Testing Report

Date:	06 June 2007
To:	Diane Herz
From:	Jim Esposito
Subject:	Results of Testing: Suggested Modifications to CPS Volunteer Supplement

The modifications to the volunteer supplement recommended in this memorandum are based on the following evaluation methods: behavior coding, interviewer debriefings and cognitive interviews. For the most part, these are qualitative evaluation methods that do not possess the attributes that would be needed to make strong assertions regarding the magnitude of measurement error; in other words, while these methods are useful in indicating where measurement error is likely (e.g., due to differences between what the sponsor intends and what respondents understand regarding particular survey questions and concepts), it is not possible to establish—in quantitative terms—the accuracy of survey estimates generated by individual supplement questions or sets of questions. That said, and given that there is a time series dating back to 2002, I am hesitant to propose major changes to the supplement at this time, either to the content and/or structure of the questionnaire or to the wording of specific supplement items. And while changes to a supplement’s instructional materials can also have an impact on the integrity of a time series, I have offered some suggestions for enhancing the interviewer manual that I hope will not negatively affect the time series.

A. Supplement Transition Statement: PRESUP [see Attachment A for wording]

A1. Observations: This transition statement, which is important for orienting the respondent to the concept of formal volunteering and to the response task, is only being read one time—when the volunteer supplement is initially administered to the respondent who completed the CPS. This information is not read to other supplement-eligible members of the household when they respond for themselves during the initial contact, nor is it read at later points in time when call-backs are conducted to gather data for household members who were not available when the initial CPS call was made.

At least two bilingual interviewers seemed to think that Spanish-speaking respondents have difficulty understanding various aspects of this statement (e.g., “volunteer activities”; “through or for an organization”).

A2. Recommendation: Supplement programming needs to be modified to allow this statement to be read to each eligible household respondent when they provide data for themselves. In those situations where one household respondent provides proxy data for other household members during a call-back, this statement should be read in its entirety at the beginning of the first proxy interview. If the respondent serves as proxy for other household members immediately thereafter, it probably would not be necessary for interviewers to read this statement again, though they should be prepared to remind the respondent of what formal volunteering entails if the respondent appears to have forgotten.

A3. Suggestions. Given the critical importance of this statement in orienting respondents to the focus of the volunteer supplement (i.e., formal volunteering), I would suggest that the design team consider revising the wording of PRESUP to make its content more informative and relevant to the task at hand. For example: “This month, we asking a few questions about volunteer activities—that is, activities for which a person is not paid. We are interested in volunteer activities that have been sponsored by an organization, or in volunteer activities that you have done for any organization that provides benefits or services to others.”

B. Primary Classification Items: S1, S2 (and S2A)

B1. Observations: Based on responses made during cognitive interviews, it appears that laypersons have a complex and varied understanding of what it means to do volunteer work, and that clarifying what the sponsor means by that concept (e.g., via PRESUP and specific question wording) is essential to minimize measurement error. Our nineteen cognitive interviewees also differ—sometimes dramatically—in their understanding of the phrase “through or for an organization”, sometimes being completely unable to articulate the difference between through an organization and for an organization. The issue under consideration here is not whether or not there is measurement error associated with supplement estimates, but rather what the magnitude of that error might be and what the sources of that error might be. These issues are not specific to the volunteer supplement; they are issues of critical importance for all sample surveys. Item S1 clearly doesn’t capture all the formal volunteering work that was performed by eligible household members during the reference period. Additional volunteers captured by items S2 and S2A substantiate that claim, and split-panel research conducted by Toppe (2005) suggests that additional volunteers might have been identified had this line of questioning been extended, though there are clearly diminishing returns and potential disadvantages to doing so (e.g., increased burden on interviewers and respondents; supplement break-offs). Considered individually, evaluation data suggest that none of these three items is perfect; considered as a set, they appear to do reasonably well in capturing individuals who do formal volunteering work. With regard to S1, interviewers mentioned several cases that were ambiguous as to volunteer status. For example, one respondent said she volunteers about 20 hour per week doing unpaid work for the school at which she teaches full time. Another respondent reported working as an unpaid intern for some unspecified organization; and a third respondent reported being trained for a volunteer position that never actually materialized. Do these examples constitute volunteering?

B2. Recommendation: I understand that the sponsor would prefer not to ask S2A in September 2008; however, the exclusion of this item from the supplement and the classification algorithm would represent a missed opportunity to reduce measurement error by a small amount, relative to the current estimate of volunteers (i.e., absent S2A, the supplement’s classification algorithm would yield a small undercount of formal volunteers). In the absence of compromising data, qualitative or quantitative—or some other persuasive rationale—it is recommended that item S2A be included as a classification item in the volunteer supplement and that it be included in the algorithm for estimating the number of persons performing volunteer work during the reference period.

