Attachment G - Advance Materials Re-evaluation

Revising the American Time Use Survey Advance Materials

Grace E. O’Neill and Margaret K. Vernon

Bureau of Labor Statistics, 2 Massachusetts Ave, N.E. Suite 4675, Washington, D.C., 20212

Abstract: Advance materials, often in the form of a letter or brochure, tend to increase response rates among survey participants. Advance materials are hypothesized to increase response rates because individuals tend to be more likely to cooperate with a survey request if they are given advance notice. Advance materials are widely utilized as a means of informing participants about an upcoming survey. However, there is very little systematic research on optimal design or content of advance materials for influencing cooperation among respondents. In addition, advance materials design often differs across surveys. As a part of the American Time Use Survey’s ongoing commitment to improve survey quality, a three-part study was undertaken to systematically reevaluate the advance materials. The three aspects of the advance materials study included a Response Analysis Survey, Interviewer Focus Groups, and an Expert Review. The reevaluation of the ATUS advance materials was undertaken to increase readership and to increase survey response. Design and content suggestions were used to make specific recommendations for revising the ATUS advance materials. Implications for both ATUS and general advance materials will be discussed.

Keywords: time use, advance letter, focus group

The American Time Use Survey (ATUS) is the first continuous, federally funded survey designed to measure people’s daily activities, including where they spend their time, what they spend their time doing, and with whom they spend their time. The ATUS is a one-time telephone interview with three main components: (1) the Current Population Survey (CPS) update questions, (2) the time diary, (3) and the summary questions. The ATUS updates three CPS sections including the designated person’s (DP)¹ employment status, his or her industry and occupation, and his or her earnings information. To complete the time diary, an interviewer asks the DP about how he or she spent his or her time over a 24-hour period starting at 4 a.m. on the day preceding the interview and ending at 4 a.m. on the day of the interview. The interviewer then asks the DP whom he or she was with and where he or she was for each activity. The summary questions include requests for additional information on secondary childcare, paid work, volunteering, and travel away from home.

The ATUS sample is drawn from households who have completed the entire CPS interview rotation. Once a CPS household is selected for ATUS, one household member is randomly selected to participate in the interview, without substitution or proxy response. The selected DP must be 15 years old or older and may or may not have been the CPS reference person. Each ATUS DP is also required to report on a pre-assigned reporting day. Fifty percent of the sample is assigned to report about a selected weekday--10 percent each for Monday through Friday. And, 50 percent is assigned to report about Saturday or Sunday--25 percent each. The specific day of the week assigned to each DP does not change, and there is no substitution of this day.

Previous advance materials research has shown that advance notice of a survey request, usually in the form of a letter, increases a respondent’s propensity to cooperate upon contact.² Advance materials often inform participants about why the survey is being conducted, who is conducting the survey, why the designated person’s participation is important, and the social usefulness of the survey. This information is thought to increase cooperation because it creates a sense of authority, legitimacy, and seriousness. The advance materials are also intended to increase the interviewer’s confidence and make the initial contact with the DP easier. All of these benefits, however, only occur if the correct person reads the letter and remembers its content upon contact.

The ATUS program, via Priority mail³, sends each DP an advance letter, an ATUS-specific brochure, and an incentive if the household is a no-telephone-number household⁴ (See appendix). The ATUS advance letter is docu-signed by the Census Bureau Director and contains information about the survey purpose, the survey sponsors, the voluntary and confidential nature of survey participation, and the target interview date. The ATUS brochure provides participants with additional survey information about the survey, the BLS web site address, and a Census Bureau monitored e-mail address in case they have further questions. Recently, the ATUS advance materials underwent a reevaluation. The reevaluation occurred in three steps, including a Response Analysis Survey, Interviewer Focus Groups, and an Expert Review.

