Attachment C - Focus Group Redesigned Final Report

Attachment C - NSDUH Focus Groups for Redesigned Contact Materials Final Report

2009 NATIONAL SURVEY ON DRUG USE AND HEALTH

Report oN the CONTACT MATERIALS FOCUS GROUP FINDINGS AND RECOMMENDATIONS

Contract No. 283-2004-00022

RTI Project No. 0209009

Authors:	Project Director: Thomas G. Virag
Doug Currivan
Emilia Peytcheva
Sonia Rodriguez
Stephanie Stolzenberg
Jim Leiman (Morpace, Inc.)
Arturo Obscura (Morpace, Inc.)

Prepared for:

Substance Abuse and Mental Health Services Administration

Rockville, Maryland 20857

Prepared by:

RTI International

Research Triangle Park, North Carolina 27709

November 2009

2009 NATIONAL SURVEY ON DRUG USE AND HEALTH

Report oN the CONTACT MATERIALS FOCUS GROUP FINDINGS AND RECOMMENDATIONS

Contract No. 283-2004-00022

RTI Project No. 0209009

Authors:	Project Director: Thomas G. Virag
Doug Currivan
Emilia Peytcheva
Sonia Rodriguez
Stephanie Stolzenberg
Jim Leiman (Morpace, Inc.)
Arturo Obscura (Morpace, Inc.)

Prepared for:

Substance Abuse and Mental Health Services Administration

Rockville, Maryland 20857

Prepared by:

RTI International

Research Triangle Park, North Carolina 27709

November 2009

Table of Contents

Chapter Page

1. Introduction 1

2. Design and Methods 3

2.1. Purpose and Goals of Focus Groups 3

2.2. Focus Group Protocol and Procedures 4

2.3. Recruitment of Participants 5

2.4. Characteristics of Participants 6

3. Results 9

3.1. Lead Letter Envelope 9

3.1.1. Considerations in Deciding Whether to Open a Piece of Mail 9

3.1.2. Reactions to the Two NSDUH Envelopes 10

3.2. Lead Letter Text 11

3.3. Lead Letter Graphics 14

3.4. Question and Answer Brochure 15

3.5. Special Concerns of Spanish-speaking Participants 17

3.6. Additional Suggestions 18

4. Summary and Recommendations 20

4.1. Lead Letter Envelopes 20

4.2. Lead Letter Text 21

4.3. Lead Letter Graphics 21

4.4. Q&A Brochure 22

5. References 23

List of Tables

Table Page

Table 1. Characteristics of English Focus Group Participants 6

Table 2. Characteristics of Spanish Focus Group Participants 7

Table 3. Preference Counts for Each Type of Contact Material 18

This page intentionally left blank.

1. Introduction

The National Survey on Drug Use and Health (NSDUH), sponsored by the Substance Abuse and Mental Health Services Administration (SAMHSA), is a national survey of the U.S. civilian non-institutionalized population aged 12 and older. The conduct of NSDUH is paramount in meeting a critical objective of SAMHSA’s mission to maintain current data on the prevalence of substance use in the United States.

Like other surveys, the potential for nonresponse bias is an ongoing concern. Nonresponse bias is a function of the nonresponse rate and differences between respondents and nonrespondents on key survey variables. On NSDUH, the key survey items are substance use prevalence rates. Historically, interview response rates have been lower among some demographic subgroups, such as those 50 and older (50+) (Murphy, Eyerman, and Kennet, 2004), and to a lesser extent those living in urban areas and males (Table B.5 in the 2006 Summary of Findings). Recent NSDUH data indicate sample members who are more likely to be underrepresented in the interview data:

• Men are somewhat less likely than women to participate.

• Adults over 50 years old, and especially those 65 and older, are somewhat less likely than adults under 50 to participate.

• Respondents who are categorized in the Other race subgroup and White respondents who are not Hispanic are somewhat less likely to participate than respondents who are Hispanic or non-Hispanic Black.

Lower response rates among subgroups may increase bias in overall estimates of prevalence rates. Increasing response rates across all groups would help minimize bias.

One way to ensure high participation across subgroups in the population would be to improve any aspects of the survey protocol that could directly influence the decision to participate. In the spring of 2005, RTI carried out 12 focus groups with potential respondents to examine the issue of nonresponse among persons 50 and over (Murphy, Schwerin, Hewitt, and Safir, 2005). This study was designed to explore how persons age 50 and over perceived the NSDUH interview request. Subjects were shown a video of an NSDUH interviewer contacting a sample member, enlisting cooperation to complete the household screening, and then enlisting cooperation for a selected household member to conduct the interview. Participants were also provided with copies of the materials given to respondents, including the lead letter, question and answer (Q&A) brochure, and other materials. Focus group members were then asked questions about the screening and interview process, as well as the contact materials being used during these processes. One set of recommendations from this study focused on examining how well the NSDUH contact materials:

• establish the legitimacy of the sponsoring and research organizations,

• clearly convey the survey objectives and importance of participation, and

• describe the selection process and importance of the selected individual’s participation.

Conveying this information to sample members is clearly dependent on the screening respondent actually seeing the materials. The results of a cursory analysis of whether the screening respondent recalled seeing the lead letter in the 2001 NSDUH¹ indicated that approximately 48 percent of respondents aged 50+, 48 percent of respondents aged 35 to 49, 45 percent of respondents aged 26 to 34, and 35 percent of respondents aged 18 to 25 recalled seeing the lead letter.

The investigation described in this report was designed to utilize feedback from focus groups to identify ways to improve the NSDUH contact materials to maximize cooperation among sample members. To this end, SAMHSA and RTI undertook two specific tasks. First, alternative versions of the lead letter envelopes, lead letters, and question and answer (Q&A) brochure were developed based on survey design principles usually associated with the decision to participate (Groves, Cialdini, and Couper, 1992).The final versions of the lead letters and Q&A brochures were also translated into Spanish.² Second, focus groups with participants from different parts of the United States were conducted to evaluate how members of the target population would react to the different versions of the contact materials. This report describes the procedures followed and results obtained from 17 focus groups involving both English and Spanish-speaking adults across the U.S. In addition to describing the key themes emerging from these focus group discussions, this report presents general recommendations for revising the contact materials based on the findings.

