders.
In terms of demographics, drivers classified as speeders tend to be younger and male, and to
have higher household incomes when compared to sometime speeders and nonspeeders.
Interestingly, 36% of all male drivers, one-half of drivers 16 to 20, and 42% of drivers with
annual household incomes of $100,000 or more were classified as speeders. The typology was
particularly useful in distinguishing self-reported behaviors and attitudes toward speeding and
toward interventions aimed at speeding among drivers.
As in the two previous NSSAB studies, approximately 10 % of drivers report being stopped by
police for speeding in the past year, and about two-thirds of these report receiving a ticket.
Overall, most drivers report driving at approximately the speeds they perceive to be safe for the
type of roads on which they are travelling. However, drivers who have been stopped for speeding
within the past year report traveling faster than their perceived safe speed limit would allow,
reflecting a willingness to accept the risks associated with speeding. Not surprisingly, drivers
classified as speeders were 4 to 5 times as likely to be stopped for speeding as sometime
speeders or nonspeeders. They were also more likely than other drivers to receive a ticket instead
of a warning if stopped for speeding. Drivers who had been stopped by police within the past
year and received a warning rather than a speeding ticket, on average, believe that driving about
11 mph over the speed limit on multi-lane divided highways and two-lane roads will not result in
a speeding ticket. Their average perceived “allowable” over-speed-limit margin was greater than
that identified by drivers who received tickets. While this suggests that tickets may be a better
deterrent to speeding than warnings, speeders who received speeding tickets in the past year
were more likely than others to report that this experience did not change their driving behavior.
Clearly, there still is much to learn about the effects of police enforcement strategies on speeding
behaviors of various types of drivers.
Only a very small portion of drivers report experiencing a speeding-related crash in the past 5
years and even fewer (about 1%) reported being in two or more speeding-related crashes in that
time period. However, 11% of drivers 16 to 20 reported at least one speeding-related crash in the
past 5 years. The percentage of drivers in speeding-related crashes in this age group is greater
than in any other age group, even though these young drivers may not have been driving for all
of the past 5 years. This age effect is not surprising, considering the high overall crash rates of
young drivers. This result continues to support further traffic safety interventions and efforts
aimed specifically at young drivers.

105

About 90% of drivers in the present study report having cell phones in their cars compared to
about 60% in 2002. Cell phone use while driving differs by driver type. Speeders are more likely
than sometime speeders, who in turn are more likely than nonspeeders, to have cell phones in
their vehicles, to talk on their cell phone while driving and to send or read text messages while
driving. Most drivers use their seat belts on all trips. Overall, 11% of drivers report that they do
not use seat belts on all of their trips. Only 1% of drivers reports never wearing seat belts while
driving. Drivers classified as speeders are less likely than other drivers to wear their seat belts
most of the time, further exhibiting their tendency to take risks.
When normative attitudes toward speeding are explored, the majority of drivers at least
somewhat agree with the statements that “Everyone should obey the speed limits because it’s the
law,” and “People should keep up with the flow of traffic.” Approximately one-half of drivers at
least somewhat agree that, “There is no excuse to exceed the speed limits.” There is general
agreement across all driver types that exceeding the speed limit by 20 mph is unacceptable. Even
among speeders, 70% agree that, “It is unacceptable to exceed the speed limits by more than 20
mph.” Among sometime speeders and nonspeeders, strong agreement with this statement is
reported by 77 % and 84 %, respectively.
In other regards, attitudes of speeders and nonspeeders are again quite different. Less than half of
speeders strongly agree that “Everyone should obey the speed limits because it’s the law,”
compared to 70% of sometime speeders and 80% of nonspeeders. Almost two-thirds of speeders
strongly agree that “People should keep up with the flow of traffic,” but only 42% of the
nonspeeders strongly agree with this statement. Nonspeeders are more than twice as likely as
speeders to strongly agree with the statement that, “There is no excuse to exceed the speed limit”
(41% versus 16%). Speeders are almost three times as likely as sometime speeders to strongly
agree with the statements, “I often get impatient with slower drivers,” (45% versus 18%,), “I
enjoy the feeling of driving fast” (19% versus 6%), and “I try to get where I am going as fast as I
can” (11% versus 3%,).
The acceptability of proposed speeding countermeasures varies among driver types, but overall,
drivers are more receptive to countermeasures if they do not include specific penalties.
Electronic signs by the road that warn drivers that they are speeding and should slow down and
increasing public awareness of the risks of speeding are considered to be good ideas for their
community by a large majority of drivers. Two-thirds of drivers indicate that more frequent
ticketing for speeding in their community is a good idea and 40% indicate that higher fines for
speeding tickets is a good idea.
When asked about automated speeding countermeasures such as speed cameras, an
overwhelming majority of drivers report that they have heard of speed cameras being used to
ticket drivers who speed. However, only about one-third of drivers report the existence of speed
cameras on their normal driving routes. The majority of drivers think that speed cameras would
be useful in school zones, places where there have been many accidents, construction zones,
areas where it would be hazardous for a police officer to stop a driver, and areas where stopping
a vehicle could cause traffic congestion. Increased use speed cameras in dangerous or high crash
locations is also considered to be a good idea by a large majority. However, drivers are more

106

likely agree with the statement that “Speed cameras are used to generate revenue” than they are
to agree that “Speed cameras are used to prevent accidents.”
Countermeasures associated with speeding tickets were less acceptable to speeders than to other
drivers. While two-thirds of sometime speeders and more than three-quarters of nonspeeders
approved of increased ticketing for speeding, only about half of speeders approved of this idea.
Higher fines for speeding were considered a good idea by about half of nonspeeders, 40% of
sometime speeders, and approximately one-third of speeders.
There was a difference by driver type in the acceptability and perceived effectiveness of invehicle speeding countermeasures. Overall, approximately 60% of all drivers indicated that invehicle countermeasures, such as a device in the motor vehicle that notifies you if you are
speeding, a device that records the speed data and reports it to the insurance company to lower
premiums, and a device that slows down the vehicle when it senses another car or object is too
close to the vehicle was a good idea and would prevent them from speeding. However, speeders
were less likely to state that a specific countermeasure would keep them from speeding. The
most promising in-vehicle countermeasure for speeders appears to be the device that records
speeding information and reports it to the insurance company. Slightly more than half of
speeders, two-thirds of sometimes speeders and almost three-quarters of nonspeeders stated that
this device would keep them from speeding.
The driver typology developed in this study appears to be useful in discriminating some driver
attitudes and behaviors. Drivers classified as speeders report more risky behaviors than other
drivers and appear to be the most resistant to conventional countermeasures and interventions
aimed at speeding. On the other hand, drivers classified as nonspeeders exhibit compliance with
traffic laws and, in general, do not speed. Finding interventions that will work on the first group
is challenging and requires continued efforts to identify effective measures. Extraordinary
interventions for the nonspeeder group are not needed, as normal public information programs
and enforcement appear to work well.
The third group identified in this study appears to hold much promise for speeding reduction
efforts. The drivers classified as sometime speeders accounts for close to 40% of drivers,
forming a group larger than either that of speeders or nonspeeders. Their self-reported speeding
behavior is not as consistent as that of speeders or nonspeeders, nor are their attitudes as extreme.
They also appear to be more amenable than speeders to countermeasures and interventions to
reduce speeding, thus offering opportunities to reduce the overall prevalence of speeding on the
nation’s roadways. While the present study did not subdivide this group further, it is highly
likely that this group is not homogenous with respect to speeding behaviors, and that further
groupings of drivers based on their behaviors can be identified. For example, some drivers from
this group may exceed the speed limit by a small amount most of the time, while others may
exceed the speed limit by a large margin, but only occasionally or on specific types of trips or
roads, or under other circumstances. Some of these questions can be explored through further
analysis of the data collected in this study, and through additional research efforts specifically
aimed at these drivers’ behavior and their acceptance and responses to various conventional and
innovative countermeasures and interventions.

107

One of the limitations of the approach used in the NSSAB studies is that all the behaviors are
self-reported and lack confirmation with more objective measures. State driver history files
contain information on licensing, citations, convictions, crashes, license revocation, and
reinstatements of all drivers in a state. Matching up driver records with their attitudes, beliefs and
self-reported behavior would be extremely informative both in understanding driving behavior
and for developing interventions. However, not all driving behavior is captured in driver history
records. As noted earlier in this report, about 10% of drivers are stopped for speeding by police
every year, and only 3% were involved in a speed-related crash in a 5-year period. Thus, some
speeding driving behavior might not be evident from the driver history file. A study that matches
a driver’s real world behavior with attitudes and beliefs about speeding, perhaps also with the
individual’s driving records, might address these shortcomings. Current technology has made it
possible to observe and record driving behaviors in naturalistic driving studies. Thus, research
that combines all three aspects: surveys of driver’s attitudes and beliefs, driver history records,
and observations of real world driving could be invaluable in advancing our knowledge of
driving behavior and in turn advancing the development of effective countermeasures to
speeding.

108

References
Blumberg, S. J., & Luke, J. V. (2012, June). Wireless substitution: Early release of estimates from the National
Health Interview Survey, July–December 2011. Atlanta: National Center for Health Statistics. Available at
www.cdc.gov/nchs/nhis.htm
De Pelsmacker, P., & Janssens, W. (2007, January). The effect of norms, attitudes and habits on speeding behavior:
Scale development and model building and estimation. Accident Analysis & Prevention, 39, Issue 1, Pages 6-15.
Elliott, M. A., Armitage, C. J., & Baughan, C. J. (2003, October). Drivers' compliance with speed limits: An
application of the theory of planned behavior. Journal of Applied Psychology, 88(5), 964-972.
National Center for Statistics and Analysis. (2013). Speeding: Traffic Safety Facts 2011 Data. (Report No. DOT HS
811 751). Washington, DC: National Highway Traffic Safety Administration.

109

APPENDIX A
QUESTIONNAIRE

CELL SAMPLE
SC1

Hello, I am _____ calling on behalf of the U.S. Department of Transportation. We are conducting
a national study on traffic safety. I know I’m calling you on your cell phone, but we are
conducting a brief survey and we would like to send you $10 if you are eligible and willing to
answer some questions.
[IF NEEDED: Any answers you give are kept strictly private. It will only take about 20 minutes.
The OMB number for this solicitation is 2127-0613]
QLAN WHICH LANGUAGE INTERVIEW CONDUCTED IN
1
2

English
Spanish

Are you currently driving?
1
2
9

Yes
No
Refused

THANK AND END, CALLBACK
THANK AND END, SOFT REFUSAL

SC1a Are you in a safe place to talk right now?
1
2
3
4
9
SC2

Yes
No, call me later
No, CB on land-line
Cell phone for business only
Refused

SCHEDULE CALLBACK
RECORD NUMBER, schedule call back
THANK AND END - BUSINESS#
THANK AND END – Soft Refusal

Are you 16 years old or older?
1
2
3
9

Yes
Yes, no time
No
Refused

SCHEDULE CALLBACK
SCREEN OUT
THANK AND END - SOFT REFUSAL

Qualified Level 1
SC2a How many people age 16 and older, live in your household?
______[ENTER NUMBER 1-10]
98
99

NONE
Don’t know/Refused

SCREEN OUT SKIP TO SCR1
THANK AND END, SOFT REFUSAL

A-2

SC3

Do any other people age 16 or older regularly ANSWER your cell phone, or just you?
[INTERVIEWER: THIS QUESTION REFERS TO THE PHYSICAL PHONE AND NOT TO
THEIR CALLING PLAN]
1
2
9

Yes, others
No, just respondent
Don’t know/Refused

SKIP TO SC4
SKIP TO SC4

SC3b How many other people age 16 or older regularly answer your cell phone?
_____[ENTER NUMBER 1-10]
99
SC4

Don’t know/Refused

Not counting any that are used strictly for business purposes, are there other cell phones that you
use regularly, or is it just the one?
1
2
9

Yes, use other cell phones
No
Don’t know/Refused

SKIP TO SC5
SKIP TO SC5

SC4b How many other cell phones do you use regularly, excluding those used only for business
purposes?
_____[ENTER NUMBER 1-10]
99
SC5

Don’t know/Refused

Not counting (this/these) cell phone(s), do you also have a regular land-line phone at home?
1
2
9

Cell is only phone
Has regular phone at home
Don’t know/Refused

SKIP TO SA3
SKIP TO D13
THANK AND END, soft refusal

A-3

LAND LINE SAMPLE
SL1

Hello, I am _____ calling on behalf of the U.S. Department of Transportation. We are conducting
a national study on traffic safety.
[IF NEEDED: If you would like to learn more about the survey, you can call our toll-free number
at 1-888-772-4269 or visit the DOT website at www.nhtsa.dot.gov. Any answers you
give are kept
strictly private. It will only take about 20 minutes. The OMB number for this
solicitation is 21270613]
How many people age 16 and older, live in this household?
______[ENTER NUMBER 1-10]
98
99

NONE
Don’t know/Refused

SCREEN OUT
THANK AND END, SOFT REFUSAL

Qualified Level 1
ASK IF SL1=1.
SL1b May I speak with that person?
1
2
3
9

SKIP TO SA3
GO TO SL1d
SCHEDULE CALLBACK
THANK AND END – Soft Refusal

Rspn on line
Rspn called to phone
Rspn unavailable
Refused

ASK IF SL1>1
SL1c

In order to select just one person to interview, may I please speak to the person in your
household, age 16 or older, who (has had the most recent/will have the next) birthday?
1
2
3
9

GO TO SA3

Rspn on line
Rspn called to phone
Rspn unavailable
Refused

SCHEDULE CALLBACK
THANK AND END – Soft Refusal

SL1d Hello, I am _____ calling on behalf of the U.S. Department of Transportation We are
conducting a national study on traffic safety. Could I please confirm that you are a
household
member age 16 or older?
1
2
9

Yes
No
Refused

SCHEDULE CALLBACK
THANK AND END – Soft Refusal

SKIP TO SA3

A-4

LANDLINE OVERSAMPLE
SO1

Hello, I am _____ calling on behalf of the U.S. Department of Transportation. We are conducting
a national study on traffic safety.
[IF NEEDED: If you would like to learn more about the survey, you can call our toll-free number
at 1-888-772-4269 or visit the DOT Web site at www.nhtsa.dot.gov. Any answers you give are kept
strictly private. It will only take about 20 minutes. The OMB number for this solicitation is 2127-0613.
How many people age 16 to 34 live in this household?
_____[ENTER NUMBER 1-10]
SCREEN OUT
THANK AND END, SOFT REFUSAL

98 NONE
99 Don’t know/Refused
Qualified Level 1
ASK IF SO1=1.
SO1b May I speak with that person?
1
2
3
9

Rspn on line
Rspn called to phone
Rspn unavailable
Refused

SKIP TO SA3
GO TO SO1d
SCHEDULE CALLBACK
THANK AND END – Soft Refusal

ASK IF SO1>1
SO1c In order to select just one person to interview, may I please speak to the person in your
household, age 16 to 34 who (has had the most recent/will have the next) birthday?
1
2
3
9

Rspn on line
Rspn called to phone
Rspn unavailable
Refused

GO TO SA3
SCHEDULE CALLBACK
THANK AND END – Soft Refusal

SO1d Hello, I am _____ calling on behalf of the U.S. Department of Transportation. We are
conducting a national study on traffic safety. Could I please confirm that you are a
household member age 16 to 34?
1
2
9
SA3