Response Analysis Survey

In January 2004, a Response Analysis Survey (RAS) was conducted to help the BLS better understand a person’s propensity to respond or not to the ATUS, and to better understand which survey features are correlated with a given participant’s propensity to respond. The RAS was a paper-and-pencil telephone survey that collected information from 54 ATUS respondents and 49 ATUS refusals. Completed interviews averaged 10-12 minutes and all interviews were conducted by ATUS-trained Census Bureau interviewers. All participants were asked their participation reason(s), their attitude towards the CPS and the ATUS advance materials, and their general attitudes towards government and non-government surveys. In addition, ATUS respondents and ATUS soft refusals were asked about their attitudes towards the ATUS interview.⁵

Interviewer Focus Groups

In September 2004, two Interviewer Focus Groups were conducted at the Census Bureau’s Jeffersonville, IN telephone call center. Each group contained about six interviewers, both men and women, and at least one supervisor. Both groups were moderated by a trained BLS employee, and observed by a member from the ATUS staff.⁶ The Interviewer Focus Groups were designed to gather the interviewers’ attitudes, opinions, and suggested modifications for the advance materials. In addition, the focus groups were designed to elicit comments that interviewers might have heard from ATUS respondents during the course of an interview. Both focus groups were audio recorded and a note taker was present in each group.

Expert Review

In addition to the Interviewer Focus Groups, two sets of expert reviews were conducted. The first set involved an in-person consultation with several BLS survey methodologists. The second set involved an e-mail and telephone consultation with several non-BLS survey experts. Their recommendations were then combined with expert feedback from earlier consultations.

Findings

The reevaluation of the ATUS advance materials was undertaken for two main reasons: to make design and concept changes that would improve readability and clarify conceptual problems. Suggestions were used to make specific recommendations for revising the ATUS advance materials. These revisions were expected to improve the advance materials. This expectation, however, will not be tested in an experimental setting as any change to the overall response rate is confounded by other survey factors. While advance material experimental testing is possible, it is not feasible at this time and still may not conclusively show if the revised advance materials are more readable and increase participation in comparison to the old materials.

Readability

A common theme across the RAS, the Interviewer Focus Group, and the Expert Review was that DPs are not reading the advance materials. As stated above, the benefits of the advance materials depend on the correct person reading the materials and remembering them when called for an interview. Thus, these studies indicate that the advance materials should be revised to increase the likelihood that the reader will open the mailer, read its content, and make a clear association between the advance materials and the subsequent telephone call. One way to increase this likelihood is by modifying the design elements of both the advance letter and the brochure.

In the RAS study, 61 percent of ATUS respondents who completed the RAS remember receiving the mailer. Of that 61 percent, all reported that they had read the letter, but only 45% of them said that they had read the brochure. Of the ATUS nonrespondents, 38% remembered receiving the advance mailer; however within that group, only two-thirds reported reading the letter, and only one-third reported reading the brochure. While the RAS did not ask respondents to make any specific suggestions of ways to increase readership, the study did demonstrate that the advance materials should be improved and that in comparison to the letter, the brochure is a less effective means of communication.

In the focus groups, interviewers suggested a brochure that looks less official, and is more eye-catching, and visually appealing. Specifically, they suggested adding statistics, shortening the text length, and adding visual cues, like bold typeface, color, and graphics. When asked, most interviewers favored the use of statistics in the advance letter. Some interviewers preferred them in the beginning of the letter, while others liked them in the letter text; however they all agreed the statistics have to be general enough to appeal to the broad survey audience. Regarding visual changes, one interviewer suggested using a colored Department of Commerce seal to draw the reader’s attention to the letter top and to include an arrow after the letter text to visually indicate that the letter is double-sided. Some interviewers suggested including the ATUS web site address on the letter bottom, in addition to mentioning it in the brochure. They also thought that different fonts and bold typeface would make the letter look less official and would emphasize they key information including that the Census Bureau is requesting ‘your help,’ the interview date has been pre-set, that it is a one-time survey, as well as the Census Bureau phone number and the ATUS e-mail address and web site. Many interviewers were also concerned that all of the interview-specific information in the last paragraph might not be read. They suggested either moving the information to the first paragraph or highlighting it in a way that would draw the reader’s attention to the information. The interviewers were also asked if they thought the letter was more effective single-sided or double-sided. Most felt that a single sheet of paper double-sided was better because they thought a bulkier letter would increase people’s likelihood to throw the letter away rather than read it. Interviewers also felt that some people were less likely to read any information that was not immediately visible to the reader, like the frequently asked questions on the back of the advance letter.