2. Design and Methods

   1. Purpose and Goals of Focus Groups

Given that the purpose of the overall investigation was to improve the design of the NSDUH contact materials in ways likely to maximize cooperation rates, a key objective of this study was to evaluate alternative versions of the materials through focus group discussions. The contact materials developed for the focus groups included multiple versions of the lead letter envelope, the lead letter text without graphics, the lead letter graphics without text, and the Q&A brochure. The revised contact materials resulting from this study will likely be used for the upcoming NSDUH redesign, which is currently slated to begin in 2013.

The feedback provided by focus groups will assist in identifying which of the lead letter envelope, lead letter, and Q&A brochure designs might maximize positive responses to requests to complete the screener and interview among households selected for NSDUH. As Fowler (1995) notes, focus groups can quickly broaden researchers’ perspective on how people think about the issues under study. As such, the focus groups conducted for the NSDUH contact materials can help in two ways:

1. To examine assumptions about the likely reactions members of the target population would have toward the materials and participation in the survey and

2. To evaluate assumptions about how members of the target population understand the key concepts and specific terms used to describe the survey.

The focus group sessions primarily provided participants’ responses to a set of structured questions and probes, but also allowed participants to add spontaneous comments on the materials. The qualitative information derived from the focus groups included both overall reactions to the contact materials as well as specific feedback on individual elements. We also planned to count participants’ preferences for specific versions of each type of contact material, to provide some quantitative context for the reactions and feedback.

Another important goal of the focus groups was to involve members of the target population who represented different regions of the country and different languages spoken. For this reason, a total of 17 focus groups were conducted across five metropolitan areas – Chicago, Dallas-Fort Worth, Los Angeles, Raleigh-Durham, and Washington, D.C. To ensure representation of U.S. residents who speak primarily English and those who speak primarily Spanish, 11 of the focus groups were conducted in English and six in Spanish.

A final objective in conducting the focus groups was to ensure the groups were conducted by experienced professionals who did not have a direct stake in the specific study materials or protocol. This facilitated a more objective treatment of the issues and ensured that detailed knowledge of the NSDUH did not unduly influence the direction or content of the focus group discussions. RTI hired professional focus group moderators from Morpace, Inc. to conduct both the English and Spanish focus groups. In addition to having extensive experience in conducting focus groups, the Morpace moderators also had experience and training in survey research methods.

2. Focus Group Protocol and Procedures

The focus group protocol guide was developed by RTI and SAMHSA, translated into Spanish by RTI, and distributed to the Morpace moderators to study. The protocol is included with this report as Appendix A. The protocol consisted of six major sections:

1. informed consent,

2. introduction,

3. lead letter envelope discussion,

4. lead letter discussion,

5. Q&A brochure discussion, and

6. concluding remarks.

The informed consent part of the protocol took about five minutes and provided an overview of the study and specifics about the participants’ rights as study volunteers. The participant informed consent form is included with this report as Appendix B. The introduction also lasted about five minutes and was intended to set up the discussion rules and acquaint the participants in each group. The discussion of the lead letter envelopes was designed to take about 15 minutes and involved discussions of (1) how respondents screen their mail and (2) how they perceive alternative versions of the envelope. About 45 minutes was allotted for the discussion of the lead letter and intended to cover both alternative versions of the letter text and graphics. The section on the Q&A brochure was planned for about 30 minutes and was aimed at comparing the current brochure (version 1) with a redesigned version (version 2). The discussion of the Q&A brochure involved participants’ preferences in terms of the content, visuals, and design. At the completion of the discussion of each type of contact material, the moderator encouraged participants to indicate which version they preferred. The concluding section of the protocol was intended to give both participants and observers a final opportunity to ask questions or make comments. At the end of each focus group session, each participant received $75 in cash and signed a receipt for the payment. The incentive receipt form is included with this report as Appendix C.

Both the English and Spanish focus groups followed the same protocol. Up to 10 participants comprised each focus group, with most sessions including seven to nine people. Upon entering the room, participants were presented with two copies of the informed consent form. The moderator briefly explained the consent form and asked participants to read over it, sign both copies, return one to the moderator, and keep the other for their records. Next, the moderator introduced some ground rules related to cell phone use, taking breaks, and side conversations among participants. He also encouraged participants to share their opinions, especially when they differed from the views expressed by other participants. The moderator explained that each session was video and audio recorded and briefly introduced himself. He then went around the room and asked each participant to briefly introduce himself or herself.

After the introduction, the moderator began following the protocol for the discussion of each type of contact material. The moderator guided participants through the protocol by asking questions about the material and following up with probes to address specific points that were not mentioned by participants. As appropriate, the moderator expanded the discussion to build on participants’ comments and asked others to share their views. A key strategy used by the moderators was to direct questions to participants who had not yet contributed to the current discussion. This ensured that most participants were heard throughout the session, so that the impressions gleaned from the groups were not based solely on a few participants who were most forthcoming.

Following the discussion of each type of contact material, the moderator asked participants to mark the version of the material they preferred with the number one. Not all participants chose to select a preferred version for each type of material. Although some participants expressed mixed feelings about specific features of different versions, in most groups a majority of participants did indicate a preference for each type of contact material.

3. Recruitment of Participants

RTI recruiters posted advertisements on craigslist.com for the Raleigh-Durham, North Carolina, Washington, District of Columbia, Dallas, Texas, Chicago, Illinois, and Los Angeles, California metropolitan areas. These advertisements explicitly noted that in order to be eligible for participation, the respondent must be age 18 or older, not currently serving on active duty in the military, not currently employed by RTI International, and not a family member of an RTI employee. The text for the recruitment advertisement for English focus groups can be found in Appendix D.

A key consideration for our recruitment efforts was to reach Spanish-only, or mostly Spanish, speakers who would not be included in the pool of potential participants in the English focus groups. Our recent experiences on recruiting this population for focus groups and cognitive interviews indicated that working with local Hispanic/Latino community organizations was an effective way to tap into this subpopulation. RTI bilingual recruiters partnered with local community organizations to legitimize the research in these target communities by assisting with recruiting and providing a facility at their site for hosting the focus groups. Advertisements were posted in highly concentrated Hispanic areas, such as Hispanic shopping centers, community centers, and La Raza, a Spanish-speaking newspaper for the Chicago area. Word of mouth was the most valuable way of getting calls from prospective participants. The local community centers assisting with these efforts were El Pueblo in Raleigh, North Carolina, and Casa Central in Chicago, Illinois.

