Census Monitoring & Tracking Research Plan

Census Monitoring & Tracking Research

Date: August 27, 2009

Background

The Census Bureau has asked DraftFCB to oversee monitoring and tracking research utilizing generic clearance for 2010 Census Communications Campaign Testing Activities, OMB#0607-0950. We, DraftFCB, plan to conduct primary research to monitor the performance of the campaign messaging in-market. The weekly monitoring and tracking results will be used to monitor potential issues with message comprehension and allow for rapid response to potential barriers to participating in the 2010 Census.

The specific marketing campaign goals are:

• Increase mail response

• Improve overall accuracy and reduce the differential undercount

• Improve cooperation with enumerators

The specific research objectives are:

• Monitor key attitudes and self-reported behaviors about Census among select segments and by geography through tracking key attitudes and self reported behaviors that are proven to be early indicators of real response

• Assess these early indicators of Census success as well as negative trends and “red flags”

• Monitor self-reported exposure to the 2010 communications campaign in totality

• Help explain low and high response to enable and direct real-time course correction of paid media, public relations, and/or Partnerships Program activities during the life of the campaign

• Provide indicators of the public's privacy and confidentiality perceptions and concerns before, during, and at the end of campaign

• Provide feedback on key messaging ideas about the Census

• Allow ability to ask new questions on a weekly basis.

Methods

Monitoring and tracking will employ a three phase approach: a benchmark phase October-Nov prior to the launch of the 2010 Census campaign; an awareness phase corresponding with awareness communication from Jan through mid-March 2010 and a motivation phase corresponding with motivation communication from March though May 2010. All interviews will be conducted online recruiting participants from a mix of online panels and river samples [See recruitment techniques on page 2 for more information on river samples]. Total sample for entire study will be approximately 21,798 people (in English only).

Target Audience

Sample Plan for Breaks by 9 Census Divisions	Diverse Mass	Asian* Augment	Black* Augment	Hispanic Augment* (in English)	Total Estimate
Benchmark Phase	1,800	164	56	56	2,076
Awareness Phase	8,100	738	252	252	9,342
Motivation Phase	9,000	820	280	280	10,380
Grand Total	18,900	1,722	588	588	21,798

*: The plan with the inclusion of augments ensures that there is a weekly read of 100 Asians, 100 Blacks and 100 Hispanics per week during the awareness and motivation phases and 100 per month for each audience during the benchmark phase.

Period of Performance and Reporting

Assuming OMB approval by September 9, 2009, as mentioned earlier, fieldwork would be scheduled from October-November 2009 and January-May, 2010. Topline reports would be issued twice during the benchmark phase, three times during the awareness phase and weekly during the motivation phase. Presentations to client and stakeholders would take place June-July 2010 and a final report will be issued in August 2010.

Report delivery dates listed below (Due by 7 P.M.):

Benchmark Period:

Begins: 10/5/10

1^st reading after 4 weeks in field – Report Due: 11/10/09

2^nd reading after 8 weeks in field - Report Due: 12/8/09

Awareness Period:

Begins: 1/4/10

1^st reading after 3 weeks - Report Due: 1/27/10

2^nd reading after 6 weeks - Report Due: 2/17/10

3^rd reading after 9 weeks - Report Due: 3/10/10

Motivation:

Begins: 3/8/10

Weekly readings: Report Due every Wed through 5/26/10

Recruitment Techniques (including incentives and length)

The vendor will utilize an online panel representing more than 800,000 households and 1.5 million individuals across the United States, along with the online sampling method of river sampling that drives a steady stream of potential participants to the survey for a one-off questioning (as opposed to being part of a panel and regularly receiving invites to participate in surveys).

The vendor is Applied Research and Consulting LLC (ARC), based out of New York, NY. Online sample for this study will be provided by DMS Research, one of the largest online survey sample providers in the industry. The survey will utilize both pre-recruited panel sample (SurveySpree®) and sample recruited at the time of the study (Opinion Place® River Sample).