For the brochure, the interviewers suggested a brochure that was brighter, had different graphics, and included statistics. The interviewers also felt that the brochure was too broad and needed to focus more on the specific designated person. Both interviewer focus groups agreed that the blue color was too dark and most interviewers suggested a brighter color like red, yellow, or purple. Most interviewers said they would prefer an image of a real clock on the front cover rather than the sundial because they felt the clock would be a friendlier graphic. They also did not like the gavel because it seemed too official and threatening. Several interviewers also did not understand the school bus, specifically thinking that it would confuse the brochure’s reader, and suggested replacing it with another graphic. One interviewer suggested a cartoon cut-away of a house with the different household members doing different activities as a replacement for the sundial or the bus. The interviewers also thought the people graphic was good but should include people eating and watching TV since those were commonly reported activities. Regarding the text, most interviewers liked the existing brochure text but would add statistics or newspaper headlines and change the section order. The interviewers also felt that the phrases “unpaid work” and “policymaker” should be dropped or reworded to make the concepts clearer to the respondent. When asked, the interviewers had mixed feelings about the confidentiality information in the brochure. Some thought that it was important but should be written in a simpler language while others thought that it should only appear in the advance letter.

In the Expert Review portion of the advance materials re-evaluation, the experts suggested several ways to increase advance material readership, including the use of statistics and simple visual modifications. Experts specifically thought statistics could be used as eye-catching headlines and suggested citing a statistic from a major paper that used ATUS data. The experts also suggested that presenting the statistics in a visual pie chart would be more effective than as text in the brochure format. Visually, the experts liked the clean and simple appearance of the brochure and, unlike the interviewers, also liked its blue color. They agreed that the school bus was confusing, especially if a respondent did not get the childcare connection; but they were unsure of what should be used in its place. The experts also liked the people graphic but were concerned that respondents might not see themselves represented in the graphic and thus might lose interest in the survey. The experts also felt that the ATUS web site should be mentioned on the brochure front rather than the back.

One expert reviewer suggested including the household’s last CPS interviewer’s name in the first line of the advance letter. Both the interviewers and other experts thought that this could confuse the DPs because each household has multiple CPS interviewers and they may not remember their last interviewer’s name. This would be especially true if the last interview was a telephone interview. Another concern mentioned was the risk of accidentally using the wrong CPS interviewer’s name. Both groups thought that the addition of another potential error source was not worth the potential increase in response and agreed that any mention of CPS should not be respondent-specific. The experts were also concerned that over-personalizing the letters might appear as a breach of confidentiality which could cause a refusal or tarnish the government’s image.

In summary, previous research has shown that advance materials increase cooperation at initial contact; however this benefit requires that the respondent read the materials and remember them when contacted. The RAS study suggested that one way to increase advance material readership is through design and content modification. In the focus groups and expert reviews, both the interviewers and the experts suggested ways to make the advance materials more visually appealing in order to increase readership. Specifically, they recommended changing the current brochure graphics and adding statistics. The interviewers also recommended shortening the letter text and adding visual cues like fonts and graphics.

Common Questions and Misconception

The second reason for the ATUS advance materials reevaluation was to clarify several misconceptions respondents had about the survey. Throughout the reevaluation, it became apparent that there were several concepts in the advance materials that confused ATUS respondents. The content of both the advance letter and the brochure should serve to communicate important survey-related information with the respondent and it is vital that this information be interpreted accurately. First, the distinction between the ATUS and the CPS was not always clear. Second, respondents would like to know more in advance about the ATUS time diary methodology prior to being contacted. Finally, respondents want to understand why they, specifically, have been selected to complete the ATUS and why it is important for them to complete the survey.

Differentiating the ATUS from the CPS

An important conceptual issue that arose was that the DPs often do not understand that the ATUS is a different survey than the CPS. This is an uncommon problem in surveys, as few government surveys use retired respondents from other surveys as sampling frame. Adding to the confusion, in ATUS, individuals are sampled from a household-level frame, rather than a person-level frame; some DPs may be unaware that their household recently participated in the CPS. Originally, it was thought that mentioning the CPS in the ATUS advance letter would increase a DPs response propensity by invoking their familiarity with CPS and the Census Bureau. Mentioning the CPS was also expected to invoke their sense of consistent survey participation and tap into their overall helpful nature. Both the RAS and Interviewer Focus Groups, however, demonstrated that mentioning the CPS in the advance materials is often more detrimental than beneficial.