Yes
No
Refused

SCHEDULE CALLBACK
THANK AND END – Soft Refusal

Record gender from observation. (Ask only if necessary)
1
2

Male
Female

Qualified Level 2

A-5

General Driving Information
1. How often do you usually drive a car or other motor vehicle? Would you say that you usually drive . . .
(NOTE: Motorcycle counts as a motor vehicle)
1
2
3
4
5
6
7

Every day, or almost every day
Several days a week
Once a week or less
Only certain times a year, OR
Never
SKIP TO D1
(VOL) Don’t know
SKIP TO D1
(VOL) Refused
SKIP TO D1

2. What kind of vehicle do you drive most often? Is it a car, van or minivan, motorcycle, SUV, pickup
truck or something else?
(NOTE: IF RESPONDENT DRIVES MORE THAN ONE VEHICLE OFTEN, ASK "What kind
of vehicle did you LAST drive?")
1
2
3
4
5
6
7
8
9

Car
Van or minivan
SUV
Pickup truck
Other truck
Motorcycle
Other (SPECIFY)
(VOL) Don’t know
(VOL) Refused

Speed Behavior
3. Which of the following statements best describes your driving? READ AND ROTATE 1&2
1
2
3
4
5

I tend to pass other cars more often than other cars pass me OR
Other cars tend to pass me more often then I pass them
(VOL) Both/About equally
(VOL) Don’t know
(VOL) Refused

4. When driving I tend to . . . READ AND ROTATE 1&2
1
2
3
4
5

Stay with slower moving traffic, or
Keep up with the faster traffic
(VOL) Both/About equally
(VOL) Don’t know
(VOL) Refused

A-6

Speed Behavior on Various Road Types
We want to find out how people may change the way they drive on different types of roads, such as multilane highways, rural routes, or residential streets. These next questions are about how you drive on some
of these different kinds of roads.
First, I am going to ask about your driving on Multi-Lane, Divided Highways. These are roads which
include interstates, freeways and other highways and have a barrier or a median separating traffic in
opposite directions.
Multi-Lane, Divided Interstate-type Highways
5a. How often do you drive on Multi-Lane, Divided Highways? Do you drive on this type of road . . .
1
2
3
4
5
6

Frequently
Sometimes
Rarely
Never
(VOL) Don’t know
(VOL) Refused

SKIP TO Q6a
SKIP TO Q6a
SKIP TO Q6a

5b. During the past seven days, approximately how many miles did you drive on Multi-Lane Divided
Highways? [IF NEEDED: Your best guess is fine.]
_________ Miles
(RANGE: 0-997, 9998, 9999)
997
9998
9999

997 miles or more
Don’t know
Refused

5c. What do you consider to be a safe speed limit for (most) Multi-Lane, Divided Highways in good
weather on roads with no congestion during the day?
____MPH
(RANGE: 0-97, 998, 999)
97 97 or more
998 (VOL) Don’t know
999 (VOL) Refused
5d. When driving on Multi-Lane, Divided Highways in good weather during the day, how fast do you
normally drive?
____MPH
(RANGE: 0-97, 998, 999)
97
998
999

97 or more
(VOL) Don’t know
(VOL) Refused

A-7

5e. How often would you say you drive 15 miles an hour over the speed limit on Multi-Lane, Divided
Highways?
1
2
3
4
5
6

Often
Sometimes
Rarely
Never
(VOL) Don’t know
(VOL) Refused

5f. How many miles per hour over the speed limit do you think the average driver can go on Multi-Lane,
Divided Highways, before he or she will receive a ticket?
____mph over the speed limit
(RANGE: 0-97, 998, 999)
97
998
999

97 or more
(VOL) Don’t know
(VOL) Refused

For this next set of questions I am going to ask you about your driving behavior on Two-Lane Highways
which are not divided. This means there is only one lane traveling in each direction and no median or
barrier separating traffic traveling in opposite directions.
Two-lane highways, one lane in each direction
6a. How often do you drive on two lane highways, one lane in each direction? Do you drive on this type
of road . . . ?
1
2
3
4
5
6

Frequently
Sometimes
Rarely, or
Never
(VOL) Don’t know
(VOL) Refused

SKIP TO Q7a
SKIP TO Q7a
SKIP TO Q7a

6b. During the past seven days, approximately how many miles did you drive on two-lane Highways, one
lane in each direction? [IF NEEDED: Your best guess is fine.]
_________ Miles
(RANGE: 0-997, 9998, 9999)
997
9998
9999

997 miles or more
Don’t know
Refused

A-8

6c. What do you consider to be a safe speed limit for (most) Two-Lane Highways, one lane in each
direction in good weather during the day?
____MPH
(RANGE: 0-97, 998, 999)
97
998
999

97 or more
(VOL) Don’t know
(VOL) Refused

6d. When driving on Two-Lane Highways, one lane in each direction in good weather during the day,
how fast do you normally drive?
____MPH
RANGE: 0-97, 998, 999
97
998
999

97 or more
(VOL) Don’t know
(VOL) Refused

6e. How often would you say you drive 15 miles an hour over the speed limit on Two-Lane Highways,
one lane in each direction?
1
2
3
4
5
6

Often
Sometimes
Rarely
Never
(VOL) Don’t know
(VOL) Refused

6f. How far above the speed limit do you think the average driver can go on Two-Lane Highways, one
lane in each direction, before he or she will receive a ticket?
____MPH over the speed limit
(RANGE: 0-97, 998, 999)
97
998
999

97 or more
(VOL) Don’t know
(VOL) Refused

A-9

Now I am going to ask you about your driving behavior on streets in neighborhoods and residential areas.
Neighborhood or Residential Streets
7a. How often do you drive on Neighborhood or Residential streets? Do you drive on this type of road . . .
1
2
3
4
5
6

Frequently
Sometimes
Rarely, or
Never
(VOL) Don’t know
(VOL) Refused

SKIP TO Q8a
SKIP TO Q8a
SKIP TO Q8a

7b. During the past seven days, approximately how many miles did you drive on Neighborhood or
Residential streets? [IF NEEDED: Your best guess is fine.]
_________ Miles
(RANGE: 0-997, 9998, 9999)
997
9998
9999

997 miles or more
Don’t know
Refused

7c. What do you consider to be a safe speed limit for (most) Neighborhood or Residential streets in good
weather during the day?
____MPH
(RANGE: 0-97, 998, 999)
97
998
999

97 or more
(VOL) Don’t know
(VOL) Refused

7d. When driving on Neighborhood or Residential streets in good weather during the day, how fast do
you normally drive?
____MPH
(RANGE: 0-97, 998, 999)
97
998
999

97 or more
(VOL) Don’t know
(VOL) Refused

A-10

7e. How often would you say you drive 10 miles an hour over the speed limit on Neighborhood or
Residential streets?
1
2
3
4
5
6

Often
Sometimes
Rarely
Never
(VOL) Don’t know
(VOL) Refused

7f. How far above the speed limit do you think the average driver can go on Neighborhood or Residential
streets, before he or she will receive a ticket?
____MPH over the speed limit
(RANGE: 0-97, 998, 999)
97
998
999

97 or more
(VOL) Don’t know
(VOL) Refused

Norms/Factors on Speeding
8. Now I'm going to read a few statements about driving and speed limits. After I read each one, please
tell me whether you agree, disagree, or neither. (READ ITEM). Would you say you strongly
(AGREE/DISAGREE) or somewhat (AGREE/DISAGREE)?
a.
b.
c.
d.
e.
f.
g.

Everyone should obey the speed limits because it’s the law.
People should keep pace with the flow of traffic.
Speeding tickets have more to do with raising money than they do with reducing speeding.
Driving over the speed limit is not dangerous for skilled drivers.
There is no excuse to exceed the speed limits.
It is unacceptable to exceed speed limits by more than 20 mph.
If it is your time to die, you’ll die, so it doesn’t matter whether you speed.

1
2
3
4
5
6
7

Strongly agree
Somewhat agree
Neither
Somewhat disagree
Strongly disagree
(VOL) Don’t know
(VOL) Refused

A-11

9. Now I'm going to read a few statements. After I read each one, please tell me whether you agree,
disagree, or neither. (READ ITEM). Would you say you strongly (AGREE/DISAGREE) or
somewhat (AGREE/DISAGREE)?
a.
b.
c.
d.
e.
f.
g.

I enjoy the feeling of driving fast.
The faster I drive, the more alert I am.
I often get impatient with slower drivers.
I try to get where I am going as fast as I can.
I worry a lot about having a crash.
I consider myself a risk taker while driving.
Speeding is something I do without thinking.

1
2
3
4
5
6
7

Strongly agree
Somewhat agree
Neither
Somewhat disagree
Strongly disagree
(VOL) Don’t know
(VOL) Refused

10. People sometimes go faster than the speed limit for different reasons. On those occasions when you
do, what do you think are the main reasons you drive faster than the speed limit? Anything else?
MULTIPLE RECORD. DO NOT READ.
1
2
3
4
5
6
7
8
9
10
11

I’m late
I am unlikely to have a crash
It’s a habit
I’m alone in the car
I’m unlikely to get a ticket
People I am with encourage it
I’m comfortable driving fast
Other, Specify
(VOL) I never speed
(VOL) Don’t know
(VOL) Refused

A-12

11. Now I'm going to read a few statements. After I read each one, please tell me whether you agree,
disagree, or neither. (READ ITEM). Would you say you strongly (AGREE/DISAGREE) or
somewhat (AGREE/DISAGREE)?
Driving at or near the speed limit . . .
a.
b.
c.
e.
f.

Reduces my chances of an accident
Makes it difficult to keep up with traffic
Makes me feel annoyed
Makes it easier to avoid dangerous situations
Uses less fuel

1
2
3
4
5
6
7

Strongly agree
Somewhat agree
Neither
Somewhat disagree
Strongly disagree
(VOL) Don’t know
(VOL) Refused

Attitudes Toward Enforcement
12. How important is it that something be done to reduce speeding by drivers? Is it . . .
1
2
3
4
5
6

Very important
Somewhat important
Not too important
Not at all important
(VOL) Don’t know
(VOL) Refused

13. How often do you think police should enforce the speed limit? Should they enforce it . . .
1
2
3
4
5
6
7

All the time
Often
Sometimes
Rarely, or
Never
(VOL) Don’t know
(VOL) Refused

A-13

14. How often do you see motor vehicles that have been pulled over by police on the streets and roads
you normally drive? Do you see motor vehicles pulled over . . . READ LIST
1
2
3
4
5
6
7

All the time
Often
Sometimes
Rarely
Never
(VOL) Don’t know
(VOL) Refused

Automated Photo Enforcement Devices
The next questions are about speed cameras. These are cameras set up at intersections or other locations
to take pictures of speeding vehicles. A traffic ticket is mailed to the owner of the vehicle along with a
photograph and information about the location and time.
15. Before today, have you ever heard of speed cameras being used to ticket drivers who speed?
1
2
3
4

Yes
No
(VOL) Don’t know
(VOL) Refused

16. Thinking about locations where speed cameras might be useful, would you find it acceptable to use
them . . . ? READ AND ROTATE A-F
A.
B.
C.
D.
E.
F.

Where it could be hazardous for a police officer to stop a driver
Where stopping a vehicle could cause traffic congestion
Where there have been many crashes
In a school zone
In a construction zone
On all roads

1
2
3
4

Yes
No
(VOL) Don’t know
(VOL) Refused

IF Q15 DOES NOT EQ 1, SKIP TO Q19
17. Along the routes you normally drive, are there speed cameras in use?
1
2
3
4

Yes, they are being used
No, there are no speed cameras along these routes
(VOL) Don’t know
(VOL) Refused

A-14

18. Have you ever received a ticket in the mail for a speed violation, identified by a speed camera?
1
2
3
4

Yes
No
(VOL) Don’t know
(VOL) Refused

19. Now I'm going to read a few statements. After I read each one, please tell me whether you agree,
disagree, or neither. (READ ITEM). Would you say you strongly (AGREE/DISAGREE) or
somewhat (AGREE/DISAGREE)?
a. Speed cameras are used to prevent accidents
b. Speed cameras are used to generate revenue
1
2
3
4
5
6
7

Strongly agree
Somewhat agree
Neither
Somewhat disagree
Strongly disagree
(VOL) Don’t know
(VOL) Refused

Attitudes Toward Speeding Countermeasures
20. How would you feel about using the following measures in your community to reduce speeding?
Please tell me whether you think each of the following is a good idea or a bad idea.
a.
b.
c.
d.
e.
f.

More frequent ticketing for speeding
Issuing higher fines for speeding tickets
Increasing public awareness of the risks of speeding
Road design changes, like speed humps and traffic circles, to slow down traffic
Electronic signs by the road that warn drivers that they are speeding and should slow
Increased use of speed cameras in dangerous or high crash locations

1
2
3
4
5

Good idea
Neither a good or bad idea
Bad idea
(VOL) Don’t know
(VOL) Refused

A-15

down

There are a number of new technologies in use to reduce the amount of speeding on our nation’s roads.
These next questions ask what you think about the use of these technologies to reduce speeding.
21. A speed governor is a device which does not allow the vehicle to go above a certain speed. Do you
think the mandatory use of a speed governor is a good idea or a bad idea for . . . . ?
a.
b.
c.
d.

Truck drivers
Drivers 18 years or younger
Drivers with multiple speeding tickets in one year
All drivers

1
2
3
4
5

Good idea
Neither a good or bad idea
Bad idea
(VOL) Don’t know
(VOL) Refused

22.

Please tell me whether you think each of the following is a good idea or a bad idea to help reduce
speeding?
a. A device in your motor vehicle that notifies you with a buzzer or a flashing light when you
drive faster than the speed limit
b. A device in your motor vehicle which records your speed data and gives you the option to
provide the information to your insurance company to lower your premiums, if you obey the
speed limits
c. A device in your motor vehicle, which slows the motor vehicle down when it senses
another
car or object is too close to your motor vehicle
1
2
3
4
5

22a.
1
2
3
4

Good idea
Neither a good or bad idea
Bad idea
(VOL) Don’t know
(VOL) Refused
Would it prevent you from speeding?
Yes
No
Not sure
(VOL) Refused

A-16

23. Now I'm going to read a few statements. After I read each one, please tell me whether you would be
likely, unlikely, or neither to use the following devices in your own vehicle(s). (READ ITEM).
Would you say you would be very (LIKELY/UNLIKELY) or somewhat (LIKELY/UNLIKELY) to
use this device?
A. A device in your motor vehicle that does not allow you to drive faster than 10 miles over the
posted speed limit.
B. A device in your motor vehicle that you can switch on or off, that prevents you from
driving
faster than the speed limit
C. A device in your motor vehicle which allows parents to limit the maximum speed of the motor
vehicle, when the teenager drives the motor vehicle
1
2
3
4
5
6
7

Very likely
Somewhat likely
Neither
Somewhat unlikely
Very unlikely
(VOL) Don’t know
(VOL) Refused

24. Some roadways use digital signs to change the speed limit on a section of road based on traffic or
weather conditions. Do you think it is a good idea or a bad idea to use these signs in the following
situations:
A.
B.
C.
D.