When a prospective respondent called RTI in response to a recruitment ad, they were screened using the recruitment scripts found in Appendix E. Specifically, all respondents were screened for age, race, level of education, total household income, distance from the focus group site, geographic area (urban/suburban/rural), whether or not they spoke English/Spanish as their native (primary) language, whether they could read English/Spanish, and how well they could read English/Spanish (very well, well, or not well). Eligible respondents were informed that the focus group would be audio and video recorded and then provided with the date and time the group would be conducted. In order to recruit the desired number of participants and encourage a sufficient number of them to show up to each focus group, an incentive of $75 was offered.

4. Characteristics of Participants

RTI used the screener in Appendix E to recruit participants and place eligible participants in the appropriate session for their age. Recruiters were asked to strive for the greatest possible variation in age, gender, and other demographic characteristics, to ensure a heterogeneous set of viewpoints in each group. The goal was to recruit ten participants for each group to ensure at least seven or eight people would attend each focus group session.

The first round of recruitment began on September, 23, 2009 and concluded on October 16, 2009. This initial round recruited subjects for focus groups conducted on October 12-13 in Durham, North Carolina, October 20-21 in Raleigh, North Carolina and October 21 in Washington, District of Columbia. Round two of recruitment included focus groups held on November 2-3 in Irvine, California, November 4-5 in Addison, Texas, and November 4-5 in Chicago, Illinois. This recruitment phase ran from October 19, 2009 through October 30, 2009. Table 1 and Table 2 presents the final demographic composition of all participants across the seventeen focus groups, including age, gender, race, education, income, and urbanicity for each participant.

Table 1. Characteristics of English Focus Group Participants

Location and Age Group	Number	Gender		Race			Education		Income		Geographic Area
		M	F	White (Non-Hisp)	Black	Other Race	Edu ≤ HS/GED	Edu > HS/GED	Inc ≤ $100,000	Inc > $100,000	Urban	Suburban	Rural
Total	96	33	63	56	24	16	11	84	81	14	35	57	3
Raleigh, NC
18-29	8	3	5	5	3	-	1	7	8	-	3	4	1
30-49	8	3	5	3	4	1	1	7	6	2	3	3	2
50+	10	4	6	7	3	-	1	9	8	2	3	7	-
Washington, DC
18-29*	9*	4	5*	4	2*	3	1	7	6	2	3	5	-
50+	10	5	5	6	4	-	2	8	8	2	4	6	-
Addison, TX
18-29	7	2	5	2	2	3	1	6	6	1	2	5	-
30-49	10	2	8	6	4	-	1	9	9	1	4	6	-
50+	7	3	3	6	-	1	3	4	7	-	3	4	-
Irvine, CA
18-29	9	1	9	4	1	4	-	9	7	2	1	8	-
30-49	10	4	6	6	1	3	-	10	8	2	4	6	-
50+	8	2	6	7	-	1	-	8	8	-	5	3	-

* One no show was replaced by an eligible participant on-site. Since this participant was not completely screened prior to arrival, not all demographic characteristics could be recorded for inclusion in this table.

Location and Age Group	Number	Gender		Country of Origin				Education		Income		Geographic Area
		M	F	Mexico	Puerto Rico	Central America	South America	Edu ≤ HS/GED	Edu > HS/GED	Inc ≤ $100,000	Inc > $100,000	Urban	Suburban	Rural
Total	49	20	29	24	8	3	14	28	21	49	0	46	2	1
Raleigh, NC
18-29	5	2	3	2	1	-	2	4	1	5	-	5	-	-
30-49	8	3	5	3	-	1	4	5	3	8	-	8	-	-
50+	7	4	3	1	1	-	5	3	4	7	-	6	-	1
Chicago, IL
18-29	9	3	6	6	1	2	-	6	3	9	-	8	1	-
30-49	10	2	8	8	2	-	-	5	5	10	-	10	-	-
50+	10	6	4	4	3	-	3	5	5	10	-	9	1	-

Table 2. Characteristics of Spanish Focus Group Participants

3. Results

   1. Lead Letter Envelope

      1. Considerations in Deciding Whether to Open a Piece of Mail

Before presenting the lead letter envelopes, the moderators asked focus group participants to think about the mail they receive each day and identify factors related to their likelihood of opening a specific piece of mail. This discussion was intended to provide background information on how people generally perceive the different types of mailings they receive and how this might affect the likelihood of opening the NSDUH lead letter envelope. The moderators probed participants by asking the following questions:

• What kinds of things do you consider in deciding whether to open a piece of mail?

• What kinds of things do you look for on an envelope to determine whether you will open it?

• What kinds of envelopes do you tend to throw out or recycle without opening?

Most participants indicated that their households receive a great amount of mail and that they spend time sorting and separating the important pieces from what they considered to be “garbage” or “junk mail.” Some participants indicated that they open all the mail they receive, often looking for interesting promotions or coupons, but others indicated that they regularly throw out at least some mail unopened. In some of their households, participants noted that there is one individual who does most of the mail sorting. In these cases, the rest of the household members would only receive pieces of mail that the sorter considers to be important mail.

Focus group participants felt the decision to keep a piece of mail and open it is usually based on what they know about the sender and the characteristics of the envelope. Participants indicated they usually try to determine whether there is some kind of “connection” between themselves and the senders of the mail. They typically assess this by examining both the return address and to whom the mail is addressed. For example, they would be more likely to open a piece of mail from a company with whom they already have a commercial relationship. Participants also stated they would be less likely to open mail addressed simply to “Resident,” and some indicated they routinely throw out mail addressed in this way. Some participants in Spanish groups who live in apartment buildings mentioned that they never open mail addressed only to “Resident” because they think this mail was sent to the landlord. This can be a problem in cases where the mail of more than one family is received in the same mailbox and people do not want to open their neighbor’s mail by mistake.

For most people, simply adding some version of “Resident of ___________ County” to the address would not significantly diminish their inclination to ignore mail addressed in this way. Some participants in Spanish groups and some participants age 50+ in English groups thought that personalizing the mailing address in this way would make them more likely to open the mail. Those who preferred adding “Resident of ___________ County” mentioned their likelihood of opening the envelope would increase because addressing the envelope this way would indicate to them that the mail contains something important for them and their local community.

In terms of the physical characteristics of the envelopes, participants stated they usually associate pieces of mail that use bright colors and glossy paper, have messages written on them, and use bulk mail postage with marketing and promotions. For this reason, they are more likely to dispose of this mail unopened. Some participants mentioned there are some law firms and companies that try to make their promotions look more “official” or “important” in order to trick people into opening them. These participants noted that they identify such efforts with the use of plain white envelopes with minimal information on the outside about the sender.