In the RAS study, 33% of the ATUS nonrespondents stated that their decision not to participate in ATUS was because of their previous CPS participation. In contrast, only 9% of ATUS respondents stated that their decision to participate in ATUS was because of their previous CPS participation. Of those that said that they did not participate because of their previous participation in CPS, some respondents said they did not understand why they were sampled for two different surveys. Others felt that participating in both surveys was doing more than their fair share. While clarifying the difference between the surveys and explaining why people are selected to do both in the advance materials is not likely to reduce respondent fatigue, it may increase cooperation, because an informed respondent may be more willing to accept the additional burden.

Similar to the RAS findings, the focus group research indicated that interviewers also felt that ATUS DPs did not understand the difference between the two surveys and refused the interview based on their feeling that they had completed the CPS. Overall, most interviewers felt that the advance materials needed to focus more on the ATUS and less on the CPS. When further probed, several interviewers felt that mentioning the CPS might cause a negative reaction by the DPs and should not be mentioned anywhere in the advance materials. Others thought that mentioning the CPS only once in the last paragraph of the letter might be a good solution. The interviewers also thought that the advance materials should better address several of the DP’s misconceptions about the two surveys. Specifically, the interviewers felt that the DPs did not understand that the ATUS is a one-time telephone survey and not a multiple-rotation personal visit survey like the CPS. DPs also did not understand that the ATUS requires both a specific respondent and a specific reporting day. In general, interviewers thought that the information in the advance materials should provide the reader with a clear understanding of the ATUS and mention the CPS only when necessary.

The expert review produced mixed results regarding role of the CPS in the ATUS advance materials. Some experts felt that if CPS participation is a strong response propensity factor then mentioning the CPS should increase response rates. Other experts, however, felt that the CPS only works as a reason not to participate and that mentioning CPS primarily increases refusals. Throughout the expert review, the participants discussed the benefits and downsides of mentioning CPS in the advance materials and decided that they did not have enough information to come to a final conclusion.

Clarifying the ATUS time diary

A second theme that emerged from these studies was that the ATUS DPs are unprepared for their time diary experience. The ATUS time diary format is one which is not commonly used in surveys. For most respondents, the ATUS is their first time-diary experience and the interview is very different from their previous CPS experience. The DPs are further confused because the ATUS begins with CPS questions and then goes into the unfamiliar time diary⁷.

Throughout the RAS, participants stated that they would have been better prepared for the ATUS interview had they known the nature of the diary questions prior to the interview. Of the RAS participants who read either the letter or the brochure, 30% of ATUS respondents and 7% of ATUS nonrespondents said that they would have liked to have had more information before starting the interview. Most participants would have liked either an advance copy of the questionnaire or a better understanding of how the diary portion of the interview was going to be conducted. This, however, has not been implemented because of a concern that more diary-specific information could alter the respondent’s daily activities or cause other participants to refuse the interview. Another way to prepare DPs for their time diary experience is through providing information on the ATUS web site and through e-mail (on request). Providing more information there offers detail to the group of respondents that seek it. When asked, however, most participants had neither visited the web site nor sent an e-mail question--even those stating that they had wanted more information about the survey. One reason for the information disconnect might be that the web site and e-mail address were listed only on the last page of the existing brochure. Thus, those participants that did not read the brochure did not get the additional information.

Similar to the RAS study, interviewers also found that respondents were surprised by and resistant to the time diary format. The interviewers stated that respondents were hesitant not only about reporting their activities but also about revealing who was with them during each activity. The diary start time also concerned some respondents because they felt it was intrusive or unnecessary for the government to know what they were doing at 4 a.m. Some respondents were unsettled by these questions, failed to understand why the government was collecting this type of information, and ended the interview at the start of the diary. Interviewers, however, thought that these break-offs were not because respondents were unwilling to complete the time diary but because they we unprepared for it. They suggested that if respondents were better prepared, they would be more likely to complete the interview. The interviewers suggested several ways to better prepare the respondents including mentioning the time diary more prominently in the advance materials, providing the respondent with example summary questions or diaries, and providing respondents with clear examples of how the data are used. The interviewers, however, also advised caution about providing too much information as it may appear overwhelming to the respondent and cause a refusal prior to contact.