Construction zones
School zones
Bad weather
Congested Roadways

1
2
3
4
5

Good idea
Neither a good or bad idea
Bad idea
(VOL) Don’t know
(VOL) Refused

Crash Experience
25. How many times have you been in a speeding related accident in the past five years?
__________ TIMES
(RANGE: 0-30)
98 (VOL) Don’t know
99 (VOL) Refused
IF Q25=0, SKIP TO Q30

A-17

26. How long ago was the most recent accident?
__________
98 (VOL) Don’t know
99 (VOL) Refused
1
2
3
4

ENTER RESPONSE IN DAYS
ENTER RESPONSE IN WEEKS
ENTER RESPONSE IN MONTHS
ENTER RESPONSE IN YEARS

27. Did you receive any injuries as a result of the most recent speeding related accident?
1
2
3
4

Yes
No
(VOL) Don’t know
(VOL) Refused

SKIP TO Q30
SKIP TO Q30
SKIP TO Q30

28. Did your injuries require you to go to the hospital?
1
2
3
4

Yes
No
(VOL) Don’t know
(VOL) Refused

SKIP TO Q30
SKIP TO Q30
SKIP TO Q30

29. How long did you stay in the hospital?
__________ DAYS
(RANGE 0-97)
0 Less than 1 day
98 (VOL) Don’t know
99 (VOL) Refused
Personal Sanctions
30. In the past TWELVE MONTHS have you been STOPPED for speeding by the police?
1
2
3
4

Yes
No
(VOL) Don’t know
(VOL) Refused

SKIP TO Q34
SKIP TO Q34
SKIP TO Q34

A-18

31. How many times have you been stopped for speeding in the past twelve months?
_______ TIMES STOPPED
(Range = 0 to 7)
8
9
32a.
1
2
3
4
32b.
1
2
3
4

(VOL) Don’t know
(VOL) Refused
Did you receive a ticket during the last time you were stopped for speeding?
Yes
No
(VOL) Don’t know
(VOL) Refused)

SKIP TO Q33

Did you receive a warning the last time you were stopped for speeding?
Yes
No
(VOL) Don’t know
(VOL) Refused)

SKIP TO Q34
SKIP TO Q34
SKIP TO Q34

33. Did you change your driving behavior as a result of receiving the (TICKET/WARNING) for
speeding?
1
2
3
4

Yes
No
(VOL) Don’t know
(VOL) Refused

Other Risky Behaviors
34. When driving your primary vehicle how often do you wear your seatbelt?
1
2
3
4
5
6
7

All of the time
Most of the time
Some of the time
Rarely
Never
(VOL) Don’t know
(VOL) Refused

35. In the past 30 days, have you driven a vehicle when you thought you might have consumed too much
alcohol to drive safely?
1
2
3
4

Yes
No
(VOL) Don’t know
(VOL) Refused

A-19

Use of Cell Phone Behaviors
36. When you drive a motor vehicle, do you usually have a cell phone or wireless phone of some type in
the vehicle with you?
1
2
8
9

Yes
No
(VOL) Don’t know
(VOL) Refused

SKIP TO D1

37. How often do you talk on the phone while you are driving? Would you say you talk on the phone
while driving during . . . ?
1
2
3
4
5
8
9

All trips
Most trips
About half your trips
Fewer than half your trips, or
None of your trips
(VOL) Don’t know
(VOL) Refused

SKIP TO Q39

38. When you are talking on the phone while driving, do you tend to …?
1
2
3
4
5
6
8
9

Hold the phone in your hand
Squeeze the phone between your ear and shoulder
Use a hands–free earpiece
Use a built-in-car system (OnStar, Sync, or built-in Bluetooth)
Use the cellular phone’s speakerphone feature
Varies
(VOL) Don’t know
(VOL) Refused

39. How often do you read OR send text messages while you are driving and the vehicle is moving?
Would you say you read OR send text messages while driving during ...?
1
2
3
4
5
8
9

All trips
Most trips
About half your trips
Fewer than half your trips, or
None of your trips
(VOL) Don’t know
(VOL) Refused

A-20

Demographics
Now, a few last questions for statistical purposes . . .
D1. How old are you?
AGE IN YEARS: ________
99 Refused (VOL)
D2. Are you currently employed full time, part time, unemployed and looking for work, retired, going to
school, a homemaker, or something else? SINGLE RECORD
1
2
3
4
5
6
7
8
9
10

Employed full time
Employed part time
Unemployed and looking for work
Retired
Going to school
Homemaker
(VOL) Disabled
Other (SPECIFY)
(VOL) Not sure
(VOL) Refused

D3. What is highest grade or year of regular school you have completed? DO NOT READ
1
2
3
4
5
6
7
8
9
10

No formal schooling
First through 7th grade
8th grade
Some high school
High school graduate
Some college
Four-year college graduate
Some graduate school
Graduate degree
(VOL) Refused

D4. Are you currently married, divorced, separated, widowed, or single?
1
2
3
4
5
6
7

Married
Divorced
Separated
Widowed
Single
(VOL) Don’t know
(VOL) Refused

A-21

D5. Do you consider yourself to be Hispanic or Latino?
1
2
3
4

Yes
No
(VOL) Don’t know
(VOL) Refused

D6. Which of the following racial categories describes you? You may select more than one.
READ LIST AND MULTIPLE RECORD
1
2
3
4
5
6
11
12

American Indian or Alaska Native
Asian
Black or African-American
Native Hawaiian or Other Pacific Islander
White
(VOL) Hispanic/Latino
(VOL) Other (SPECIFY)
(VOL) Refused

ASK IF D5=2 AND D6=6
D6a.
1
2
3
4
D7a.

Just to confirm, do you consider yourself to be Hispanic or Latino?
Yes
No
(VOL) Don’t know
(VOL) Refused
How many persons live in your household?

(NOTE: This includes children under the age of 16.)
______ persons
98 (VOL) Don’t know
99 (VOL) Refused
IF D7a=1, SKIP TO D8
D7b.

How many persons live in your household who are under 16 years old?

______ persons under 16
00 None
98 (VOL) Don’t know
99 (VOL) Refused

A-22

D8. Do you own or rent your home?
1
2
3
4
5

Own
Rent
Some other arrangement
(VOL) Don’t know
(VOL) Refused

D9. Which of the following categories best describes your total household income before taxes in 2010?
Your best estimate is fine. READ LIST
1
2
3
4
5
6
7
8
9

Less than $5,000
$5,000 to $14,999
$15,000 to $29,999
$30,000 to $49,999
$50,000 to $74,999
$75,000 to $99,999
$100,000 or more
(VOL) Not sure
(VOL) Refused

D10. Let me just confirm that the number I reached you at was: [qphone] READ PHONE NUMBER
1
2
3

Yes
No
(VOL) Refused

D10a.May I please have your zip code?
ENTER ZIP CODE: _________
98 (VOL) Don’t know
99 (VOL) Refused
IF DK OR REF IN D10a, ASK D10b
D10b. Do you live in a rural, suburban, or urban area?
1
2
3
4
5
6

Rural
Suburban
Urban
(VOL) Other (Specify)
(VOL) Don’t know
(VOL) Refused

A-23

ASK ONLY FOR LANDLINE SAMPLE
D11.Is this the only telephone number for this household?
1
2
9

Yes, this is the only number
No, there is more than one number
(Don’t know/Refused)

ASK ONLY FOR LANDLINE SAMPLE
D12.Do you have a cell phone in addition to the line we are speaking on right now?
1
2
9

This is only phone
Also has cell phone
(Don’t know/Refused)

CELL SAMPLE ONLY: SKIP TO C1
ASK ONLY IF (SC5=2) OR (D12=2)
D13. Of all the telephone calls that you or your family receives, are . . . (Read List)
1
2
3
8
9

All or almost all calls received on cell phones
Some received on cell phones and some on regular phones
(IF CELL: SCRN OUT: NOT CELL MOSTLY) SKIP TO SCR1
Very few or none on cell phones
(IF CELL: SCRN OUT: NOT CELL MOSTLY) SKIP TO SCR1
(VOL) Don’t know
(IF CELL: SCRN OUT: NOT CELL MOSTLY) SKIP TO SCR1
(VOL) Refused
IF CELL: SCRN OUT: NOT CELL MOSTLY) SKIP TO SCR1

ASK ONLY IF (SC5=2) OR (D12=2)
D14.Thinking about just your LAND LINE home phone, NOT your cell phone, if that telephone rang
when someone was home, under normal circumstances, how likely is it that the phone would be
answered? Would you say it is … (Read List)
1
2
3
4
5
8
9

Very likely the land line phone would be answered,
Somewhat likely,
Somewhat unlikely,
Very Unlikely, or
Not at all likely the land line phone would be answered
(VOL) Don’t know
(VOL) Refused

A-24

CELL SAMPLE ONLY: GO TO SA3
FOR LANDLINE AND LANDLINE OVERSAMPLE ONLY
Those are all the questions I have for you. Thank you for your participation.
FOR CELL SAMPLE ONLY
C1. May I please have your name, street address, city, and state and zipcode so I can send you your $10
incentive check?
ENTER NAME:
ENTER ADDRESS:
ENTER CITY:
ENTER STATE:
ENTER ZIP:
Those are all the questions I have for you. Thank you for your participation.
SCR1. I am sorry but you are not eligible to participate in the survey today. Thank you for your
cooperation and I hope you have a pleasant evening.

A-25

APPENDIX B
SURVEY METHODOLOGY

Methodology for the 2011 National Survey of Speeding Attitudes and Behavior
The goal of the 2011 National Survey of Speeding Attitudes and Behavior was to obtain a
“snapshot” of the attitudes and behaviors regarding speeding of the population of drivers in the
United States using a telephone survey of U.S. drivers 16 years and older. Only surveys based on
probability samples can be used to create mathematically sound statistical inferences about a
larger target population. Most statistical formulas for specifying the sampling precision
(estimates of sampling variance), given particular sample sizes, are premised on simple random
sampling. However, random sampling requires an enumeration of all of the elements in the
population. Since no enumeration of the total population of the United States (or its subdivisions)
is available, all surveys of the general public are based upon complex sample designs that may
employ stratification and two or more stages of sampling.
A sampling design using geographic stratification, an oversample of young drivers, sampling
frames of households with landlines and cell phones, together with an overall sample size of
6,000 was developed and implemented for this survey. The final sample consisted of 6,144
respondents, which included an independent cell phone sample of cellphone only and cell phone
mostly households as well as an oversample of 500 drivers 16 to 34. Weights were developed to
yield national estimates of the target population within specified limits of expected sampling
variability. This appendix describes the methods of sample construction and survey
administration, and shows the sample disposition and computation of weights.
Sample Construction
Strata - The initial stage in the construction of this sample required the development of a
national probability sample of the non-institutionalized population of the United States 16 and
older. Stratification (i.e., division of the population into collectively exhaustive and mutually
exclusive homogenous groups), an efficient way of achieving high statistical precision with a
smaller overall sample size, was employed. NHTSA has 10 regional offices with each regional
office providing services to the States within its Region. Therefore, for the sample, the country
was stratified into 10 strata, each consisting of the States in NHTSA’s 10 Regions.
The estimated distribution of the target population by stratum was calculated on the basis of the
U.S. Census Bureau, Population Estimates by State by Single Year of Age, Sex, Race, and
Hispanic Origin: 2008. The population estimates were taken for the population 16 and older.
Based on these Census estimates of the geographic distribution of the target population, the total
sample was proportionately allocated by stratum.
Oversample of respondents 16 to 34 - Given the overrepresentation of young drivers in traffic
crashes, it was very important that the subsample of drivers 16 to 34 years old in this survey be
large enough for meaningful statistical analysis. However, the population prevalence of this age
group was not large enough to generate the desired sub-sample size, given a total sample of
6,000 for the survey, so an oversample was included. Based on year 2008 Census Bureau
estimates of the civilian non-institutionalized population, we estimated that in a population-based
sample, about 33% of drivers should be 16 to 34. Our experience with recent telephone surveys
using only conventional random digit dialing (RDD) of landline households indicates that the
subsample of respondents 16 to 34 obtained by this method would fall short of the desired 33
B-2

percent of the total sample. For example, in the 2007 Motor Vehicle Occupant Safety Survey
(MVOSS) that relied on RDD of landline phones, respondents 16 to 34 made up only 18 % of
the entire sample.
Table B.1 shows the national population figures and projected sample distribution by age for the
total sample of 6,000 respondents. The fourth column shows the desired sample from a
population-based sample, and the last two columns show what could be expected from a
conventional RDD landline approach, such as that used in the MVOSS 2007 study.
Table B.1.
EXPECTED POPULATION AND SAMPLE DISTRIBUTION**
BASED ON June 1, 2008 CENSUS BUREAU ESTIMATES

AGE

Sample Distribution
Target Population

Population
based

Expected Based on
MVOSS response

(N in 1000s)

Total (16+)

233,627

100%

6,000

100%

16-24

37,476

16.0%

962

366

6.1%

25-34

39,960

17.1%

1,026

732

12.2%

35-44

41,735

17.9%

1,072

1,086

18.1%

45-64

77,397

33.1%

1,988

2,406

40.1%

65+

37,060

15.9%

952

1,410

23.5%

2007

U.S. Bureau of the Census, Population Estimates, Age Category Estimates, 6/01/08
Source: www.census.gov/popest/national/asrh/files/NC-EST2007-ALLDATA-N-File19.csv
** Sample distribution from MVOSS 2007 with RDD landline survey
The reasons for this discrepancy include a lower response rate among younger adults, a higher
proportion of persons 16 to 34 living in group quarters (e.g., dormitories), and a higher
proportion of this age group living in cell phone only households. Hence, a simple proportionate
sample of the adult driver population based on RDD landline methodology would not meet the
needs of this study design. Consequently, an oversample of 500 respondents 16 to 34 was
included in the sample design at the start of the study.
Landline and Cell Phone RDD samples - As noted above, RDD landline telephone sampling
has been the conventional approach for conducting surveys of the U.S. household population for
the past few decades. However, households are increasingly turning to cell phones, and many
households have abandoned landline phones altogether. For example, in the second half of 2010,
the percentage of cell phone only households (households with no landline, but accessible by cell

B-3

phone) was 29.7 percent, according to the National Health Interview Survey (Blumberg & Luke,
2012). Current RDD landline sampling procedures exclude telephone exchanges and banks of
telephone numbers used exclusively for cell phones. This makes it difficult to reach people in
subpopulations with high cell phone only usage. For example, almost 7 out of 10 (69.4%) adults
living with unrelated roommates and over half (53.5%) of adults 25 to 29 years old live in cellphone-only households. These are some of the same groups that are increasingly underrepresented in conventional RDD landline telephone surveys. As the percentage of cell-phoneonly households continues to grow, the conventional RDD landline sampling model can no
longer reliably provide adequate population coverage required for sampling the U.S. household
population. To overcome this challenge and to account for drivers that rely solely or mostly on
cell phones, this survey used both a RDD sample of landline phones and a RDD sample of cell
phones.
Cell Phone Households - A stratified random sample of cellular phone numbers was drawn and
used to contact potential respondents. This was feasible because the 10 strata used in this study
are defined in terms of states and cellular phone codes are also defined by states. However, cell
phones are portable and some respondents could be living in states other than that indicated by
their cell phone area code. To address this possible scenario, all cell phone respondents were
asked their address so they could be classified into one of the NHTSA regions.
Two types of cell phone households were identified through screening: cell phone only
households and cell phone mostly households. Cell phone only households do not have a
landline phone. Cell phone mostly households have both landline and cellular telephone service
(dual service), but the landline is not often used for receiving calls and, therefore, the probability
of reaching such a household through the landline sample is greatly diminished. Because cell
phone mostly households are also included in the sample frame of land line households, the
estimation procedures that account for the overlapping dual service sample are more complicated
than those that use non overlapping (mutually exclusive) samples of cell phone only households
and landline households (with or without cell phone). Indeed, most surveys conducted to date
with cell phone samples used strictly cell phone only households. However, it was important to
include the cell phone mostly households in the study sample for the representativeness of the
population and to capture respondents in the critical group of 16 to 34-year-olds.
Cell phones were treated as personal devices and only the person with the cell phone was
screened for eligibility. A $10 incentive was offered to respondents to complete an interview via
their cell phone A total of 783 interviews were conducted with respondents from cell phone only
households and 354 interviews with respondents from cell phone mostly households. The
number of interviews to be achieved for these groups was derived using a formula (Cochran,
1977) for the optimal allocation to strata when unit costs differ between the strata. A check was
mailed to the respondents who accepted the incentive and provided a complete mailing address
within 10 business days after the interview was completed.
Landline Households - A stratified sample of landline telephone numbers was drawn and
potential respondents were contacted using conventional RDD methods. The households were
screened for eligibility, and an eligible driver was selected for the interview. Landline
respondents were not offered any incentives. A total of 5,007 interviews were conducted with
respondents from the landline sample. This includes the oversample of 500 respondents 16 to 34.