Overall, focus group participants indicated the mail they are most likely to open would have the following characteristics:

• mail addressed to a specific person in the household,

• mail that includes the name of a company with which the recipient is familiar, and

• mail that uses mostly white envelopes with familiar and official-looking logos.

2. Reactions to the Two NSDUH Envelopes

Following the discussion of what factors influence the decision to open a piece of mail, the moderators distributed two versions of the lead letter envelope. The first version was a white standard number 10 size window envelope, printed as shown in Appendix F. The second envelope was a white 9 x 12 inch catalog envelope, with the return address and address window arranged in portrait orientation. The image for this larger envelope is exhibited in Appendix G. Both envelopes included the same agency logo and return address.

Initial Reactions and Likelihood of Opening the Two Versions of the Envelopes

Table 3 on page 18 presents focus group participants’ preferences with respect to each type of contact material. In the tabulation of participants preferences for the lead letter envelopes, slightly more than half of the English group participants preferred the larger envelope. Two-thirds of the Spanish group participants preferred the larger envelope.

Overall, initial reactions to the envelopes were very positive. Based on their appearance, most participants indicated they would likely open either the standard size or the larger size version of the envelope if they received it in their mail. A key factor stated almost universally by participants in the English language groups was that they would open the envelopes because of the Department of Health and Human Services (DHHS) return address. People emphasized that the envelope should look as “business-like” and “official” as possible, and this is generally accomplished by using the DHHS logo and title. For this reason, most participants stated they would open the envelope, regardless of the size and the use of the generic addressee of “Resident.” Because the mailing would be coming from the DHHS, many participants suggested that there could be important information concerning recent developments in health care. The Rockville, Maryland return address did not seem to bother participants in any of the cities. Inclusion of the RTI project number also did not seem to bother anyone, but it was mentioned by a few that this information did suggest that the mailing might be in reference to a survey.

In the Spanish groups, most participants were not familiar with the DHHS, but did indicate that they would still open the envelopes because they look “official” and different from marketing materials. The white envelope and the logo with an eagle in it clearly communicated to these participants that the mail came from the U.S. government. Similar to the English language groups, the Rockville, Maryland return address reinforced the idea that the mail is legitimate because people recognized that many government offices are located in that part of the country.

Specific Reactions to the Different Sizes of the Envelope

Reactions from participants suggested that the larger envelope would certainly garner more initial attention than the standard size envelope. Across the groups, people suggested that the larger envelopes are often used to send legal documents, so they look more “official.” Among the Spanish group participants the larger envelope communicated that the documents inside are important and cannot be folded. Some of the 18 to 29 year old participants in the Durham English group preferred the smaller envelope because they felt it looked “more business-like.” Also most participants in the 50+ Spanish group in Chicago felt there was no need to use a large envelope, perhaps because it would be a waste of resources.

In some of the English language groups, after discussing the letters the moderator returned to the question of which of the two envelopes should be used. Once participants realized that both envelopes would contain only a single sheet of paper, some reversed their earlier preference for the larger envelope suggesting that it would be a waste of paper and postage. Some of these people were motivated by a concern for the environment. Others were motivated by a concern with government spending.

Use of the “Official Business” Endorsement on the Envelopes

The text on the envelopes that reads “Official Business. Penalty for private use $300” tended to enforce the official nature of the mailing for focus group participants. Despite this view, the great majority of participants did not understand exactly what this statement actually means. In the Spanish groups, some people suggested that this statement means that the envelope must only be opened by the person to whom it is addressed and no one else. This perception could be problematic, given that the mail is addressed to “Resident” and in some cases the household receives mail in a depository shared with other households. There were also a few Spanish group members who felt it was intimidating to have the penalty note on the envelope.

2. Lead Letter Text

Three versions of the lead letter with text only and no graphics were presented to participants and labeled as versions 1a, 2a, and 3a. These versions varied in how or whether various aspects of the survey request and protocol were mentioned and the emphasis given to each element. For example, version 3a provided the study name in the first paragraph, while versions 1a and 2a did not. Versions 1a and 2a provided the URL for the NSDUH website. Version 3a of the lead letter text was based on a letter used by the National Survey of Family Growth and only included a signature from SAMHSA staff (with the RTI signature omitted). This version thereby attempted to develop a more “personal” approach and appeal to recipients. The three versions of the letters with text only are presented in numerical order as Appendices H, I, and J.

Initial Reactions to the Three Alternative Versions of the Letter Text

As Table 3 indicates, focus group participants varied somewhat in their preferences for each version of the lead letter text. Slightly more than half of the English group participants preferred version 1a, but Spanish group participant preferences were even more mixed. A plurality of about 43 percent of Spanish group participants preferred version 3a and another 35 percent preferred version 1a. Version 2a was the least popular among both English and Spanish group participants. Versions 1a and 3a were viewed by participants as being better organized, shorter, and more direct than version 2a. Participants also felt both of these versions contained most of the information participants wanted to know. In addition, the way the text was distributed on the page made versions 1a and 3a easier to read, understand, and communicate to other members of the household.

One of the main differences between version 1a and version 3a was that the latter mentions the specific topic of the survey (drug use, alcohol, and tobacco). Many participants felt that being specific about the topic would increase their interest in the survey and would make them more likely to participate. In addition, knowing the topic would also prevent surprises at the time of the interview when the questions start asking about sensitive subjects. Spanish group participants, who preferred version 3a, suggested that mentioning in the letter (as in the brochure) that both users and non-users of drugs and alcohol needed to participate would be a good idea. However, there were strong differences among the Spanish group participants in terms of reading skills and ability to understand the text of the letters. Only the more educated participants with a college education were able to talk about differences that were too subtle for the rest. This may be one reason why the two shorter versions of the letter were preferred.

Version 1a was particularly strong among the younger participants who believed that the text in the letter was shorter, more professional, and straight to the point. Version 3a was particularly popular among the older participants in the English language groups, and nearly half of the Spanish group participants. One of the reasons cited was the more personal tone of the letter. However, many of the younger and middle aged participants in the English language groups really disliked this tone particularly mentioning the introduction, “My agency…”

Other Specific Features of the Letter Text

One feature used only in version 2a viewed favorably by participants was how the “Members of the Household” and “Resident of the [city, county, or state]” made the letter more seem more personal. These participants were particularly thinking about situations when more than one person in the household would read the letter.