The experts thought that the advance materials should prepare the DPs for the ATUS interview by informing DPs about the exact nature of the interview in order to alert the respondent about what to expect upon contact. The experts liked the explanation of time use in the brochure over that in the letter and thought that the materials should avoid using vague terms like policy makers or unpaid work. Experts also like the brochure’s activity list and the advance letter’s sampling methods explanation. In conclusion, the experts thought that the advance letter should inform the DPs about the ATUS, while the FAQs and the brochure should reduce their anxiety about the interview.

Respondent importance

The final conceptual theme involved making the DP feel personally important to the ATUS and making the ATUS personally important to the DP. The advance letter’s purpose is to request the DPs help in completing the survey request upon contact. This request must be crafted in a way that not only conveys all the necessary information to the reader, but also conveys it in an appealing manner. Most advance letters are written using one of two main appeal types: altruistic, which emphasizes the survey’s benefits to a larger group or community, and egoistic, which emphasizes the direct gain respondents receive from completing the survey.⁸ The current ATUS advance letter incorporates both appeal types, however it only indirectly mentions more egoistic appeals.

The RAS study found that the ATUS needs to emphasize respondent–level participation reasons and use a more egoistic appeal in the advance materials. Specifically, the RAS study suggested that language needs to convey to the respondent why he or she was selected, over a different individual. Overall, the RAS study showed that the advance materials not only need to clearly explain concepts to DPs, but also explain them in a way that appeals to respondents personally.

The interviewers’ suggestions also blended altruistic appeals with more respondent-oriented egoistical appeals. One suggestion was to explain sample selection in statistical representation terms and demonstrate that each DP was selected for specific statistical reasons and not random or haphazard reasons. Interviewers were especially concerned about addressing the needs of the elderly and of minors. Currently, the advance materials do not mention activities specific to the elderly, and many elderly do not feel that their responses are important because they do not have jobs or children. Equally, many parents do not understand why their minor child was selected to participate, and think that the Census Bureau contacted them by mistake. Interviewers also suggested ways to make the ATUS personally important to the DPs, including explaining data on a more egoistic level and mentioning the ways other countries use their time-use data.

In line with findings from the RAS and the Interviewer Focus Group, the experts also felt that an egoistic appeal was necessary in order to encourage respondents to participate in the ATUS. The experts felt that DPs need to know the ATUS is conducted on their behalf. The experts echoed several of the interviewers’ suggestions and also suggested emphasizing “you,” “your time” and “your participation,” in the text of both the advance letter and the brochure. The experts, however, cautioned against over personalizing the letter, which could make the government appear too knowledgeable or solely interested in specific groups.

In addition to the advance materials needing to appeal to respondents in altruistic and egoistic ways, the advance materials need to lend a sense of legitimacy to the study. Previous research (i.e., Groves and Couper, 1998), has found that government agencies often obtain higher response rates than other organizations because of the high amount of authority and legitimacy respondents perceive government agencies to have. Findings from the RAS study also suggest that government sponsorship positively affects some DP’s response propensities. This is important for the ATUS, because government sponsorship immediately makes the survey important to some respondents, regardless of appeal types or graphic design. Thus, the ATUS advance materials will need to strike a balance between catchy marketing-type graphics and official government seals and mentions of the agencies in text.

In summary, the ATUS advance material reevaluation was conducted to determine how to increase survey quality through redesigning the advance materials to be more appealing and convincing. Recommendations to increase readership included making the advance materials more visually appealing, adding statistics, and shortening the letter text. Recommendations to reduce conceptual issues included ways to help the DPs differentiate between the ATUS and the CPS, ways to better prepare DPs for their ATUS experience, and ways to make the DPs feel more personally important to the ATUS. All three reevaluation methods provided useful insight into these content areas. The RAS clarified types of changes that may need to be addressed, while the Interviewer Focus Group and the Expert Review provided more concrete suggestions for doing so. These suggestions were used to redesign the advance materials.