For highly targeted online sample needs, the DMS Research SurveySpree® panel is deeply profiled across a range of topics. Representing more than 800,000 households, and 1.5 million individuals, members are recruited via banner ads across Internet sites (a more expensive, yet more representative, method) plus other opt-in techniques. Profiling includes data about decision-making, automobiles, health conditions, employment, leisure activities, shopping habits, and dining preferences, to name a few.

Opinion Place® River Sample

Opinion Place® is an online sampling method that utilizes millions of online advertisements to drive a steady stream of potential survey participants to DMS’s research portal. Visitors are identified, screened, surveyed, rewarded, and released, engaging all types of people, including the elusive casual survey taker. By building the concept of RDD purity into Opinion Place®, this unmatched online “River” provides a truly fresh, scientifically-managed sampling frame as an alternative to common online panels that rely on repeat respondents.

Qualification to Participate

Respondent must be between the ages of 18-64 and reside in the US. Sample will be Census-balanced (based on ASC 2007) by age, gender, income and education. In addition, at a bare minimum, regional balances will reflect the four Census regions [Northeast, Midwest, South and West].

We will also be able to read data by the 9 Census divisions [New England, Middle Atlantic, East North Central, West North Central, South Atlantic, East South Central, West South Central, Mountain and Pacific].

Payment to Respondents

It is standard practice in market research to offer recruited respondents some form of reimbursement for their time and effort. Respondents will receive the equivalent of approximately $2.00 in compensation for participating—a standard payment from this vendor for the 15-20 minutes anticipated time commitment for participating in this survey. For online panels, the reimbursement is usually provided as $2.00 in “points” given to each respondent by their online panel’s proprietary reward program. Respondents can accrue and/or redeem these points for a wide variety of items such as merchandise and gift cards for use as payment at physical and online merchants.

Assurance of Confidentiality Provided to Respondents

Respondents are informed prior to participation that their responses will be anonymous. They will also be advised of the nature of the activity, the length of time required, and that their participation is purely voluntary. Respondents are told that no penalties will occur if they wish to not respond, either to the information collection as a whole or to any specific question.

All presentation of data in analysis and reporting is in aggregate form with no links to individuals being preserved. Reporting is used only by project staff for purposes of analysis. Although some personal information is collected (e.g., gender, race, age), no personal identifiers (e.g., full name, address or phone, social security number, etc.) will be collected or maintained. We will also mention to respondents that the survey is being conducted by DraftFCB on behalf of the Census Bureau.

In order to ensure quality respondents, a combination of security locks, known as Security Suite, is used:

– IdentityLock: Validates a person’s identity through the use of both internal

and third-party databases and digitally fingerprints a visitor’s computer.

– UsageLock: Once an identity is established and confirmed, UsageLock

prevents too frequent participation and duplicates.

– In addition, data traps can be embedded in the survey to weed out

undesirable respondents.

Screening Burden

We expect to screen approximately 185,000 potential participants in order to get 21,798 qualified participants.

The total amount of screening burden hours, based upon all contacts made whether qualified and unqualified, is estimated at 9,250 based on 3 minutes of screening time.

Survey Burden

For the main survey portion of this research, the estimated burden hours for participants who qualify (21,798 people) would be of 7,266 based on 20 minutes of research time.

Questionnaire

A draft questionnaire is attached to this package for reference. Since messaging is not finalized yet, any changes made to this questionnaire will be reflected in a finalized version that would be submitted no later than one week prior to launch of the survey.

Total Burden (Screening & Survey Combined)

The total burden hours for all participants regardless of qualification: 16,516

The total burden cost approximate: $43,596 [Paid by sample provider to participants, neither Census Bureau nor DraftFCB, for being part of panel/river sample]

Contact

The contact person for questions regarding the design of this research as follows:

Eric Charles Newburger

U.S. Census Bureau

4600 Silver Hill Road

Mail Room

8H486F

Suitland, MD 20746

Phone: 301.763.3649

email:Eric.charles.newburger@census.gov

5