Participants were asked to compare how the three letters handled the issue of survey confidentiality. A large number of the participants in the English language groups preferred the single sentence used in version 3a as it was short and to the point. Some participants pointed out that the mention of the random selection of the address, not the particular person (version 2a) further enhanced confidentiality. Furthermore, they liked the fact that the sentence was its own paragraph that made the topic “stand out.” There was no discussion about the small font text at the bottom of the letter that further addressed confidentiality.

Participants from the Spanish groups did not notice big differences across letters in terms of information about confidentiality. To them, the statements in the three versions were sufficient to make the reader comfortable about participating in the survey. Stating that the confidentiality of the responses will be protected by federal law was one of the key elements for the Spanish group participants. Only a few noticed the small font text at the bottom of the page addressing confidentiality. Once it was pointed out, however, most agreed that it was a positive element and should be included in the letter.

There were mixed responses to the handling of signatures. Some suggested that two signatures made the letter appear more official and important; others said they were not very interested in how the government was conducting the survey and a single signature would suffice. However, if a single signature would be used, it should be the one from DHHS rather than RTI.

There were mixed reactions to the concept in version 2a of using other organizations to endorse the survey. The 30 to 49 year old English participants in Durham and Dallas said an endorsement by the American Medical Association (AMA) might help give some credibility to the survey but an endorsement by the American Association of Retired Persons (AARP) would not carry much weight. Adults in the 18 to 29 year old group in Durham did not seem to think endorsements would add much. The youngest group of adults in D.C. suggested that maybe an endorsement by the Centers for Disease Control (CDC) might enhance the appeal. In California there was little support for the concept of using endorsements. There, it was emphasized that the merits of the study itself should be enough to convince people to participate. In Dallas it was suggested that the use of endorsements from other organizations carried some risk since some organizations could be perceived as detractors and this might cause less favorable disposition towards survey participation. Some of the frequently mentioned organizations were the American Health Association, the Red Cross, and the March of Dimes.

Organization like AMA and AARP had a low level of awareness among the Spanish group participants, and their endorsement would not have a strong effect on people’s willingness to participate in the survey. At the same time, it was suggested that mentioning the endorsement of a health institution would be more beneficial than endorsements from other types of institutions. Further, mentioning that the survey was required by the U.S. Congress reinforced that the survey was official, serious, and relevant, and some of the Spanish group participants said it would make them feel more obligated to participate. Most of the participants said that mentioning of the specific government code enabling the survey in version 2a was not needed, and if anything, might be intimidating.

Nearly everyone in all the English language groups said they would use the internet to access the RTI website indicated in the first two versions of the letters. Few said they would call the toll-free number, but only after having visited the website first, and only if they had additional questions about the survey. The lack of a web site address in version 3a was often cited by participants as a reason why they did not choose that version of the letter.

Spanish group participants also believed that having a web site address was a very positive element in the text. Most said that they would immediately go to the web to find more information and make sure that the survey was legitimate. However, for Spanish group participants, having a telephone number was also important because there were many participants who did not have access to the internet, or they did not feel comfortable using computers.

In general, participants were not familiar with RTI. The three versions of the letter did a good job explaining the role of RTI – most participants understood that it was the institution that would conduct the interviews, and they thought it was good to mention that RTI is a non-profit organization. However, there was little interest in including more details about RTI.

The bolding and highlighting of the incentive sentence caught the attention of the readers. It did not seem to have a negative connotation and participants admitted that this line would increase their interest in the letter, and because of it, they would be more likely to pay attention and read the complete text more carefully. The fact that this information was not bolded in Version 3 was often cited as a reason for not choosing that version of the letter.

After reading the texts, most participants had a good understanding of the ideas communicated in the letters. The language was appropriate and most of the information people wanted to know was included. However, some participants believed the letter should include more specific information about the interviewer’s visit – specifically, when the interviewer will call on their house, and who in their household they would want to talk to. Spanish group participants from the two younger groups (18-29 and 30-49 years old) in Chicago expressed interest in being able to determine whether they would be eligible to participate in the survey by including eligibility age range in the letter.

3. Lead Letter Graphics

In addition to text, the graphics for the lead letters also included three alternative versions, labeled 1b, 2b, and 3b. These versions varied in the size, format, and content of the graphics used to “package” the letters. For example, the image on the example identification badge varied from a dark gray silhouette (versions 1b and 3b) to an actual photo of a person, printed in color (version 2b). Participants were asked to comment on the graphical layout of the letter independent of its content (this was accomplished by providing letters with graphics, but no text). The three versions of the letter graphics are exhibited in numerical order as Appendices K, L, and M.

Initial Reactions to the Three Alternative Versions of the Letter Graphics

As Table 3 shows, focus group participants indicated mixed preferences for which version of the lead letter graphics they preferred. None of the versions garnered majority approval in either the English or Spanish groups, but half of participants in the Spanish groups preferred version 2b. In contrast, about 45 percent of English group participants preferred version 1b.

Participants immediately noticed the differences in the headers and sizes of the logos. They believed the logo was important because it reinforced the official nature of the letter. Adding Rockville, Maryland and a zip code in the header was also a positive element and helped to make the letters appear more “business-like” and official. Some participants liked the larger DHHS logo in version 1b as it allowed them to read it and that was why they preferred that version over the others.

Other Specific Features of the Letter Graphics

Across the groups there was a difference in opinion regarding the line in the header of version 1a, “An Important Request from The U.S. Department of Health & Human Services.” Some participants liked how it emphasized the importance of the survey while others suggested that it made the letter look “less business-like.” In the Irvine groups it was suggested that the importance of the survey should be obvious without having to explicitly state it.

Overall, greater support for the use of this tagline in version 1a was evidenced in the 50+ age groups than in the younger-aged groups. Among the 50+ year olds in the Addison groups the appeal of this tag line was consistent with their preference for the text in letter version 3a, which begins with, “My agency, …, needs your help.” In Irvine, however, the older adults did not like the header on version 1b saying that it looked less professional than the other versions. They also suggested that it was best if the recruitment materials did not make DHHS sound “too desperate.” Some Addison group participants similarly indicated that the materials should not appear to be using “hard sell” approaches.

Among the Spanish group participants, the tagline did not have any negative effect, but it was not considered a significant positive element. However, it is important to mention that a couple of younger participants (who could also speak some English) thought that the word “solicitud” was usually associated with an application they had to fill out and suggested using the word “petición” instead.