B-4

Table B.2 shows the number of interviews from each sample type by age. Age quotas were not
used during data collection except for the 500 person land line oversample for the 16 to 34-yearold group.
Table B.2. Sample Size by Type and Age
Age

Landline

Landline
Oversample

Cell
Only

Phone Cell Phone
TOTAL
Mostly

16-34

472

500

424

119

1,515

35+

3,970

355

235

4,560

Not Reported

TOTAL

4,507

500

783

354

6,144

Survey Administration
The objective of survey administration is to conduct the data collection portion of the survey in a
systematic, uniform and consistent manner. Survey administration includes the pretest of the
instrument and survey procedures, monitoring of the interviews, and tracking of the sample
disposition.
Cognitive Testing
On December 10, 2009, two interviewers conducted nine cognitive interviews at Abt Associates’
Bethesda Cognitive Testing Laboratory (CTL) with licensed drivers from the Washington, DC,
area. There was a mix of respondents by age, education, and gender. Each respondent signed an
informed consent form and was paid $75 in appreciation for participation in the pretest.
The cognitive interview protocol consisted of a description of the cognitive interviews’ general
objective to identify question flaws that may affect the validity or reliability of answers,
instructions to the respondent, and guidelines for the cognitive interviewer. Respondents were
asked to think aloud during the interview, saying what they were thinking as they answered the
survey questions and to also volunteer any additional comments about the clarity or other aspects
of the questions. In addition, the interviewers asked follow-ups to some of the survey questions
to determine details about the response process, and to check on the presence of potential
problems noted when reviewing the draft instrument.
Pretest of CATI Instrument
Once the questions for the survey instrument were developed, the Computer Assisted Telephone
Interview (CATI) instrument was programmed. Interviewers were briefed about the survey, the
questionnaire and trained on the interviewing procedures. A survey pretest was conducted using
the CATI programs and interviewed 34 respondents from the target population. The pretest was
conducted over two evenings, and was monitored by NHTSA staff and the project director. The
purpose of the pretest was to ensure all of the interviewing systems were working properly and to
B-5

also test the survey instrument to ensure that the respondents did not have any trouble
understanding the questions or the language.
Calling Protocol
The calling protocol used in this study consisted of a maximum of 15 attempts for the land line
sample, including the oversample of drivers 16 to 34. If someone in the household was contacted
on one of these attempts, then the overall maximum attempts for that household was 25. For the
cell phone sample, the maximum number of attempts to reach someone was 10. If contact was
made during one of those 10 attempts with someone in that household, then the maximum
number of attempts was set at 20.
If a person selected for the sample refused to participate in the survey and was classified as a
Soft Refusal, he or she was re-contacted approximately one to two weeks after the initial refusal,
giving them a “cooling off” period before the re-contact.
Spanish Language Interviewing
A Spanish language version of the survey instrument was developed in order to eliminate
language barriers for a small proportion of the U.S. adult population. The questionnaire was
translated into Spanish by a professional translation firm. The Spanish questionnaire was then
reviewed next to the English questionnaire by a different translator and checked for errors. Any
translations that were not comparable were revised to be in line with the intent of the English
questionnaire.
If the interviewer encountered a language barrier during the initial contact, either with the person
answering the phone or with the designated respondent, the interviewer thanked the person and
terminated the call. If the case was designated as Spanish language, it was turned over to the next
available Spanish-speaking interviewer.
All households which were designated as “Foreign Language-Spanish” were assigned to a
Spanish-speaking interviewer. These bilingual interviewers re-contacted each Spanish-speaking
household to screen for eligibility and conducted the interview with the target respondent.
Monitoring of Telephone Interviewers
For quality control, the telephone interviews were monitored by field supervisory staff using a
silent line and screen monitoring.
Answering Machines
The strategy for handling answering machines with a 20- or 25-call protocol has to balance the
objectives of reaching the household and avoiding annoyance of the household. Thus, messages
were left on the answering machine or voice mail on the fifth, seventh and ninth calls, if an
answering machine or voice mail was encountered on those attempts. The first answering
machine message explained that the household had been selected as part of a USDOT study of

B-6

American driving habits and attitudes, and asked the respondent to call a toll-free number to
schedule an interview. The subsequent answering messages also included this information.
Follow-Up Letter for Refusals, Non-Contacts and Callbacks
A quasi experiment was performed to test the effectiveness of a follow-up letter in obtaining a
response. Follow-up letters were sent to 1,000 people who did not respond to the telephone
interview by the tenth contact attempt, regardless of whether it is a non-contact, callback or
refusal. The telephone numbers of these non-respondents were matched to an address database
(with a 60% match rate), and letters were sent asking them to call the toll-free survey number
and complete the survey. The follow-up letter did not have an effect on the refusal conversion or
completion rate when we compared those who were sent a letter to those who were not sent a
letter. Figure B1 shows the follow up letter.

B-7

Figure B.1. Follow up Letter
DATE

PIN #: PINNUM

NAME
ADDRESS 1
ADDRESS 2
CITY, ST ZIP
Dear FNAME LNAME:
I am contacting you on behalf of the National Highway Traffic Safety Administration of the U.S.
Department of Transportation. We are currently conducting a national study on traffic safety and
you were selected to participate in our survey. The information you provide will be a big help to
us in improving the safety of America’s highways.
Unfortunately, we have not been able to reach you at the following number: PHONE. Please call
us at your earliest convenience to schedule your phone interview. Our toll-free number is
[redacted]. You can contact us any day of the week between the hours of 9 a.m. and 9 p.m.,
Eastern Time. Ask for extension 4548. When you contact us, you will need to provide your
personal identification number (PIN) to complete the survey. Your PIN is: [redacted].
The interview only takes 20 minutes to complete. It is voluntary and you don’t have to answer
any questions that you don’t want to answer. This study has been reviewed and approved by the
U.S. Office of Management and Budget under OMB control number 2127-0613.
Your opinions about highway safety are very important to us. The information you provide will
help the National Highway Traffic Safety Administration continue to improve motor vehicle
safety for everyone on America’s highways. Thank you in advance for your participation.
Sincerely,

Paul Schroeder
Project Director

B-8

Sample Dispositions
The final dispositions for each of the three independent samples are given in the following
tables: Table B-3: Landline Cross-Section, Table B-4: Cell Sample, and Table B-5: Landline
Oversample.

B-9

Table B-3: Final Disposition for Landline Cross-Section
Estimated Estimated
Original Qualified
Response
Count Household* Eligible
64,154

TOTAL

47,946

A1
A2
A3
A4

NON-Usable Numbers
Not in
service/Disconnected//DIS/Change#/Intercepts
Non-residential #
Computer/Fax tone
Line problem

T2
B
B1
B2
C
C1
C2
C3
D
D1
D2
D3
E
E1
E2
E3
E4

Total Usable Numbers
UNKNOWN ELIGIBLE HOUSEHOLD*^
No answer/Busy
Answering machine
NOT ELIGIBLE RESPONDENT^
Language barrier
Health/Deaf
Respondent away for duration
UNKNOWN ELIGIBLE RESPONDENT^
Callback
Spanish Callback not screened
Refusals not screened
CONTACTS SCREENED
Qualified callback
Refusals – Qualified
Terminates
Screen-outs

16,208
3,294
1,307
1,987
2,773
547
1,952
274
3,724
2,807
0
917
1,910
376
342
0
1,192

COMPLETE

ESTIMATED ELIGIBLE HH RATE =T2/T1

B'
C'
D'

39,951
4,502
2,570
923

4,507

ELIGIBLE RESPONSE RATE = E+FE4/(E+F)
SUM RESPONSE ELIGIBLE COUNT
RESPONSE RATE = F/C'

*Estimated Qualified HH=Original Count * A'
^Response Eligible=Qualified Household Count * B'
B-10

832

678

2,773

2,258

3,032

376
342
0

4,507

25.26%
81.42%
40.27%

11,193

Table B-4. Final Disposition for the Cell Phone Sample

Cell Phone Sample, 2011
T1 TOTAL
A NON-Usable Numbers
Not in Service/Disconnected
A1 /Change#/Intercepts
A2 Non-residential #
A3 Computer/Fax tone
A4 Line problem
T2 Total Usable Numbers
B UNKNOWN ELIGIBLE HOUSEHOLD*^
B1 No answer/Busy
B2 Answering machine
C NOT ELIGIBLE RESPONDENT^
C1 Language barrier
C2 Health/Deaf
C3 Respondent away for duration
D UNKNOWN ELIGIBLE RESPONDENT^
D1 Callback
D2 Spanish Callback not screened
D3 Refusals not screened
E CONTACTS SCREENED
E1 Qualified callback
E2 Refusals – Qualified
E3 Terminates
E4 Screen-outs
F
COMPLETE
A'
B'
C'
D'

ESTIMATED ELIGIBLE HH RATE =T2/T1

ELIGIBLE RESPONSE RATE = E+FE4/(E+F)
SUM RESPONSE ELIGIBLE COUNT
RESPONSE RATE = F/C'

*Estimated Qualified HH=Original Count * A'
^Response Eligible = Qualified Household Count *
B'

B-11

Original
Count
19000
7616
6101
1315
41
159
11384
765
692
73
1621
825
660
136
6526
5982
544
1336
186
45
0
1104
1137

Estimated
Qualified
Household*

Estimated
Response
Eligible

458

253

1621

896

3609

186
45

1137

59.92%
55.30%
18.55%

6127

Table B-5. Final Disposition for the Landline Oversample (Age 16 to 34)
Original
Count
55,588

T1 TOTAL
A
A1
A2
A3
A4
T2
B
B1
B2
C
C1
C2
C3
D
D1
D2
D3
E
E1
E2
E3
E4
F

NON-Usable Numbers
Not in Service/Disconnected/
Change#/Intercepts
Non-residential #
Computer/Fax tone
Line problem
Total Usable Numbers
UNKNOWN ELIGIBLE HOUSEHOLD*^
No answer/Busy
Answering machine
NOT ELIGIBLE RESPONDENT^
Language barrier
Health/Deaf
Respondent away for duration
UNKNOWN ELIGIBLE RESPONDENT^
Callback
Spanish Callback not screened
Refusals not screened
CONTACTS SCREENED
Qualified callback
Refusals – Qualified
Terminates
Screen-outs
COMPLETE**

ESTIMATED ELIGIBLE HH RATE =T2/T1

B'
C'
D'

ELIGIBLE RESPONSE RATE = E+FE4/(E+F)
SUM RESPONSE ELIGIBLE COUNT
RESPONSE RATE = F/C'

Estimated
Qualified
Household*

Estimated
Response
Eligible

40,343
34,038
3,331
2,183
791
15,245
1,950
1,565
385
916
275
548
93
3,332
2,862
21
449
8,464
188
85
0
8,191
583

535

916

188
85

583

27.42%
9.46%
44.55%

*Estimated Qualified HH=Original Count * A'
^Response Eligible = Qualified Household Count * B'
** 83 Respondents were excluded from the final sample due to the fact that they screened as
eligible but reported that their age was outside the 16 to 34 range in the demographics section.
B-12

315

1,309

Precision of Sample Estimates
The confidence interval for an estimate derived from the survey sample is:
𝑦� ± 𝑧1−𝛼⁄2 �𝑉𝑎𝑟(𝑦�)
where:

𝑦� = an estimate of the population proportion;
𝑉𝑎𝑟(𝑦�) = is the simple random sampling variance 1 of 𝑦�; and
𝑧1−𝛼⁄2 = (1 − 𝛼 ⁄2)th percentile of the standard normal distribution (95%: 𝛼 = 5%, 𝑧 =
1.96; 90%: 𝛼 = 10%, 𝑧 = 1.645).

For best results, data users should use statistical software such as SAS, SPSS, STATA or
SUDAAN to calculate the confidence intervals for a complex sampling design. However, data
users can use the tables that follow to approximate the confidence interval based on a simple
formula.

Sampling Error
The sampling variance for an estimate is a measure of uncertainty that reflects the fact that the
estimate is derived from a sample drawn from the population. If one were to draw a second
sample in the exact same manner, the estimate would be different from the first simply due to the
fact that our sample contains different members of the population. A third sample would be
different from the first two, and so on. The sampling variance measures how different the
estimates would be had we drawn different samples.
The sampling error for a complex survey depends on three things:
1. 𝜎𝑦2 =the population variance for the characteristic: the sampling variance is higher when
there is a lot of variability in the population (large 𝜎𝑦2 ) and lower when there is little
variability in the population.
2. n = The sample size: the sampling variance is higher when the sample size is small and lower
when the sample size is large. The sampling variance for estimates of subgroups is based on
the sample size for those subgroups.
3. DEFF = design effect: 2 Sampling design features such as stratification, clustering, dualframe sampling, and survey weighting all contribute to the sampling variability. The design
effect is a measure of inefficiency (or efficiency) of the complex sample relative to a simple
random sample, calculated as 𝐷𝐸𝐹𝐹 = 𝑉𝑎𝑟(𝑦�)⁄𝑉𝑎𝑟srs (𝑦�).
Using this relationship, we can write the sampling variance of the complex design as: 𝑉𝑎𝑟(𝑦�) =
𝑉𝑎𝑟srs (𝑦�) × 𝐷𝐸𝐹𝐹 = 𝜎𝑦2 ⁄𝑛 × 𝐷𝐸𝐹𝐹 . Therefore, one can calculate the sampling variance with
the population variance (or an estimate of the population variance); the sample size; and the
design effect.
1

A simple random sample is a sample on n units drawn directly from a population of N units.

Kish, L. (1965). Survey Sampling. New York: John Wiley & Sons.
B-13

Estimating the population variance
The population variance is often estimated from the survey data, 𝑠 2 = ∑𝑛(𝑦𝑖 − 𝑦�)2 ⁄𝑛. In the case of

percentages, the population variance 𝜎𝑦2 = P×(1-P) and can be estimated from the survey estimate
𝑠 2 = 𝑝̂ × (1 − 𝑝̂ ). An alternative is to use the variance estimates based on the percentages
presented in Table B.6. Rounding the estimated percentage up to the nearest 5 percentage points
(e.g., 17% to 20%, 34% to 35%) is a conservative estimate of the population variance. The
variance for a percentage is low when a small percentage of the population has the characteristic
(or a large percentage of the population has the characteristic) and high when the percentage of
the population with the characteristic is equal (50/50).
Estimating Design effects
The sampling design impacts the variance for each data item differently. Therefore the design
effect for one survey estimate might be higher or lower than the design effect of another survey
estimate. The design effect will also vary for different subpopulations represented in the sample,
such as males and females. To simplify the calculations of the sampling error, design effect
approximations are presented in Table B.6 below. These approximations are based on the
average design effect for over 100 data items.