Revisions made to the ATUS advance materials

The ATUS advance materials were revised based on the findings from the Response Analysis Survey, Interviewer Focus Groups, and Expert Review. The advance material revision was divided into two sections: (1) changes made to improve readability and (2) changes made to address common misconceptions about the ATUS. Hopefully, these revisions will increase DP cooperation rates and ultimately increase the survey’s overall response rates. The paper concludes with a discussion of general implications for other surveys’ advance materials.

Readability

In general, the reevaluation showed that the advance materials required several aesthetic changes. First, the letter was drastically shortened and the information on the letter’s second side was moved to the brochure. This revision made the letter look more inviting, and less daunting to read.⁹ It was assumed that the shortened letter would increase overall brochure readership. Restricting the letter to one page alleviated the concern that respondents were not reading the letter’s second side. Second, the interviewers suggested that the letter should look less formal. The RAS study, however, found that many people participated because the ATUS is an official government survey. Therefore, the brochure was designed to look informal, while the letter remained relatively formal. Finally, certain key words, such as “one-time-only” were highlighted in the letter with bold typeface.

Several aesthetic changes were also made to the brochure. First, the BLS graphics department changed the overall orientation of the brochure. The brochure was changed from a horizontally opening tri-fold format to a vertically opening accordion format. Second, the brochure was changed from a dark (primarily blue) colored brochure to a brighter, multicolored brochure. Next, new graphics were included to show a wider variety of people including teens and the elderly, engaging in a broad variety of activities. In order to provide further legitimacy to the survey, the last brochure panel now includes an American flag background in addition to the official government names and emblems. Finally, several changes were made to the brochure text including format, content, and wording changes. The format was changed to a question-and-answer format, which allowed respondents to skim the brochure and to read only the topics of personal interest. Content was included to highlight ATUS estimates and publications featuring the survey. Wording was also modified to eliminate discipline-specific jargon and to simplify the brochure language.¹⁰

Common Questions and Misconceptions

The advance materials were also modified to clarify several misconceptions respondents had about the survey. The reevaluation showed that some respondents do not understand the difference between the CPS and the ATUS. This misconception often leads to negative affect towards the survey and possibly a refusal. The main change was to remove all mentions of the CPS from the letter and to include only a brief mention in the brochure. In addition, the advance letter’s first sentence states that the ATUS is a “one-time survey,” and implies that it is not a multiple-rotation survey like the CPS. In the brochure, questions and answers were added to explain why the respondent was selected for another survey and why it is important for them to respond. An FAQ was also added to explain the ATUS sampling rules and why respondent proxies and substitution were not permitted.

Some respondents were also overwhelmed by the ATUS time-diary task. Originally, it was believed that advance information about the time diary, including sending advance questionnaires, might influence the respondent’s behavior or cause a refusal to the survey request. The reevaluation, however, suggested that interview break-offs might be prevented if DPs were better prepared for the survey. A balance, therefore, was struck: the brochure generally explains what the survey is about and what kind of questions the respondent will be asked. The brochure also more prominently displays the survey’s e-mail address, website, and phone number.

Finally, the new advance materials were modified to convey the survey information in a more personal manner. In the advance letter, the phrases “your help” and “one-time only” were boldfaced in order to emphasize the DPs personal role in the survey process. Several of the brochure’s FAQs were added in order to show respondents that their participation is import to both their community and to people like themselves. These FAQs include information about specific sources where the data have been published, specific results from 2003 data, who might use the data to develop policies and programs that will help people like the respondents, and what types of questions the data might help to answer in the future. Several FAQs were also added to the brochure to address the importance of elderly and teen participation.

In summary, revisions to the advance materials were undertaken to address two major themes. First, visual and practical changes, such as shortening the letter length to decrease burden, were made in the hopes that respondents would be more willing to read the advance materials. Changes were also made to address common questions and misconceptions that respondents had about the ATUS. The majority of changes made to help respondents better understand ATUS were content changes, for example, adding information in the brochure about common questions, and taking away information that confused respondents. The changes made to the advance materials were informed by the RAS study, and addressed specific suggestions made in the Interviewer Focus Groups and the Expert Review.