There was a consistent response to the image of the identification badge using an actual photograph, as in version 2b. Nearly all participants said their first impression was that this was the individual who would “show up” at their home. When informed otherwise, nearly everyone said if the badge had some indication that it was being used for illustrative purposes only, like having “sample” written across it, they would still prefer to see an actual face on the badge. Some respondents also thought it was too much to show Ilona Johnson’s signature on the identification badge and that her signature on the letters (in versions 1b and 2b) was sufficient. Participant feedback indicated that the way the identification badge is identified must be clear and somehow visual for people who do not read English. There were a couple of instances where Spanish group respondents thought that the person in the photo was Ilona Johnson because they read the name in text on the badge, but they did not understand the words around it.

After the discussion about the badge, most participants agreed that having the hand written name of the interviewer that would visit their home was a good idea. However, some Spanish group participants were skeptical about the ability of RTI to actually send the person who signed the letter.

4. Question and Answer Brochure

Focus group participants were shown two versions of the question and answer (Q&A) brochure. Version 1 was the current Q&A brochure, updated to reflect planned for the NSDUH redesign. Version 2 was an updated version which used a variety of background colors and photographs. The brochures included some similar questions and answers, but also had significant differences in both content and format. The two versions of the Q&A brochures are displayed in numerical order as Appendices N and O.

Initial Reactions to the Two Alternative Versions of the Q&A Brochure

Overall, version 2 was preferred by participants over version 1. These preferences differed somewhat between the English and Spanish group participants. Over 80 percent of participants in the Spanish groups preferred version 2 and 51 percent of English group participants preferred version 2. Participants in the English groups (24 percent) were also much more likely to decline to indicate a preference between the two versions compared to those in the Spanish groups (0 percent).

Most participants indicated that version 2 was more appealing and something that most people would find more inviting to read. Older participants remarked that a slightly larger type font made this version easier to read. Participants generally felt that the photos included in version 2 of the brochure do a good job communicating that different types of people of different ages, occupations, and walks of life are participating in the survey. This feature led people to note that version 2 appeared more friendly and personal. Participants also liked the colors used for this brochure.

The only image that people had some trouble with was the map. Participants weren’t sure why this image was included and what the different shades of blue and colors signified. Some people believed that the map was communicating the fact that NSDUH is a national survey, but no one associated this image with the selection of sample units.

Version 1 of the brochure was preferred by a minority of the participants. A negative comment among 18 to 29 year olds suggested that this version looked like “something from the 80s.” A common complaint among all participants was that version 1 contained too much information. People suggested this made the brochure somewhat overwhelming and, therefore, less compelling for them to read. At the same time, a minority of participants in each group indicated they liked the greater content in version 1. Some of these participants suggested that this version of the brochure might be more effective for describing the survey to another household member. These people suggested that dividing the information into more specific topics in version 1 was a better approach than the way the topics were organized than in version 2.

One section in version 1 of the brochure that was identified as being particularly effective by some participants was the section with the title “What If I Do Not Smoke, Drink, or Use Illegal Drugs?” One section that was considered unimportant by most people was “How Does the Government Conduct the Study?” Words that summarize a number of participants’ reactions that SAMHSA uses a competitive bid process to select a vendor included “I don’t care.” In each of the Irvine groups it was pointed out that the way in which version 1 addressed the question of how the survey data would be used is very good. It was emphasized, particularly in the 50+ age group, that it is important to tell people how this information will be used.

Overall, participants felt that the information in the brochures is more understandable and complete, especially compared to the more limited information presented in the letter. The technical and administrative details of the survey seemed less relevant to participants than knowing the purpose and utility of the survey. The brochures also do a much better job explaining that not only drug and alcohol users need to participate in the survey. Some participants said this will make people more willing to participate, although they also acknowledged that this might discourage participation among those who have considerable substance use to report.

Other Specific Features of the Q&A Brochures

When asked if any of the information in either version of the brochure was “confusing,” participants in the English groups generally indicated they did not have difficulty with most of the phrases used to describe the survey process, such as “randomly selected” and “chosen at random through scientific methods.” One concept that some people did have difficulty with was the juxtaposition of “random” with “scientific” in the phrase “…chosen at random through scientific methods.” It was suggested by some participants that “random” and “scientific” is a contradiction. In the Spanish groups, explanations of how participants are selected randomly and scientifically were not understood by participants with lower education levels. At the same time, these participants did not express much concern about why they would be asked to participate.

Although much the same information is contained in both versions of the brochures, participants seemed to have an easier time identifying the listing of relevant web sites in version 1. Since the web site is likely to be used by people to validate information about NSDUH, use of the format for the web addresses in version 1 of the brochure should be considered.

The mention of computers in the brochure was only mentioned as a concern in the Durham group of 18 to 29 year olds. This concern was not expressed for them personally, but instead they suggested some older adults might be concerned about having to use a computer and therefore be less interested in participating. None of the older adults in the English or Spanish language groups expressed any such concern.

In the Spanish groups some participants mentioned that not using a computer would make them doubt the legitimacy of the survey because today everybody uses computers for everything. On the other hand, older participants in the Spanish groups who did not have much experience working with computers appreciated the information included in the brochures explaining that knowledge of computers is not necessary. These participants did not quite understand whether they would actually have to use the computer themselves or whether the interviewer would enter their responses for them.

Overall, participants felt the brochures do a better job than the letters in explaining the role of RTI in the study. People noted that the brochures provide more detailed information and history about both the NSDUH and the institutions involved. Some participants suggested that this information would cause them to be more favorably disposed towards participating in the survey.

Participants generally thought it is important to include the logos of SAMHSA and RTI in the way they are presented in version 2 of the brochure. Even though sample members may or may not be aware of these institutions, people indicated the logos would make them more comfortable by reinforcing the importance and legitimacy of the survey.

The 50+ participants in the English groups seemed sincerely motivated to participate in such a survey if it helped the government’s health planning and related public policy initiatives. Messages focused in terms of how survey participation would support a worthy endeavor seem likely to be received favorably by older adults.

5. Special Concerns of Spanish-speaking Participants

Overall, Spanish-speaking participants shared many of the same impressions of the materials as the English group. There are only some special concerns with the Spanish version of the materials. The most important concern is that the reading level of the letters is only appropriate for Spanish-speakers with at least a high school education or greater. During the focus groups strong differences across participants were noticeable in terms of reading skills and ability to understand the letters text. There was one participant in Chicago who said she did not understand any of the letters, and others just repeated part of other people's comments. Only the more educated participants with a college education were able to talk about differences that seemed too subtle for the rest.