B-14

Table B.6. Estimated 95% Error Margins Overall and Various Population Subgroups
P=
50, 50 45, 55 40, 60 35, 65 30, 70 25, 75 20, 80 15, 85 10, 90 5, 95
2
DEFF
n 𝜎 = 0.2500 0.2475 0.2400 0.2275 0.2100 0.1875 0.1600 0.1275 0.0900 0.0475
Total
NHTSA Region
1
2
3
4
5
6
7
8
9
10
Age group
16-20
21-24
25-34
35-44
45-54
55-64
65+

1.76

6144

1.7%

1.6%

1.5%

1.4%

1.3%

1.2%

1.0%

0.7%

1.72
1.77
1.65
1.74
1.64
1.67
1.90
1.63
1.74
1.64

361
868
682
856
1175
581
404
245
668
304

6.8%
4.4%
4.8%
4.4%
3.7%
5.3%
6.7%
8.0%
5.0%
7.2%

6.7%
4.4%
4.8%
4.4%
3.6%
5.2%
6.7%
7.9%
5.0%
7.2%

6.6%
4.3%
4.7%
4.3%
3.6%
5.2%
6.6%
7.8%
4.9%
7.0%

6.5%
4.2%
4.6%
4.2%
3.5%
5.0%
6.4%
7.6%
4.8%
6.9%

6.2%
4.1%
4.4%
4.1%
3.4%
4.8%
6.2%
7.3%
4.6%
6.6%

5.9%
3.8%
4.2%
3.8%
3.2%
4.6%
5.8%
6.9%
4.3%
6.2%

5.4%
3.5%
3.9%
3.5%
2.9%
4.2%
5.4%
6.4%
4.0%
5.8%

4.8%
3.2%
3.4%
3.2%
2.6%
3.8%
4.8%
5.7%
3.6%
5.1%

4.1%
2.7%
2.9%
2.7%
2.2%
3.2%
4.0%
4.8%
3.0%
4.3%

3.0%
1.9%
2.1%
1.9%
1.6%
2.3%
2.9%
3.5%
2.2%
3.1%

1.54
1.57
1.85
1.65
1.53
1.42
1.47

295
281
939
835
1,185
1,211
1,329

7.1%
7.3%
4.3%
4.4%
3.5%
3.4%
3.3%

7.0%
7.3%
4.3%
4.3%
3.5%
3.3%
3.2%

6.9%
7.2%
4.3%
4.3%
3.4%
3.3%
3.2%

6.7%
7.0%
4.1%
4.2%
3.4%
3.2%
3.1%

6.5%
6.7%
4.0%
4.0%
3.2%
3.1%
3.0%

6.1%
6.3%
3.8%
3.8%
3.0%
2.9%
2.8%

5.7%
5.9%
3.5%
3.5%
2.8%
2.7%
2.6%

5.1%
5.2%
3.1%
3.1%
2.5%
2.4%
2.3%

4.2%
4.4%
2.6%
2.6%
2.1%
2.0%
2.0%

3.1%
3.2%
1.9%
1.9%
1.5%
1.5%
1.4%

B-15

1.72
1.77

2,696
3,448

2.5%
2.2%

2.4%
2.2%

2.4%
2.1%

2.3%
2.0%

2.1%
1.9%

2.0%
1.8%

1.8%
1.6%

1.5%
1.3%

1.1%
1.0%

Race/Ethnicity
Hispanic
NH white
NH black
NH Asian
NH AIAN
NH other

1.57
1.60
1.50
1.46
1.59
1.87

440
4,750
483
134
75
164

5.9%
1.8%
5.5%
10.2%
14.3%
10.5%

5.8%
1.8%
5.4%
10.2%
14.2%
10.4%

5.7%
1.8%
5.3%
10.0%
14.0%
10.3%

5.6%
1.7%
5.2%
9.7%
13.6%
10.0%

5.4%
1.6%
5.0%
9.4%
13.1%
9.6%

5.1%
1.6%
4.7%
8.8%
12.4%
9.1%

4.7%
1.4%
4.4%
8.2%
11.4%
8.4%

4.2%
1.3%
3.9%
7.3%
10.2%
7.5%

3.5%
1.1%
3.3%
6.1%
8.6%
6.3%

2.6%
0.8%
2.4%
4.5%
6.2%
4.6%

Educational attainment
LT HS
1.44
HS grad
1.51
Some coll
1.54
Coll grad
1.51
Grad school
1.54

464
1,666
1,661
1,231
1,094

5.5%
2.9%
3.0%
3.4%
3.7%

5.4%
2.9%
3.0%
3.4%
3.7%

5.4%
2.9%
2.9%
3.4%
3.6%

5.2%
2.8%
2.8%
3.3%
3.5%

5.0%
2.7%
2.7%
3.1%
3.4%

4.7%
2.6%
2.6%
3.0%
3.2%

4.4%
2.4%
2.4%
2.7%
2.9%

3.9%
2.1%
2.1%
2.5%
2.6%

3.3%
1.8%
1.8%
2.1%
2.2%

2.4%
1.3%
1.3%
1.5%
1.6%

Driver type cluster
Speeder
Sometime Speeder
Nonspeeder

1,572
2,148
1,579

3.3%
2.7%
3.2%

3.3%
2.7%
3.1%

3.2%
2.7%
3.1%

3.1%
2.6%
3.0%

3.0%
2.5%
2.9%

2.8%
2.4%
2.7%

2.6%
2.2%
2.5%

2.3%
2.0%
2.3%

2.0%
1.6%
1.9%

1.4%
1.2%
1.4%

1.77
1.68
1.64

B-16

Table B.6. Estimated 95% Error Margins Overall and Various Population Subgroups
P=
50, 50 45, 55 40, 60 35, 65 30, 70 25, 75 20, 80 15, 85 10, 90 5, 95
2
DEFF
n 𝜎 = 0.2500 0.2475 0.2400 0.2275 0.2100 0.1875 0.1600 0.1275 0.0900 0.0475
Urban
No
1.73
Yes
1.77
Frequent driver (q1=1,2)
No
1.77
Yes
1.76
Type of vehicle
Car
Van or minivan
Pickup truck
SUV

1.78
1.69
1.65
1.77

2,903
3,030

2.4%
2.4%

2.3%
2.3%

2.2%
2.2%

2.1%
2.0%

1.9%
1.9%

1.7%
1.7%

1.4%
1.4%

1.0%
1.0%

323
5,821

7.3%
1.7%

7.2%
1.7%

7.1%
1.7%

6.9%
1.6%

6.7%
1.6%

6.3%
1.5%

5.8%
1.4%

5.2%
1.2%

4.4%
1.0%

3.2%
0.7%

3,507
563
816
1,131

2.2%
5.4%
4.4%
3.9%

2.2%
5.3%
4.4%
3.9%

2.2%
5.3%
4.3%
3.8%

2.1%
5.1%
4.2%
3.7%

2.0%
4.9%
4.0%
3.6%

1.9%
4.7%
3.8%
3.4%

1.8%
4.3%
3.5%
3.1%

1.6%
3.8%
3.1%
2.8%

1.3%
3.2%
2.6%
2.3%

1.0%
2.3%
1.9%
1.7%

B-17

Table B.7. Estimated 95% Error Margins Overall and Various Sample Sizes
P
50, 50 45, 55 40, 60 35, 65 30, 70
2
DEFF
n 𝜎 0.2500 0.247 0.240 0.227 0.210
5
0
5
0
1.76 6,000
1.7%
1.7%
1.6%
1.6%
1.5%
5,500
1.8%
1.7%
1.7%
1.7%
1.6%
5,000
1.8%
1.8%
1.8%
1.8%
1.7%
4,500
1.9%
1.9%
1.9%
1.8%
1.8%
4,000
2.1%
2.0%
2.0%
2.0%
1.9%
3,500
2.2%
2.2%
2.2%
2.1%
2.0%
3,000
2.4%
2.4%
2.3%
2.3%
2.2%
2,500
2.6%
2.6%
2.5%
2.5%
2.4%
2,250
2.7%
2.7%
2.7%
2.6%
2.5%
2,000
2.9%
2.9%
2.8%
2.8%
2.7%
1,750
3.1%
3.1%
3.0%
3.0%
2.8%
1,500
3.4%
3.3%
3.3%
3.2%
3.1%
1,250
3.7%
3.7%
3.6%
3.5%
3.4%
1,000
750
500
400
300
200
150
100
50

4.1%
4.7%
5.8%
6.5%
7.5%
9.2%
10.6%
13.0%
18.4%

4.1%
4.7%
5.8%
6.5%
7.5%
9.1%
10.6%
12.9%
18.3%

4.0%
4.7%
5.7%
6.4%
7.4%
9.0%
10.4%
12.7%
18.0%

B-18

3.9%
4.5%
5.5%
6.2%
7.2%
8.8%
10.1%
12.4%
17.5%

3.8%
4.4%
5.3%
6.0%
6.9%
8.4%
9.7%
11.9%
16.9%

25, 75
0.187
5
1.5%
1.5%
1.6%
1.7%
1.8%
1.9%
2.1%
2.3%
2.4%
2.5%
2.7%
2.9%
3.2%

20, 80
0.160
0
1.3%
1.4%
1.5%
1.6%
1.6%
1.8%
1.9%
2.1%
2.2%
2.3%
2.5%
2.7%
2.9%

15, 85
0.127
5
1.2%
1.3%
1.3%
1.4%
1.5%
1.6%
1.7%
1.9%
2.0%
2.1%
2.2%
2.4%
2.6%

10, 90
0.0900
1.0%
1.1%
1.1%
1.2%
1.2%
1.3%
1.4%
1.6%
1.6%
1.7%
1.9%
2.0%
2.2%

5, 95
0.047
5
0.7%
0.8%
0.8%
0.8%
0.9%
1.0%
1.0%
1.1%
1.2%
1.3%
1.4%
1.5%
1.6%

3.6%
4.1%
5.0%
5.6%
6.5%
8.0%
9.2%
11.3%
15.9%

3.3%
3.8%
4.7%
5.2%
6.0%
7.4%
8.5%
10.4%
14.7%

2.9%
3.4%
4.2%
4.6%
5.4%
6.6%
7.6%
9.3%
13.1%

2.5%
2.8%
3.5%
3.9%
4.5%
5.5%
6.4%
7.8%
11.0%

1.8%
2.1%
2.5%
2.8%
3.3%
4.0%
4.6%
5.7%
8.0%

Testing for Statistical Differences
Sampling error is also used to determine whether two population subgroups (or domains) are
significantly different with respect to a certain statistic, that is, the difference in the sampled
subgroup estimates is large enough that it would be unlikely to randomly occur if the statistics
were the same for the subgroups. Consider the hypothesis test for comparing two domains:
H0: Y1 = Y2 or Y1 – Y2 = 0
H1: Y1 ≠ Y2 or Y1 – Y2 ≠ 0
One method to test whether Y1 is different from Y2 is to calculate a confidence interval around
the difference in the sample estimates, 3 (𝑦�1 − 𝑦�2 )±𝑧1−𝛼⁄2 �𝑉𝑎𝑟(𝑦�1 − 𝑦�2 ). If the interval does
not contain 0, we conclude that Y1 is different from Y2 –the observed difference in the sample
estimates is not likely to randomly occur if Y1 was equal to Y2, therefore there is evidence to
indicate a difference in the population statistics. If the interval does contain 0, we cannot
conclude that Y1 is different from Y2 – there is insufficient evidence to indicate a difference in
the population statistics.
𝑉𝑎𝑟(𝑦�1 − 𝑦�2 ) = 𝑉𝑎𝑟(𝑦�1 ) + 𝑉𝑎𝑟(𝑦�2 ), the sum of the variances for two population subgroups.
The subgroup variances are estimated as described above. Table B.8 includes the estimated 95%
error margins for the differences between subgroups of various size. If the observed difference is
less than or equal to the error margin, the difference is not statistically significant at the α = 0.05
significance level. If it is greater than the error margin, the difference is statistically significant at
the α = 0.05 significance level.

This method should only be used for large sample sizes. One rule of thumb is n1 and n2 both greater than 30.

B-19

Table B.8. Estimated 95% Error Margins for the Difference Between Two Subgroups
DEFF
n1
P n2 = 6000 5000 4000 3000 2000 1500 1000 500
400
300
200
1.76
6,000 50,50
2.4% 2.5% 2.7% 2.9% 3.4% 3.8% 4.4% 6.1% 6.7% 7.7% 9.3%
40,60
2.3% 2.4% 2.6% 2.8% 3.3% 3.7% 4.4% 5.9% 6.6% 7.5% 9.2%
30,70
2.2% 2.3% 2.4% 2.7% 3.1% 3.4% 4.1% 5.5% 6.2% 7.0% 8.6%
20,80
1.9% 2.0% 2.1% 2.3% 2.7% 3.0% 3.6% 4.8% 5.4% 6.2% 7.5%
10,90
1.4% 1.5% 1.6% 1.7% 2.0% 2.3% 2.7% 3.6% 4.0% 4.6% 5.6%
5,000 50,50
2.5% 2.6% 2.8% 3.0% 3.4% 3.8% 4.5% 6.1% 6.8% 7.7% 9.4%
40,60
2.4% 2.5% 2.7% 2.9% 3.4% 3.8% 4.4% 6.0% 6.6% 7.6% 9.2%
30,70
2.3% 2.4% 2.5% 2.8% 3.2% 3.5% 4.1% 5.6% 6.2% 7.1% 8.6%
20,80
2.0% 2.1% 2.2% 2.4% 2.8% 3.1% 3.6% 4.9% 5.4% 6.2% 7.5%
10,90
1.5% 1.6% 1.7% 1.8% 2.1% 2.3% 2.7% 3.7% 4.1% 4.6% 5.6%
4,000 50,50
2.7% 2.8% 2.9% 3.1% 3.6% 3.9% 4.6% 6.2% 6.8% 7.8% 9.4%
40,60
2.6% 2.7% 2.8% 3.1% 3.5% 3.9% 4.5% 6.0% 6.7% 7.6% 9.2%
30,70
2.4% 2.5% 2.7% 2.9% 3.3% 3.6% 4.2% 5.7% 6.2% 7.1% 8.6%
20,80
2.1% 2.2% 2.3% 2.5% 2.8% 3.1% 3.7% 4.9% 5.5% 6.2% 7.5%
10,90
1.6% 1.7% 1.7% 1.9% 2.1% 2.4% 2.8% 3.7% 4.1% 4.7% 5.7%
3,000 50,50
2.9% 3.0% 3.1% 3.4% 3.8% 4.1% 4.7% 6.3% 6.9% 7.9% 9.5%
40,60
2.8% 2.9% 3.1% 3.3% 3.7% 4.0% 4.7% 6.2% 6.8% 7.7% 9.3%
30,70
2.7% 2.8% 2.9% 3.1% 3.4% 3.8% 4.4% 5.8% 6.3% 7.2% 8.7%
20,80
2.3% 2.4% 2.5% 2.7% 3.0% 3.3% 3.8% 5.0% 5.5% 6.3% 7.6%
10,90
1.7% 1.8% 1.9% 2.0% 2.3% 2.5% 2.8% 3.8% 4.2% 4.7% 5.7%
2,000 50,50
3.4% 3.4% 3.6% 3.8% 4.1% 4.4% 5.0% 6.5% 7.1% 8.0% 9.6%
40,60
3.3% 3.4% 3.5% 3.7% 4.0% 4.4% 4.9% 6.4% 7.0% 7.9% 9.4%
30,70
3.1% 3.2% 3.3% 3.4% 3.8% 4.1% 4.6% 6.0% 6.5% 7.4% 8.8%
20,80
2.7% 2.8% 2.8% 3.0% 3.3% 3.6% 4.0% 5.2% 5.7% 6.4% 7.7%
10,90
2.0% 2.1% 2.1% 2.3% 2.5% 2.7% 3.0% 3.9% 4.3% 4.8% 5.8%