Advance material revisions for other surveys

Other surveys, both federal and private, can also benefit from the ATUS advance material reevaluation. In order to increase respondent cooperation, the advance materials should be eye catching, interesting, and easy to read. Visual cues like bold typeface, graphics, bright colors, and survey-related statistics all aid readability. Graphics and statistics, however, require careful selection since they visually represent the survey to the reader.¹¹ Thus, it is important to select graphics and statistics that are general enough and interesting enough to appeal to all potential readers. Brevity is assumed to improve cooperation. By only including information essential to a respondent, a survey can make the advance letter appear less daunting and, hopefully, increase a DP’s propensity to read the letter. Any additional information should be provided in complimentary advance materials (such as a brochure), or via a web site, an email address, or a telephone number. Web site addresses, email addresses, and telephone numbers should be prominently displayed in the advance materials. Deciding which information is essential and which information is secondary can be complicated. A conservative assumption is that a respondent will only read the first few sentences completely and will then scan the remaining letter and any additional materials for important information. One reason the ATUS staff decided to shorten the advance letter was to increase the reader’s propensity to read the entire letter and at least scan the brochure for information that might be of interest. By adding bold typeface and off-setting the questions in the brochure from the answers, the advance materials are not only easier read, but are also easier to scan. Another way advance materials help to increase respondent cooperation is by reducing misconceptions about the survey, preparing the respondent for the survey task, and providing the information in everyday language.

While most samples are not drawn directly from another survey, respondents may still confuse surveys or misunderstand different methodologies. Therefore, it is important to mention not only attractive parts of a survey’s methodology, like a Web option, one-time data collection, or an option of personal visits, but also methods that might confuse a respondent, like proxy rules and special sample requirements. Another important facet of the advance materials is preparing the respondent for the survey task and easing any respondent concerns. This may include sending the respondents sample questionnaires or web sites where they can learn more about the survey. Finally, it is important that the survey sell itself on a personal level. While many respondents participate for more altruistic and community based reasons, it is also important to focus on the individual and what they gain from participation. This is especially important for particular groups like the elderly. Again, deciding which information is essential and which information is secondary can be complicated. However, it is important to keep the survey information simple and brief and include further contact information for interested respondents.

Advance materials increase respondent cooperation by introducing them to the survey, explaining their role in the survey, and demonstrating how their response is important to both individual respondents and the larger community. In order for these three goals to reach the respondent, a survey organization has to provide a readable document that contains information in a manner which is appealing to a general audience. Often the most difficult challenge facing advance material authors is what to include and what not to include. Conducting focus groups with respondents, interviewers, and survey experts provides valuable insight for these content decisions.

1 A designated person is the selected household member prior to interview contact.

2 Groves and Couper, Nonresponse in Household Survey, New York, John Wiley & Sons, 1998.

3 Pre-production testing showed that response rates were significantly higher among households that received the advance materials via priority mail (71%) compared with first class mail (58%). See Piskurich et al., 2002 AAPOR proceedings

4 Households that did not provide telephone numbers in CPS are sent a $40 debit card and asked to call the Census Bureau (toll free) in order to complete their interview

5 For more information on the RAS study, see O’Neill and Sincavage, 2004, Response Analysis Survey: A Qualitative look at Response and Nonresponse in the American Time Use Survey (http://www.bls.gov/ore/pdf/st040140.pdf).

6 In general, it is not recommended that survey-related personnel conduct or attend focus groups; however, and exception was made because the topic was not sensitive.

7 Immediately following the RAS, some wording changes were made to the section introductions to minimize confusion. Changing question order to place the diary first was also considered and rejected, as the CPS questions provide important input to the diary section.

8 For a review of advance letter appeals, see Redline et al., AAPOR 2004 proceedings

9 In shortening the letter’s length, the extensive and technical, but legally required, Office of Management and Budget (OMB) statement was placed in a footnote at the bottom of the letter.

10 Both the letter and brochure will be translated into Spanish and the Spanish version will appear on one side of the page while the English version will appear on the other side.

11 For more information on the impact of visual images on survey response, see Couper, M.P., Tourangeau, R., & Kenyon, K. (In press). “Picture this! Exploring visual effects in Web surveys.” Public Opinion Quarterly

10