Like English group participants, the text on the envelopes that reads “OFFICIAL BUSINESS. PENALTY FOR PRIVATE USE $300” was frequently misunderstood by most Spanish speakers. In addition, many Spanish group participants indicated that they also felt uncomfortable opening an envelope with this note. Some felt intimidated by it. Participants felt it meant that the envelope must only be opened by the person that it is addressed to, and no one else. This was considered to be confusing, particularly if the mail is addressed to “Resident” and the household receives its mail in a depository with other households. Many participants thought they would have to pay a fine of $300 if they opened the envelope.

The use of acronyms in Spanish is not as common as in English, and it has always shown to be problematic for Spanish group participants who are unfamiliar with U.S. government agencies, associations or U.S. code. Many Spanish-speaking participants pointed out their unease with acronyms, and recommended avoiding them all together.

Explanations of how participants are selected randomly and scientifically were not really understood by those participants with lower education levels. And the Spanish term ‘al azar’ (at random) was confusing and many did not understand what that meant.

Spanish group participants said their first impression of the picture on the identification badge was that it was either the individual who would “show up” at their home or the project director. When informed otherwise, some participants said they would feel distrustful if the field interviewer who showed up at their door was not the same person shown in the picture.

Letters addressed to “Resident” made people in the Spanish groups think that it was not necessarily sent to them. Many thought the letter was addressed to landlords, such as in cases where people live in rented homes.

6. Additional Suggestions

In addition to the discussion of the features of the envelopes, letters, and brochures already detailed, several further suggestions were identified by focus groups participants:

• Participants suggested the use of a regular stamp to make the envelope look more official and distinguish it from junk mail. This recommendation is in sync with the Tailored Design Method for survey mailings, proposed by Dillman (Dillman, 2000).

• Many participants suggested putting the incentive amount on the envelope or somehow suggesting that the recipients can get paid for participation in an official government survey. Such a message would make sample persons more likely to open the envelope and read the letter.

• Generally, participants did not like the fact that the envelope might be addressed to “Resident”. They preferred “Resident of ____________ County” as this placed them in a group and suggested the letter contained something that applied to them. Several suggestions were made to better address this issue. Among the proposed addressees were “Head of household at [ADDRESS]”, “Randomly selected resident at [ADDRESS]”, and “Survey to resident at [ADDRESS]”.

• In most groups there was some misunderstanding about the statements in letter versions 2a and 3a regarding the interviewer visit. On first read, many participants believed the letters stated they would be given a $40 incentive for answering a few questions (in version 2a) or a five-minute interview (in version 3a). These statements caused some confusion about the survey protocol. During the discussions, participants eventually realized the letters were referring to the screening process that would determine their eligibility to participate in the survey interview.

• Several participants suggested mailing the brochure with the initial letter to make the mailing look more official overall.

Table 3. Preference Counts for Each Type of Contact Material

	English Groups				English Totals		Spanish Groups		Spanish Totals
	Durham, NC	Washington, DC	Irvine, CA	Addison, TX	Count	%	Raleigh, NC	Chicago, IL	Count	%
Lead letter envelopes
Lead letter envelope, regular size	9	9	8	13	39	40.6%	4	6	10	20.4%
Lead letter envelope, 9x12	17	10	14	10	51	53.1%	15	18	33	67.3%
No preference expressed	0	0	5	1	6	6.3%	3	3	6	12.2%
TOTAL	26	19	27	24	96	100.0%	22	27	49	100.0%
Lead letter text only
Lead letter, Version 1a	9	12	18	11	50	52.1%	5	12	17	34.7%
Lead letter, Version 2a	7	3	1	3	14	14.6%	7	4	11	22.4%
Lead letter, Version 3a	8	3	1	8	20	20.8%	8	13	21	42.9%
No preference expressed	2	1	7	2	12	12.5%	0	0	0	0.0%
TOTAL	26	19	27	24	96	100.0%	20	29	49	100.0%
Lead letter graphics only
Lead letters, Version 1b	12	2	17	13	44	45.8%	10	5*	15	30.0%
Lead letters, Version 2b	8	8	7	6	29	30.2%	9	16	25	50.0%
Lead letters, Version 3b	4	8	2	4	18	18.8%	1	9*	10	20.0%
No preference expressed	2	1	1	1	5	5.2%	0	0	0	0.0%
TOTAL	26	19	27	24	96	100.0%	20	30	50*	100.0%
Q& A brochure
Q & A brochure, Version 1 (current)	6	3	5	10	24	25.0%	1	7	8	16.3%
Q & A brochure, Version 2 (new)	10	16	17	6	49	51.0%	19	22	41	83.7%
No preference expressed	10	0	5	8	23	24.0%	0	0	0	0.0%
TOTAL	26	19	27	24	96	100.0%	20	29	49	100.0%

* One Spanish group participant selected both versions 1b and 3b

This page intentionally left blank.

4. Summary and Recommendations

The 17 focus groups conducted in various regions of the U.S. confirmed the importance of much of the content and formatting elements used in the contact materials, but also highlighted some key issues that might not be fully addressed in each type of contact materials. This section synthesizes the focus group reactions into specific recommendations for improving each type of contact materials. Recommendations accepted by SAMHSA will then be used in creating a new version of each type of materials for the NSDUH redesign scheduled for 2013.

1. Lead Letter Envelopes

A majority of focus group participants stated a preference for the larger 9x12 envelope. This majority was greater in the Spanish focus groups. At the same time, participants in the English groups did express concern about using large envelope for only a letter and no additional study materials. In addition, the overwhelming majority of participants indicated that they would open either envelope, primarily because the DHHS logo led them to believe that the mailing was important. The main benefit of the larger envelope appears to be that it will attract greater attention, at least in some households. One benefit for NSDUH interviewers is that they would not have to fold the letters to insert them into the envelopes, which would increase the likelihood that the address information is visible in the address window.

Based on these results, the recommendation would be to carefully consider the costs and benefits of using 9x12 envelopes versus the standard number 10 size. Using these larger envelopes would require additional material costs of about $4,800 and additional postage costs of approximately $116,000 annually, assuming 200,000 lead letters are mailed each year. Although focus group participants generally preferred the larger envelope, this preference was tempered by knowing that only a single letter would be included in the mailing. Given that the standard size envelope did not generate negative reactions, this size envelope with logo and return address presented in the focus groups may be equally effective as the larger envelope.