B-20

100
13.1%
12.8%
12.0%
10.5%
7.9%
13.1%
12.9%
12.0%
10.5%
7.9%
13.2%
12.9%
12.1%
10.5%
7.9%
13.2%
12.9%
12.1%
10.6%
7.9%
13.3%
13.1%
12.2%
10.7%
8.0%

50
18.5%
18.1%
16.9%
14.8%
11.1%
18.5%
18.1%
16.9%
14.8%
11.1%
18.5%
18.1%
17.0%
14.8%
11.1%
18.5%
18.2%
17.0%
14.8%
11.1%
18.6%
18.2%
17.1%
14.9%
11.2%

Table B.8. Estimated 95% Error Margins for the Difference Between Two Subgroups (Continued)
DEFF
n1
P n2 = 6000 5000 4000 3000 2000 1500 1000 500
400
300
1.76
1,500 50,50
3.8% 3.8% 3.9% 4.1% 4.4% 4.7% 5.3% 6.7% 7.3% 8.2%
40,60
3.7% 3.8% 3.9% 4.0% 4.4% 4.7% 5.2% 6.6% 7.2% 8.1%
30,70
3.4% 3.5% 3.6% 3.8% 4.1% 4.4% 4.9% 6.2% 6.7% 7.5%
20,80
3.0% 3.1% 3.1% 3.3% 3.6% 3.8% 4.2% 5.4% 5.9% 6.6%
10,90
2.3% 2.3% 2.4% 2.5% 2.7% 2.8% 3.2% 4.0% 4.4% 4.9%
1,000 50,50
4.4% 4.5% 4.6% 4.7% 5.0% 5.3% 5.8% 7.1% 7.7% 8.6%
40,60
4.4% 4.4% 4.5% 4.7% 4.9% 5.2% 5.7% 7.0% 7.5% 8.4%
30,70
4.1% 4.1% 4.2% 4.4% 4.6% 4.9% 5.3% 6.5% 7.0% 7.8%
20,80
3.6% 3.6% 3.7% 3.8% 4.0% 4.2% 4.7% 5.7% 6.2% 6.8%
10,90
2.7% 2.7% 2.8% 2.8% 3.0% 3.2% 3.5% 4.3% 4.6% 5.1%
500 50,50
6.1% 6.1% 6.2% 6.3% 6.5% 6.7% 7.1% 8.2% 8.7% 9.5%
40,60
5.9% 6.0% 6.0% 6.2% 6.4% 6.6% 7.0% 8.1% 8.5% 9.3%
30,70
5.5% 5.6% 5.7% 5.8% 6.0% 6.2% 6.5% 7.5% 8.0% 8.7%
20,80
4.8% 4.9% 4.9% 5.0% 5.2% 5.4% 5.7% 6.6% 7.0% 7.6%
10,90
3.6% 3.7% 3.7% 3.8% 3.9% 4.0% 4.3% 4.9% 5.2% 5.7%
400 50,50
6.7% 6.8% 6.8% 6.9% 7.1% 7.3% 7.7% 8.7% 9.2% 9.9%
40,60
6.6% 6.6% 6.7% 6.8% 7.0% 7.2% 7.5% 8.5% 9.0% 9.7%
30,70
6.2% 6.2% 6.2% 6.3% 6.5% 6.7% 7.0% 8.0% 8.4% 9.1%
20,80
5.4% 5.4% 5.5% 5.5% 5.7% 5.9% 6.2% 7.0% 7.4% 7.9%
10,90
4.0% 4.1% 4.1% 4.2% 4.3% 4.4% 4.6% 5.2% 5.5% 6.0%
300 50,50
7.7% 7.7% 7.8% 7.9% 8.0% 8.2% 8.6% 9.5% 9.9% 10.6%
40,60
7.5% 7.6% 7.6% 7.7% 7.9% 8.1% 8.4% 9.3% 9.7% 10.4%
30,70
7.0% 7.1% 7.1% 7.2% 7.4% 7.5% 7.8% 8.7% 9.1% 9.7%
20,80
6.2% 6.2% 6.2% 6.3% 6.4% 6.6% 6.8% 7.6% 7.9% 8.5%
10,90
4.6% 4.6% 4.7% 4.7% 4.8% 4.9% 5.1% 5.7% 6.0% 6.4%

B-21

200
9.8%
9.6%
9.0%
7.8%
5.9%
10.1%
9.9%
9.2%
8.1%
6.0%
10.9%
10.7%
10.0%
8.7%
6.5%
11.3%
11.0%
10.3%
9.0%
6.8%
11.9%
11.6%
10.9%
9.5%
7.1%

100
13.4%
13.2%
12.3%
10.7%
8.1%
13.6%
13.4%
12.5%
10.9%
8.2%
14.2%
14.0%
13.1%
11.4%
8.5%
14.5%
14.2%
13.3%
11.6%
8.7%
15.0%
14.7%
13.8%
12.0%
9.0%

50
18.7%
18.3%
17.1%
15.0%
11.2%
18.8%
18.5%
17.3%
15.1%
11.3%
19.3%
18.9%
17.7%
15.4%
11.6%
19.5%
19.1%
17.9%
15.6%
11.7%
19.9%
19.5%
18.2%
15.9%
11.9%

Table B.8. Estimated 95% Error Margins for the Difference Between Two Subgroups (Continued)
DEFF
n1
P n2 = 6000 5000 4000 3000 2000 1500 1000 500
400
300
1.76
200 50,50
9.3% 9.4% 9.4% 9.5% 9.6% 9.8% 10.1% 10.9% 11.3% 11.9%
40,60
9.2% 9.2% 9.2% 9.3% 9.4% 9.6% 9.9% 10.7% 11.0% 11.6%
30,70
8.6% 8.6% 8.6% 8.7% 8.8% 9.0% 9.2% 10.0% 10.3% 10.9%
20,80
7.5% 7.5% 7.5% 7.6% 7.7% 7.8% 8.1% 8.7% 9.0% 9.5%
10,90
5.6% 5.6% 5.7% 5.7% 5.8% 5.9% 6.0% 6.5% 6.8% 7.1%
100 50,50
13.1% 13.1% 13.2% 13.2% 13.3% 13.4% 13.6% 14.2% 14.5% 15.0%
40,60
12.8% 12.9% 12.9% 12.9% 13.1% 13.2% 13.4% 14.0% 14.2% 14.7%
30,70
12.0% 12.0% 12.1% 12.1% 12.2% 12.3% 12.5% 13.1% 13.3% 13.8%
20,80
10.5% 10.5% 10.5% 10.6% 10.7% 10.7% 10.9% 11.4% 11.6% 12.0%
10,90
7.9% 7.9% 7.9% 7.9% 8.0% 8.1% 8.2% 8.5% 8.7% 9.0%
50
50,50
18.5% 18.5% 18.5% 18.5% 18.6% 18.7% 18.8% 19.3% 19.5% 19.9%
40,60
18.1% 18.1% 18.1% 18.2% 18.2% 18.3% 18.5% 18.9% 19.1% 19.5%
30,70
16.9% 16.9% 17.0% 17.0% 17.1% 17.1% 17.3% 17.7% 17.9% 18.2%
20,80
14.8% 14.8% 14.8% 14.8% 14.9% 15.0% 15.1% 15.4% 15.6% 15.9%
10,90
11.1% 11.1% 11.1% 11.1% 11.2% 11.2% 11.3% 11.6% 11.7% 11.9%

B-22

200
13.0%
12.7%
11.9%
10.4%
7.8%
15.9%
15.6%
14.6%
12.7%
9.6%
20.6%
20.1%
18.8%
16.4%
12.3%

100
15.9%
15.6%
14.6%
12.7%
9.6%
18.4%
18.0%
16.9%
14.7%
11.0%
22.5%
22.1%
20.6%
18.0%
13.5%

50
20.6%
20.1%
18.8%
16.4%
12.3%
22.5%
22.1%
20.6%
18.0%
13.5%
26.0%
25.5%
23.8%
20.8%
15.6%

Weighting Methodology
Base sampling Weights
For the cross-sectional landline sample, the base sampling weight equals the population count of
land line telephone numbers in the list-assisted sampling frame, divided by the total count of
sample telephone numbers for the replicates that were released:
•

If STATUS = 1 and FPROJ = 4548, then BSW = 4414.828.

For the cell phone sample, the base sampling weight equals the population count of telephone
numbers in the cellular sampling frame, divided by the total count of sample telephone numbers
for the replicates that were released:
•

If STATUS = 1 and FPROJ = 4548c, then BSW = 21779.137.

A separate landline sample was used to screen for households containing one or more persons 16
to 34 years old. This oversample leads to an overrepresentation of persons 16 to 34 years old in
the combined landline sample. Therefore, for the cross-sectional landline interviews, we
calculated the sum of the base sampling weights for respondents 16 to 34 years old. Call this
SUM16-34. We then obtained the unweighted count of respondents with (STATUS = 1 and
FPROJ = 4548 and (D1) = 16 to 34) or (STAUS = 1 and FPROJ = 4548o and D1 = 16 to 34).
Call this count N16-34. For the respondents 16 to 34 years old, the base sampling weight equals
SUM16-34 / N16-34.
The base sampling weights were assigned to the 6,144 completed interviews
Design Weights
For the cross-sectional landline sample one person 16 or older was randomly selected from the
household. The base sampling weight was multiplied by the number of age-eligible people in the
household, with the maximum value capped at five:
•

If FPROJ = 4548, BSW_NUMADULT = BSW x SL1o_R.

•

Recodes: SL1o_R = SL1o values of 1 to 5. If SL1o = missing, SL1o_R = 2. If SL1o = 6
to 10, SL1o_R = 5.

For the cell phone sample, the cell phone was treated as a personal device and no respondent
selection took place:
•

If FPROJ = 4548c, BSW_NUMADULT = BSW.

For the oversample of households containing one or more persons 16 to 34 years old, one person
in this age range was randomly selected. The base sampling weight was multiplied by the
number of age-eligible persons in the household, with the maximum value capped at five:

B-23

•

If FPROJ = 4548o, BSW_NUMADULT = BSW x SO1o_R.

•

Recodes: SO1o_R = SL1 values of 1 to 5. If SO1o = missing, SO1o_R = 2. If SO1o = 6
to 10, SO1o_R = 5.

If the cell phone respondent reported that they had two or more personal-use cell phones, the
weight from the prior step was divided by two:
•
•

If FPROJ = 4548c and SC4 = 1, then BSW_NUMPHONE = BSW_NUMADULT / 2.
Otherwise, BSW_NUMPHONE = BSW_NUMADULT.

If the landline respondent reported that they had two or more voice-use landline telephone
numbers in the household, the weight from the prior step were divided by two:
•
•

If FPROJ = 4548 or 4548oandD11 = 2, then BSW_NUMPHONE = BSW_NUMADULT
/ 2.
Otherwise, BSW_NUMPHONE = BSW_NUMADULT.

Compositing Weights
The 6,144 completed interviews were first divided into four telephone status categories:
•
•
•

FPROJ = 4548c.
IF SC5 = 1, TELEPHONE_STATUS = 1 (cell only).
IF SC5 ≠ 1, TELEPHONE_STATUS = 3 (cell sample, dual user).

•
•
•

FPROJ = 4548 or 4548o.
IF D12 = 2, TELEPHONE_STATUS = 4 (landline sample, dual user).
IF D12 ≠ 2, TELEPHONE_STATUS = 2 (landline only).

From telephone status, we created a second telephone status variable:
•

If TELEPHONE_STATUS = 1, TELEPHONE_STATUS2 = 1 (cell only).

•

If TELEPHONE_STATUS = 2, TELEPHONE_STATUS2 = 2 (landline only).

•

If TELEPHONE_STATUS = 3 and D13 = 1, TELEPHONE_STATUS2 = 3 (cell sample,
dual users, cell mostly).

•

If TELEPHONE_STATUS = 4 and D13 = 1, TELEPHONE_STATUS2 = 4 (landline
sample, dual users, cell mostly).

•

If TELEPHONE_STATUS = 4 and D13 ≠ 1, TELEPHONE_STATUS2 = 6 (landline
sample dual users, not cell mostly).

B-24

For each of telephonestatus2 = 3 and 4, we calculated the mean and standard deviation of
BSW_NUMPHONE. For each of telephonestatus2 = 3 and 4, we then calculated:
•

1+ [SD3/Mean3]2 = DEFF3 and 1+ [SD4/Mean4]2 = DEFF4.

Next, we divided the unweighted number of interviews in each telephonestatus2 = 3by Deff3 to
obtain neffective3, and the unweighted number of interviews in each telephonestatus2 = 4 by
Deff4 to obtain neffective4.
The dual frame compositing factors for dual users who are cell mostly equal:
Lambda3 = neffective3/(n effective3 + neffective4).
Lambda4 = neffective4/(neffective3 + neffective4).
•
•
•
•
•

If telephonestatus2 = 1, BSW_COMPOSITED = BSW_NUMPHONE.
If telephone_status2 = 2, BSW_COMPOSITED = BSW_NUMPHONE.
If telephone_status2 = 3, BSW_COMPOSITED = BSW_NUMPHONE * Lambda3.
If telephone_status2 = 4, BSW_COMPOSITED = BSW_NUMPHONE * Lambda4.
If telephone_status2 = 6, BSW_COMPOSITED = BSW_NUMPHONE.

Raking to Population Control Totals
A survey sample may cover segments of the target population in proportions that do not match
the proportions of those segments in the population itself. The differences may arise, for
example, from sampling fluctuations, from nonresponse, or because the sample design was not
able to cover the entire target population. In such situations, one can often improve the relation
between the sample and the population by adjusting the sampling weights of the cases in the
sample, so that the marginal totals of the adjusted weights on specified characteristics, referred to
as control variables, agree with the corresponding totals for the population. This operation is
known as raking ratio estimation, raking, or sample-balancing, and the population totals are
usually referred to as control totals. Raking is most often used to reduce biases from nonresponse
and noncoverage in sample surveys. The term “raking” suggests an analogy with the process of
smoothing the soil in a garden plot, by alternately working it back and forth with a rake in two
perpendicular directions.
Raking usually proceeds, one variable at a time, applying a proportional adjustment to the
weights of the cases that belong to the same category of the control variable. The initial design
weights in the raking process are often equal to the inverse of the selection probabilities and may
have undergone some adjustments for unit nonresponse and noncoverage. The weights from the
raking process are used in estimation and analysis.
The adjustment to control totals is sometimes achieved by creating a cross-classification of the
categorical control variables (e.g., age categories × gender × race × household-income
categories) and then matching the total of the weights in each cell to the control total. This
approach, however, can spread the sample thinly over a large number of adjustment cells. It also
requires control totals for all cells of the cross-classification. Often, this is not feasible (e.g.,
control totals may be available for age × gender × race, but not when those cells are subdivided
B-25

by household income). The use of raking with marginal control totals for single variables (i.e.,
each margin involves only one control variable) often avoids many of these difficulties.
The procedure known as raking adjusts a set of data so that its marginal totals match control
totals on a specified set of variables. In a simple 2-variable example, the marginal totals in
various categories for the two control variables are known from the entire population, but the
joint distribution of the two variables is known only from a sample. In the cross-classification of
the sample, arranged in rows and columns, one might begin with the rows, taking each row in
turn, and multiplying each entry in the row by the ratio of the population total to the weighted
sample total for that category, so that the row totals of the adjusted data agree with the
population totals for that variable. The weighted column totals of the adjusted data, however,
may not yet agree with the population totals for the column variable. Thus, the next step, taking
each column in turn, multiplies each entry in the column by the ratio of the population total to
the current total for that category. Now the weighted column totals of the adjusted data agree
with the population totals for that variable, but the new weighted row totals may no longer match
the corresponding population totals.
This process continues, alternating between the rows and the columns, and close agreement on
both rows and columns is usually achieved after a small number of iterations. The result is a
tabulation for the population that reflects the relation of the two control variables in the sample.
Raking can also adjust a set of data to control totals on three or more variables. In such
situations, the control totals often involve single variables, but they may involve two or more
variables.
Ideally, one should rake on variables that exhibit an association with the key survey outcome
variables and that are related to nonresponse and/or noncoverage. This strategy will reduce bias
in the key outcome variables. In practice, other considerations may enter. A variable such as
gender may not be strongly related to key outcome variables or to nonresponse, but raking on it
may be desirable to preserve the “face validity” of the sample.
For this survey, nine raking control variables were used:
1.
2.
3.
4.
5.
6.
7.
8.
9.

Telephone status
Census Region
Number of children under 16 in the household
Number of persons in the household
Marital status
Education
Tenure status (Rent or Own Home)
Race/ethnicity
Age group by gender

The population control totals were obtained from the 2009 American Community Survey Public
Use Microdata Sample (PUMS), except for telephone status which was obtained from the
National Health Interview Survey. The population control totals are for people living in
households 16 and older. Population control totals do not exist for drivers. The survey, therefore,

B-26

included non-drivers in the landline sample, but for cost reasons, they were not interviewed in
the cell phone sample.
An SAS raking macro (Izrael et al., 2009) was used to develop the raked weights for the 6,144
completed interviews. BSW_COMPOSITED was used as the input weight for the raking. During
the raking process, a weight trimming procedure was implemented. A reduction in the variability
of the weights, as measured by the coefficient of variation of the weights, can be achieved by
reducing a few large weight values and increasing a few low weight values. A weight-trimming
procedure (Izrael et al., 2009) was, therefore, implemented during the raking iterative process, in
order to ensure that: (1) a limit was placed on high and low weight values in the final weights;
(2) the convergence criteria were satisfied and (3) the weights summed to the correct population
total (243,680,923). The raking output is presented in Appendix C. The raked weight is
FINAL_WEIGHT. The interviews with drivers represent a domain of the population and,
therefore, the estimated population of drivers is referred to as a domain estimate.

B-27

Non-response Bias Analysis
Comparison of Characteristics of Completed Landline Sample Interview Telephone
Numbers With Nonrespondent Landline Sample Telephone Numbers
Unit Nonresponse in a probability sample encompasses sampling units that do not complete the
survey. For a random-digit-dialing sample, not all telephone numbers in the sample are
residential numbers and, among the residential sample numbers, not all will yield a completed
interview. For list samples, the sampling frame may contain considerable information on all
population elements allowing for the comparison of the characteristics of respondents and
nonrespondents. For RDD samples, the sampling frame contains very little information on the
characteristics of the residential telephone numbers in the sample. We can determine the
residential directory-listed status of each sample telephone number in the landline sample. We
can also assign each landline telephone number to a NHTSA region and to Nielsen county size
categories. 4 Socio-demographic characteristics of ZIP Codes can be mapped into landline
telephone exchanges to create exchange-level socio-demographic characteristics. These are not
socio-demographic characteristics of telephone numbers; they are ecological variables.
Exchange-level variables were obtained from Survey Sampling Inc. for the percent of the
population in the exchange that is African-American, Hispanic, and Asian. Mean household
income for the telephone exchange was also obtained from SSI. These variables do not exist for
the cellular sample. The nonresponse analysis presented below is therefore limited to the crosssectional landline telephone sample.
The cross-sectional landline telephone numbers were divided into five categories: (1) completed
interviews, (2) known residential numbers that did not yield a completed interview, (3) likely
residential telephone numbers, (4) undetermined residential numbers (i.e., residential status not
determined), and (5) nonresidential numbers. We removed the nonresidential numbers from the
analysis. Appendix A provides a mapping of the final disposition codes into the first three
categories.
We created three dichotomous dependent variables: (1) completed interviews versus known
residential numbers that did not yield a completed interview (n = 5,907), (2) completed
interviews versus known residential numbers that did not yield a completed interview and likely
residential numbers (n = 12,953), and (3) completed interviews versus known residential
numbers that did not yield a completed interview, likely residential numbers and undetermined
residential numbers (n = 19,509). We then fit three unweighted logistic regression models using
4

Nielsen Code
A
B
C
D

Description
All counties belonging to the largest metropolitan areas
which account for 40% of all U.S. households
All counties in the next largest set of metropolitan
areas that account for 30% of all U.S. households
All counties in the next largest set of metropolitan
areas that account for 15% of all U.S. households
All remaining counties

B-28

the variables discussed above as the predictors in the model. The first model compares
respondents and known nonrespondents. The second model expands the nonrespondent
definition to include likely residential telephone numbers. The third model takes the broadest
view of nonresponse in that it includes the undetermined residential numbers as nonrespondents.
Table B.9 contains a column for each dependent variable. The rows list the categorical and
continuous predictors in the model. For each categorical predictor we indicate the reference
category. The cell entries in the table give the statistical significance of each predictor. For the
comparison of completed interviews with known residential numbers that are nonrespondents,
the logistic regression coefficients for Nielsen county size category B, C and D are statistically
significant at the 0.05 level. The logistic regression coefficients for NHTSA Regions 6 and 10
are significant at the 0.05 level. For the continuous exchange-level predictor variables we find
significant coefficients for the percent African-American population, the percent Hispanic
population, and mean household income.
Table B.10 shows the adjusted odds ratios from the logistic regression models. For the
comparison of completed interviews with known residential numbers that are nonrespondents we
find that for the statistically significant predictors:
•
•
•
•
•

known residential numbers from Nielsen county category B, C and D are more likely to
be respondents than known residential numbers from category A.
Known residential numbers from NHTSA Regions 6 and 10 are more likely to be
respondents than known residential numbers from Region 1. Judging by the adjusted
odds ratios this is the strongest effect.
The lower the percent African-American population in the telephone exchange the more
likely a known residential number will yield a response.
The lower the percent Hispanic population in the telephone exchange the more likely a
known residential number will yield a response.
Although the mean household income variable is statistically significant, the relationship
with nonresponse is negligible (i.e., odds ratio = 1.0).

B-29

Table B.9. Analysis of Maximum Likelihood Estimates: Statistical Significance
Predictor

DIRECTORY LISTED NUMBER (Reference
category = Not Directory listed)

Completed Interviews Compared With
Nonresponse Group:
(1)
(2)
(3)
Completed
Completed
Completed
interview
interview
interview
versus
versus
versus known
known
known
residential
landline
residential residential
telephone
landline
landline
telephone
telephone
numbers, not
numbers,
numbers,
completed,
not
not
likely
completed
completed,
residential
and likely
landline
residential numbers, and
landline
undetermined
numbers
residential
status
landline
numbers
0.2209
0.0225
<.0001
0.0058

<.0001

SIZE C (Reference

0.0485

<.0001

SIZE D (Reference

0.0028

<.0001

(Reference category = 1)

0.9979

0.4748

0.8805

NHTSA REGION 4 (Reference category = 1)

0.084

0.7647

0.4388

NIELSEN COUNTY
category = A)
NIELSEN COUNTY
category = A)
NIELSEN COUNTY
category = A)
NHTSA REGION 2

SIZE B (Reference

NHTSA REGION 3 (Reference category = 1)

0.2573

NHTSA REGION 5 (Reference category = 1)

0.816

NHTSA REGION 6 (Reference category = 1)

0.0184

NHTSA REGION 8 (Reference category = 1)

0.6318

NHTSA REGION 7 (Reference category = 1)

0.0712

NHTSA REGION 9 (Reference category = 1)

0.1288

NHTSA REGION 10 (Reference category =
1)
PERCENT BLACK POPULATION IN TELEPHONE
EXCHNAGE
PERCENT HISPANIC POPULTION IN TELEPHONE
EXCHNAGE
PERCENT ASIAN POPULATION IN TELEPHONE
EXCHNAGE
AVERAGE HOUSEHOLD INCOME IN TELEPHONE
EXCHANGE

0.0316

B-30

0.5644
0.4651

0.6839
0.546

0.6776

0.3078

0.7465

0.6491

0.2246
0.7801
0.1469

0.0656
0.6945
0.1005

<.0001

0.079

<.0001

0.0049

0.3861

0.3155

When we expand the definition of residential telephone numbers to include likely residential
numbers, we examine the results in Tables B.9 and B.10 for the dependent variable defined by
completed interviews versus known residential numbers that did not yield a completed interview
and likely residential numbers. The residential directory-listed status of the telephone number is
statistically significant, along with the three Nielsen county size categories. None of the NHTSA
Regions are significant, while all three race/ethnicity exchange variables are significant.
Examining the adjusted odds ratios we find that:
•
•
•
•
•

Known and likely residential numbers that are directory listed are more likely to be
respondents than numbers that are not directory listed.
Known and likely residential numbers from Nielsen county category B, C and D are more
likely to be respondents than numbers from category A. Judging by the adjusted odds
ratios this is the strongest effect.
The lower the percent African-American population in the telephone exchange the more
likely a known or likely residential number will yield a response.
The lower the percent Hispanic population in the telephone exchange the more likely a
known or likely residential number will yield a response.
The lower the percent Asian population in the telephone exchange the more likely a
known or likely residential number will yield a response.

A further expansion of the nonrespondent group to include undetermined residential numbers
yields essentially the same findings as for the expansion from known residential numbers to
likely residential numbers. The nonresponse bias analysis across all three dependent variables
finds consistent Nielsen county size effects, and black and Hispanic race/ethnicity exchange
effects. The weighting methodology for the NHTSA Speeding Survey included poststratification
by race/ethnicity of the respondent. Future surveys should give consideration to determining
county of residence in order to form an urban/rural continuum variable, consistent with Census
Bureau population data sources such as the American Community Survey, that can be used in
poststratification.

B-31

Table B.10: Adjusted Odds Ratio Estimates for Modeling Completed Interviews
Predictor

DIRECTORY LISTED NUMBER (Reference
category = Not Directory listed)
NIELSEN COUNTY SIZE B (Reference
category = A)
NIELSEN COUNTY SIZE C (Reference
category = A)
NIELSEN COUNTY SIZE D (Reference
category = A)
NHTSA_REGION 2 (Reference category =
1)
NHTSA_REGION 3 (Reference category =
1)
NHTSA_REGION 4 (Reference category =
1)
NHTSA_REGION 5 (Reference category =
1)
NHTSA_REGION 6 (Reference category =
1)
NHTSA_REGION 7 (Reference category =
1)
NHTSA_REGION 8 (Reference category =
1)
NHTSA_REGION 9 (Reference category =
1)
NHTSA_REGION 10 (Reference category =
1)
PERCENT BLACK POPULATION IN TELEPHONE
EXCHNAGE
PERCENT HISPANIC POPULTION IN
TELEPHONE EXCHNAGE
PERCENT ASIAN POPULATION IN TELEPHONE
EXCHNAGE
AVERAGE HOUSEHOLD INCOME IN TELEPHONE
EXCHANGE

1.253

1.212

1.228

1.395

1.292

1.410

1.522

1.291

1.000

0.938

0.988

1.196

0.946

0.965

1.307

0.972

0.935

1.034

1.067

1.050

1.502

1.044

0.908

1.407

1.144

1.206

1.105

0.960

1.054

1.288

0.972

0.964

1.553

1.185

1.192

0.986

0.992

0.985

0.988

0.986

0.989

0.984

0.980

1.000

B-32

Comparison of Early Versus Late Responders
One limitation of the analysis presented above is that it does not use any substantive variables
included in the survey because this information is not obtained for nonrespondents. For the
cross-sectional landline interviews we can however divide the completed interviews into two
groups: early versus late responders. The two groups are formed by examining the distribution of
the number of call attempts required to complete the interview. Examining the distribution we
find that 90% of the interviews were completed within the first 13 call attempts and 10% of the
interviews were completed at attempt 14 to 35. We will use the interviews completed at 14 to 35
call attempts as the late responder group. The concept behind this approach is that late
responders may be more similar to nonrespondents than the early responders.
We identified eight key substantive survey questions to include in the nonresponse bias analysis:
1. How often do you usually drive a car or other motor vehicle? Would you say that you usually
drive . . . (NOTE: Motorcycle counts as a motor vehicle)
1 Every day, or almost every day
2 Several days a week
3 Once a week or less
4 Only certain times a year, OR
5 Never
SKIP TO D1
6 (VOL) Don’t know
SKIP TO D1
7 (VOL) Refused
SKIP TO D1
3. Which of the following statements best describes your driving? READ AND ROTATE 1&2
1 I tend to pass other cars more often than other cars pass me OR
2 Other cars tend to pass me more often then I pass them
3 (VOL) Both/About equally
4 (VOL) Don’t know
5 (VOL) Refused
4. When driving I tend to . . . READ AND ROTATE 1&2
1 Stay with slower moving traffic, or
2 Keep up with the faster traffic
3 (VOL) Both/About Equally
4 (VOL) Don’t know
5 (VOL) Refused
5e. How often would you say you drive 15 mph over the speed limit on Multi-Lane, Divided
Highways?
1 Often
2 Sometimes
3 Rarely
4 Never
5 (VOL) Don’t know
6 (VOL) Refused

B-33

6e. How often would you say you drive 15 miles an hour over the speed limit on Two-Lane
Highways, one lane in each direction?
1 Often
2 Sometimes
3 Rarely
4 Never
5 (VOL) Don’t know
6 (VOL) Refused
7e. How often would you say you drive 10 miles an hour over the speed limit on Neighborhood
or Residential streets?
1 Often
2 Sometimes
3 Rarely
4 Never
5 (VOL) Don’t know
6 (VOL) Refused
30. In the past TWELVE MONTHS have you been STOPPED for speeding by the police?
1 Yes
2 No
SKIP TO Q34
3 (VOL) Don’t know SKIP TO Q34
4 (VOL) Refused
SKIP TO Q34
31. How many times have you been stopped for speeding in the past twelve months?
_______ TIMES STOPPED
Range = 0 to 7
8 (VOL) Don’t know
9 (VOL) Refused
We produced an unweighted tabulation of the early versus late responder variable by each of the
eight substantive survey variables. There is a statistically significant difference between early
and late responders only for questions 1 and 4. 5 For question 1, 86% of late responders drive
every day while 81% of early responders drive every day. For question 4, 47% of late responders
reported that they keep up with faster traffic while among early responders 42% keep up with
faster traffic.
Our analysis of nonresponse has focused on the cross-sectional landline sample. The analysis of
early versus late responders was also implemented for the cell phone sample. We conducted a
similar analysis for the cell phone sample completed interviews but for the cell phone sample we
defined late responders as those interviews completed at attempt 9 to 16. We find that none of
the eight variables are statistically significant. Our variable-specific analysis of nonresponse bias
5

For a two-variable contingency table the null hypothesis for the Chi Square test is that the two variables are
independent. If the Chi Square statistic is significant at the 0.05 level we reject the null hypothesis that early versus
late responders is independent of the substantive survey question.

B-34

finds evidence of modest bias only for frequency of driving and driving pattern while in traffic,
but only for the landline sample.