Including the endorsement “Official Business. Penalty for private use $300” appears to have advantages and drawbacks. Focus group participants felt this statement made the envelopes look more official, but almost no participants actually understood what this statement means. The phrase “Penalty for private use $300” was particularly confusing to many participants. These reactions suggest that it may be advisable to keep the phrase “Official Business,” but drop the second part of the statement if possible. United States Postal Service guidelines indicate that this full statement would have to be included on the envelope. The current envelope used for mailing the lead letters and frequently asked questions for the National Immunization Survey, a study sponsored by DHHS, includes only the “Official Business” part of this statement. The potential for using only the first part of this statement should be investigated further, so that a final decision can be made on whether to include this statement. Given that most participants indicated they would open the envelope because of the DHHS logo, including this statement does not seem critical to the effectiveness of the lead letter envelope.

2. Lead Letter Text

A majority of participants in the English groups preferred version 1a of the lead letter text, but a plurality of Spanish group participants preferred version 3a. In most groups, participants made compelling arguments for either version. Version 2a was generally viewed as too lengthy and complicated, and participants therefore felt they could recommend few parts of the text in this letter as preferable.

Based on these results, the primary recommendation would be to create a hybrid of the text in version 1a and 3a that would combine the preferred text of each letter and avoid any text considered to be problematic. For example, many participants indicated version 1a was well-organized and covered the most important information about the study. Participants also thought it was useful that version 3a specifically mentions the topic of the survey and indicates (as in the Q&A brochure) both users and non-users of drugs and other substances are needed to participate. These reactions can be used to update version 1a to include some of the content and phrasing of version 3a to produce a stronger letter overall.

One element of the version 2a letter text that could be used in the new letter was addressing the letter to “Resident of ____________ County.” Some participants did not feel this would significantly increase their likelihood of opening the letter, but many did feel this would indicate to them that the mailing is important for them and their local community. For this reason, it might be worthwhile to investigate the costs and logistics of adding the county, parish, or district for each addressee.

On the issue of two signatures (included in versions 1a and 2a) versus a single signature (used in version 3a), participants did not indicate a strong preference. Most felt including both signatures was the better approach, so recipients would more clearly understand both SAMHSA’s and RTI’s role in conducting the study. Given that there were really no negative reactions to including both signatures, it seems like the letter should continue to provide both the SAMHSA and RTI signatures.

3. Lead Letter Graphics

Focus group participants offered rather mixed preferences for which version of the lead letter graphics they preferred. None of the versions garnered majority approval in either the English or Spanish groups, but half of the participants in the Spanish groups preferred version 2b. In contrast, about 45 percent of English group participants preferred version 1b.

Specific elements of the lead letter graphics seemed to heavily influence participant preferences. The larger DHHS logo on version 1b was often cited as preferable to the smaller version displayed on versions 2b and 3b. Another key element was the use of a gray silhouette versus an actual picture in the image of the field interviewer’s identification badge. Overall, participants preferred the actual picture on the identification badge, even when it was pointed out to them that the picture could not be tailored to show the actual field interviewer assigned to each selected household. One qualification on this point is that participants agreed that a watermark or other graphical feature should be used to indicate the identification badge is only a sample. Many participants were initially unclear that the picture was just a sample, and therefore they would have expected the person in the picture to be the actual field interviewer assigned the recipients’ household. One aspect of the graphics in version 1b that was not received favorably by participants was including the phrase “An Important Request from The U.S. Department of Health & Human Services” in the header. Most participants viewed this as superfluous.

These reactions suggest that the lead letter graphics should incorporate various elements used across the three versions, including:

• the larger DHHS logo used in version 1b

• the sample picture on the identification badge used in version 2b, with a watermark indicating that the badge is just a sample

• the line for the interviewers’ name under the identification badge used in all three versions.

Participants did not express strong preferences for how the return address was presented in the header. This feature of the header seems unlikely to have a significant impact on recipients’ reaction to the letter.

4. Q&A Brochure

Although a majority of focus group participants preferred version 2 of the Q&A brochure, preferences did differ significantly between the English and Spanish group participants. Whereas over 80 percent of participants in the Spanish groups preferred version 2, 51 percent of English group participants preferred version 2. Participants in the English groups were also much more likely to decline to indicate a preference between the two versions compared to those in the Spanish groups.

The primary appeals of version 2 of the brochure appeared to be the use of colors and pictures, as well as the layout and amount of text presented. Participants who perceived version 1 as providing more detail overall than version 2 viewed this alternatively as either a positive or negative feature. Some felt the additional details were informative and useful, but others thought these details were overwhelming and would discourage people from reading the brochure.

Feedback on the Q&A brochures indicates development of version 2 should continue, but useful elements from version 1 should be incorporated into the brochure. For example, some participants suggested the way the topics were organized in version 1 was a better than in version 2. In addition, the brochure should retain sections viewed as particularly useful by participants and consider reducing or dropping sections viewed as less important. Participants felt the section “What If I Do Not Smoke, Drink, or Use Illegal Drugs?” in version 1 of the brochure was particularly effective. One section that was identified as less important by most people was “How Does the Government Conduct the Study?” Creating a new brochure should address these concerns.

Preliminary cost estimates gathered during the design phase for the contact materials indicated that version 2 of the Q&A brochure should not cost significantly more than version 1 to print. A final cost estimate can be determined once the brochure design is finalized.

This page intentionally left blank.

5. References

Dillman, Don. (2000). Mail and Internet Surveys: The Tailored Design Method. NY: John Wiley and Sons

Fowler, F.J. (1995). Improving Survey Questions: Design and Evaluation. Sage: Thousand Oaks, CA.

Groves, R. M., Cialdini, R. B., and Couper, M. P. (1992). "Understanding the Decision to Participate in a Survey." Public Opinion Quarterly 56(4): 475-495.

Murphy, J., Eyerman, J., and Kennet, J. (2004). Nonresponse among persons age 50 and older in the National Survey on Drug Use and Health. Proceedings of the eighth conference on Health Survey Research Methods, Atlanta, GA.

Murphy, J., Schwerin, M., Hewitt, D., and Safir, A. (2005) “Nonresponse among Respondents Aged 50 and Older Potential Respondents Focus Group Report.” Prepared by RTI for Substance Abuse and Mental Health Services Administration.

This page intentionally left blank.

Appendices

1 The NSDUH was titled the “National Household Survey on Drug Abuse” (NHSDA) prior to 2002.

2 For the focus groups, the return address or other text on the lead letter envelopes was not translated into Spanish. NSDUH lead letters that include the Spanish translation of the letter on the reverse side are currently mailed in the same envelopes as English-only letters.