References
Blumberg, S. J., & Luke, J. V. (2012, June). Wireless substitution: Early release of estimates
from the National Health Interview Survey, July–December 2011. Atlanta: National Center for
Health Statistics. June 2012. Available at www.cdc.gov/nchs/nhis.htm.
Cochran, W. G. (1977). Sampling techniques (3rd ed.). New York: John Wiley & Sons.
Izrael, D., Battaglia, M., & Frankel, M. 2009. Extreme Survey Weight Adjustment as a
Component of Sample Balancing (a.k.a. Raking), 2009 SAS Global
Forum.www.abtassociates.com/Page.cfm?PageID=40858&FamilyID=8600.

B-35

APPENDIX C
Output for Raking with Trimming Weight by
Individual and Global Cap Value Method

Raking with Trimming Weight by Individual and Global Cap Value Method
Sample size of completed interviews: 6144
Raking input weight adjusted to population total: BSW_COMPOSITED_ATPT
Mean value of raking input weight adjusted to population total: 39661.61
Minimum value of raking input weight: 3063.71
Maximum value of raking input weight: 97303.55
Coefficient of variation of raking input weight: 0.64
Global low weight cap value (GLCV): 4957.70
Global low weight cap value factor: Mean input weight times 0.125
Global high weight cap value (GHCV): 317292.87
Global high weight cap value factor: Mean input weight times 8.0
Individual low weight cap value (ILCV) factor: Respondent's weight times 0.25
Individual high weight cap value (IHCV) factor: Respondent's weight times 4
Number of respondents who have an individual high weight cap value less than the global low weight cap value
(GLCV used in weight trimming): 0
Number of respondents who have an individual low weight cap value greater than the global high weight cap value
(GHCV used in weight trimming): 0

Weighted Distribution Prior to Raking. Iteration 0
Input Weight
Sum of
Weights

Target
Total

1 cell only

71714362.47

69205382

2508980.34

29.430

28.400

2 landline only

31024855.43

26561221

4463634.82

12.732

10.900

1.832

3 cell/landline sample dual users cell mostly

29877314.96

43375204 -13497889.3

12.261

17.800

-5.539

6 landline sample dual users not cell mostly

111064390.1 104539116

45.578

42.900

2.678

Census Region
1 Northeast
2 Midwest
3 South
4 West

Input Weight
Sum of
Weights
46899337.74
62673746.57
86662498.14
47445340.55

Imputed value I_D7B_R4 : Number of persons
in HH under 16 years

Input Weight
Sum of
Weights

1 0 Children under 16 in HH

166176782.0 158543587

2 1 Child under 16 in HH
3 2 Children under 16 in HH
4 3+ Children under 16 in HH

TELEPHONE_STATUS2_R

Target
Total
44625674
53188804
89418582
56447862

Sum of
% of
Weights
Input Target % of Difference
Difference Weights
Weights
in %

6525274.18

Sum of
Weights
Difference
2273663.32
9484942.66
-2756084.30
-9002521.67

Target
Total

% of
Input
Weights
19.246
25.720
35.564
19.470

Target % of
Weights
18.313
21.827
36.695
23.165

1.030

Difference
in %
0.933
3.892
-1.131
-3.694

Sum of
% of
Weights
Input Target % of Difference
Difference Weights
Weights
in %
7633194.82

68.194

65.062

3.132

36192967.91

39249100 -3056132.48

14.853

16.107

-1.254

25835431.51

29331619 -3496187.51

10.602

12.037

-1.435

15475741.63

16556616 -1080874.82

6.351

6.794

-0.444

C-2

Imputed value I_D7A_R5 : Number of persons
in HH

Input Weight
Sum of
Weights

1 Person in HH

32563801.31 32697180

-133379.05

13.363

13.418

-0.055

2 Persons in HH

87285446.38 79192398

8093048.11

35.820

32.498

3.321

3 Persons in HH

45776939.31 47772503 -1995563.36

18.786

19.605

-0.819

4 Persons in HH

43333609.26 43201945

131664.39

17.783

17.729

0.054

5+ Persons in HH

34721126.74 40816897 -6095770.10

14.249

16.750

-2.502

Target
Total

Sum of
% of
Weights
Input Target % of Difference
Difference Weights
Weights
in %

Imputed value I_D4_R4 : Marital Status

Input Weight
Sum of
Weights

1 Married

146512959.8 125507011 21005948.71

60.125

51.505

8.620

2 Divorced/Separated

25848963.71

31851392 -6002428.20

10.608

13.071

-2.463

3 Widowed

14452161.05

14535520

-83358.62

5.931

5.965

-0.034

4 Never married

56866838.44

71787000 -14920161.9

23.337

29.459

-6.123

Imputed value I_D3_R4 : Education

Input Weight
Sum of
Weights

1 Less than HS

19324485.73 42604849 -23280363.0

7.930

17.484

-9.554

2 HS/GED

68581526.23 67538321

1043205.11

28.144

27.716

0.428

3 Some college

66833247.75 72138583 -5305335.01

27.427

29.604

-2.177

4 College graduate

88941663.29 61399170 27542492.88

36.499

25.197

11.303

Target
Total

Imputed value I_D8_R2 : Tenure

Input Weight
Sum of
Weights

1 Own

176238592.3 169105527

2 Rent

67442330.68

C-3

Target
Total

Sum of
% of
Weights
Input Target % of Difference
Difference Weights
Weights
in %

Sum of
% of
Weights
Input Target % of Difference
Difference Weights
Weights
in %
7133065.43

72.324

69.396

2.927

74575396 -7133065.43

27.676

30.604

-2.927

Imputed value I_RACEETHNICITY_R7 : Race
Ethnicity

Input Weight
Sum of
Weights

1 Hispanic

20498693.19

Target
Total

Sum of
% of
Weights
Input Target % of Difference
Difference Weights
Weights
in %

33864671 -13365977.3

13.897

-5.485

2 AIAN NonHispanic

3433270.00

1941363.24

1.409

0.612

0.797

3 Asian/NHOPI NonHispanic

6234290.21

11445418 -5211127.46

2.558

4.697

-2.139

4 Black NonHispanic

20315910.48

27738307 -7422396.80

8.337

11.383

-3.046

6 White NonHispanic

185705689.8 165570080 20135609.95

76.209

67.945

8.263

3.075

1.465

1.610

7 Other NonHispanic

1491907

8.412

7493069.32

3570541

Imputed value I_D1_R7_SA3 : Agegroup by
Sex

Input Weight
Sum of
Weights

Target
Total

11 16-24, Male

14034891.56 19083440 -5048548.79

5.760

7.831

-2.072

12 16-24, Female

11830961.38 18285697 -6454735.14

4.855

7.504

-2.649

8719871.52 11001548 -2281676.82

3.578

4.515

-0.936

21 25-29, Male
22 25-29, Female

3922528.40

Sum of
% of
Weights
Input Target % of Difference
Difference Weights
Weights
in %

8682039.43 10883200 -2201160.36

3.563

4.466

-0.903

31 30-39, Male

15477585.68 20568127 -5090540.87

6.352

8.441

-2.089

32 30-39, Female

18221883.64 20774954 -2553070.81

7.478

8.525

-1.048

41 40-49, Male

20413972.52 22285092 -1871119.59

8.377

9.145

-0.768

42 40-49, Female

25057499.91 23065503

1991996.88

10.283

9.465

0.817

51 50-64, Male

34293426.62 28119033

6174394.07

14.073

11.539

2.534

52 50-64, Female

42332765.51 30228822 12103943.32

17.372

12.405

4.967

61 65-74, Male

13016619.24

9866415

3150204.06

5.342

4.049

1.293

62 65-74, Female

15678959.48 11543400

4135559.24

6.434

4.737

1.697

71 75 PLUS, Male

7281649.26

186024.82

2.988

2.912

0.076

72 75 PLUS, Female

8638797.26 10880067 -2241270.03

7095624

3.545

4.465

-0.920

**** Program terminated at iteration 7 because all current percentages differ from target percentages by less than 0.10
****

C-4

Weighted Distribution After Raking

TELEPHONE_STATUS2_R

Output
Weight Sum
of Weights

Sum of
% of
Target Weights Output Target % of Difference
Total Difference Weights
Weights
in %

1 cell only

69201177.11

69205382

-4205.02

28.398

28.400

-0.002

2 landline only

26556126.67

26561221

-5093.93

10.898

10.900

-0.002

3 cell/landline sample dual users cell mostly

43375691.13

43375204

486.83

17.800

0.000

6 landline sample dual users not cell mostly

104547928.1 104539116

8812.12

42.904

42.900

0.004

Sum of
Weights
Difference
-30686.56
30554.75
20848.97
-20717.16

% of
Output
Weights
18.301
21.840
36.704
23.156

Target % of
Weights
18.313
21.827
36.695
23.165

Difference
in %
-0.013
0.013
0.009
-0.009

Census Region
1 Northeast
2 Midwest
3 South
4 West

Output
Weight Sum
of Weights
44594987.86
53219358.66
89439431.42
56427145.07

Imputed value I_D7B_R4 : Number of persons
in HH under 16 years

Output
Weight Sum
of Weights

1 0 Children under 16 in HH

158695796.4 158543587 152209.26

65.124

65.062

0.062

2 1 Child under 16 in HH

39212123.27

39249100 -36977.12

16.092

16.107

-0.015

3 2 Children under 16 in HH

29268727.91

29331619 -62891.11

12.011

12.037

-0.026

4 3+ Children under 16 in HH

16504275.43

16556616 -52341.02

6.773

6.794

-0.021

Target
Total
44625674
53188804
89418582
56447862

Sum of
% of
Target
Weights Output Target % of Difference
Total Difference Weights
Weights
in %

Imputed value I_D7A_R5 : Number of persons
in HH

Output
Weight Sum
of Weights

1 Person in HH

32650819.85 32697180 -46360.51

13.399

13.418

-0.019

2 Persons in HH

79154296.61 79192398 -38101.66

32.483

32.498

-0.016

3 Persons in HH

47790478.91 47772503

17976.24

19.612

19.605

0.007

4 Persons in HH

43238921.24 43201945

36976.37

17.744

17.729

0.015

5+ Persons in HH

40846406.39 40816897

29509.55

16.762

16.750

0.012

C-5

Sum of
% of
Target
Weights Output Target % of Difference
Total Difference Weights
Weights
in %

Imputed value I_D4_R4 : Marital Status

Output
Weight Sum
of Weights

Sum of
% of
Target
Weights Output Target % of Difference
Total Difference Weights
Weights
in %

1 Married

125440317.0 125507011 -66694.06

51.477

51.505

-0.027

2 Divorced/Separated

31801674.75

31851392 -49717.16

13.051

13.071

-0.020

3 Widowed

14484089.30

14535520 -51430.37

5.944

5.965

-0.021

4 Never married

71954841.93

71787000 167841.59

29.528

29.459

0.069

Imputed value I_D3_R4 : Education

Output
Weight Sum
of Weights

Sum of
% of
Target
Weights Output Target % of Difference
Total Difference Weights
Weights
in %

1 Less than HS

42624175.95 42604849

19327.24

17.492

17.484

0.008

2 HS/GED

67533345.22 67538321

-4975.90

27.714

27.716

-0.002

3 Some college

72140634.36 72138583

2051.60

29.605

29.604

0.001

4 College graduate

61382767.47 61399170 -16402.94

25.190

25.197

-0.007

Imputed value I_D8_R2 : Tenure

Output
Weight Sum
of Weights

1 Own

168977785.2 169105527 -127741.74

69.344

69.396

-0.052

2 Rent

74703137.85

30.656

30.604

0.052

Imputed value I_RACEETHNICITY_R7 : Race
Ethnicity

Output
Weight Sum
of Weights

1 Hispanic

33920284.58

2 AIAN NonHispanic

Target
Total
74575396

Sum of
% of
Weights Output Target % of Difference
Difference Weights
Weights
in %
127741.74

Sum of
% of
Target
Weights Output Target % of Difference
Total Difference Weights
Weights
in %
33864671

55614.06

13.920

13.897

0.023

1490477.28

1491907

-1429.48

0.612

-0.001

3 Asian/NHOPI NonHispanic

11456722.17

11445418

11304.50

4.702

4.697

0.005

4 Black NonHispanic

27748187.81

27738307

9880.53

11.387

11.383

0.004

6 White NonHispanic

165492646.3 165570080 -77433.52

67.914

67.945

-0.032

1.466

1.465

0.001

7 Other NonHispanic

3572604.83

C-6

3570541

2063.90

Imputed value I_D1_R7_SA3 : Agegroup by
Sex

Output
Weight Sum
of Weights

11 16-24, Male

19083440.34 19083440

0.00

7.831

0.000

12 16-24, Female

18285696.52 18285697

0.00

7.504

-0.000

21 25-29, Male

11001548.34 11001548

0.00

4.515

-0.000

22 25-29, Female

10883199.79 10883200

0.00

4.466

0.000

31 30-39, Male

20568126.55 20568127

0.00

8.441

-0.000

32 30-39, Female

20774954.44 20774954

-0.00

8.525

-0.000

41 40-49, Male

22285092.11 22285092

-0.00

9.145

-0.000

42 40-49, Female

23065503.02 23065503

0.00

9.465

0.000

51 50-64, Male

28119032.55 28119033

-0.00

11.539

-0.000

52 50-64, Female

30228822.18 30228822

61 65-74, Male

0.00

12.405

0.000

9866415

-0.00

4.049

-0.000

11543400.24 11543400

0.00

4.737

0.000

7095624

0.00

2.912

0.000

10880067.29 10880067

0.00

4.465

0.000

9866415.18

62 65-74, Female
71 75 PLUS, Male

7095624.43

72 75 PLUS, Female

Sum of
% of
Target Weights Output Target % of Difference
Total Difference Weights
Weights
in %

Iteration
Number

Maximum Absolute Value
of Difference in Sum of
Weights

Maximum Absolute Value
of Difference in %

Coefficient of Variation of
Weights at the Completion
of the Iteration

13852461.44

5.6847

0.96179

4215616.04

1.7300

0.91393

2704257.36

1.1098

0.89457

1515962.64

0.6221

0.88412

775853.60

0.3184

0.87880

371418.76

0.1524

0.87635

167841.59

0.0689

0.87524

Mean

Min

Max

BSW_COMPOSITED_ATPT

Weight

39661.61

3063.71

97303.55

0.645

FINAL_WEIGHT

39661.61

4957.70 314634.25

0.875

C-7

Number of Respondents Who Had Their Weights Decreased by the Trimming: 101.
Number of Respondents Who Had Their Weights Increased by the Trimming: 123.
Raking output weight: RAKED_WT2

NHTSA Region: ACS and Weighted Sample Comparison
NHTSA
Region
1

ACS (16+)
11703417

WGT Sample
12521600

ACS %
4.8%

WGT
Sample %
5.1%

Difference in %
0.30%

32876263

32073388

13.5%

13.2%

-0.30%

24695552

25603780

10.1%

10.5%

0.40%

35308383

35904871

14.5%

14.7%

0.20%

40822356

39991792

16.7%

16.4%

-0.30%

29504175

26985086

12.1%

11.1%

-1.00%

13089479

14992540

5.4%

6.2%

0.80%

9682344

10738238

4.0%

4.4%

0.40%

35220310

32424118

14.4%

13.3%

-1.10%

10
Total

10943895

12445510

4.5%

5.1%

0.60%

243846174

243680923

100.0%

0.00%

C-8

DOT HS 811 865
December 2013

10081-120613-v4

File Type	application/pdf
Author	Administrator
File Modified	2013-12-11
File Created	2013-11-19