B3. Suggestions: Since it has been noted in the supplement instructional manual that the phrase “through or for an organization” can sometimes be misheard by respondents (i.e., “through a foreign organization”)—a result perhaps of the rapid pace at which telephone interviews are conducted—one simple suggestion for eliminating this problem would be to reorder the prepositions to read “for or through an organization”. Also, at some point in the future, the supplement design team may wish to consider dropping the preposition “through” altogether—for two reasons: First, this term has already been discarded from item S5 and from all subsequent questions for which it might apply; and second, the preposition “for”, alone, appears broad enough in connotation to carry the full meaning of the longer phrase. For example, Statistics Canada in their 2003 General Social Survey on Social Engagement asks about formal volunteering as follows: “In the past 12 months, did you do unpaid volunteer work for any organization?”

B4. Other Issues/Comments: CPS respondents are unpaid for their time, complete the CPS survey and its various supplements repeatedly (i.e., four times a year in most cases), and provide a benefit to at least two Federal agencies/organizations (i.e., the BLS and the Census Bureau)—not the mention the Congress of the United States: It there a reason why the time and effort they have expended on our behalf should not count as a formal volunteering activity?

C. Secondary Classification Items: S3, S4A, and S4A1

C1. Observations: I refer to these three items as secondary classification questions because if respondents cannot name the organization for which they volunteered or describe what type of organization it is, in theory, they will not be classified and counted as a volunteer. In reality, it appears that a codable answer (i.e., any substantive response other than refused or don’t know) to S3 and to either S4A or S4A1 will yield a volunteer classification. If true, this would mean that volunteering through or for an organization is respondent-determined: In other words, if a respondent can provide an organization name or an organization type, that person would qualify as a volunteer—regardless of whether the sponsor would agree (if all details were known). Given that over three-fifths of respondents mention only one organization, there is the potential for some false positives here—although the definition of “organization” is so liberal (e.g., “… an association, society or group of people who share a common interest”), it would be very hard to disqualify any response as inconsistent with the definition. Another issue of potential concern here is the absence of specific guidance to respondents and interviewers as to how the respondent should answer S3 (and S4 or S4A1) if, in performing volunteer activities, she worked through an organization (e.g., she participated in a church-sponsored clothing drive to benefit a local homeless shelter by sorting and packaging clothes at the church) and for an organization (e.g., she later delivered clothes to the homeless shelter and helped to distribute packages to residents). How should the respondent answer S3 (and S4A or S4A1)? For example, should the answer to S3 be “one” (either the church or the homeless shelter) or “two” (the church and the shelter)?

C2. Recommendations: Seek clarification from the sponsor as to how these issues should be addressed and include relevant information in the volunteer-supplement interviewer manual.

C4. Other Issues/Comments: Several interviewers mentioned that respondents would sometimes recall that they volunteered for an organization well after moving past this series of questions and that they would experience difficulty either in skipping back (via F1) to change the answer to S3 or experience doubts as to whether they should do so and risk messing up the data file. Is skipping back to change the answer to S3 a documented problem and, if so, what should interviewers do in such a situation?

D. Identification of the Principal Volunteer Organization: S5A, S6A and S7A

D1. Observations: Interviewers report (and my observations in Tucson and Hagerstown confirm) that this is a very difficult series of questions about which to gather data. With regard to S5A, some respondents appear to presume that this question is requesting that a respondent total up how many days or hours they have volunteered in the past twelve months and convert that total to weeks (e.g., “Probably one and a half weeks, ten days.”). If a respondent does not volunteer at a fairly regular rate during the weeks she does volunteer, S7A requests the she compute a total for the entire twelve months. This is an impossible task for persons who volunteer frequently but unevenly throughout the year.

D2. Recommendations: Several interviewers noted that the intent of S5A would be much easier to comprehend if one word was added to this question: “During how many different weeks in the last year did you do volunteer activities for the New World elementary school?” This seems like a reasonable suggestion, and one that could be made with little difficulty. Also, because the term “last year” is ambiguous (e.g., does “last year” mean from January onward or from last September onward), I would recommend that the words “past twelve months” be used instead. To be consistent, and after checking for possible awkwardness in wording, I would suggest making this change wherever the words “last year” appear in the supplement.

D3. Suggestions: Given the dubious quality of hours data being collected from items S6A and S7A, the sponsor may wish to test a different periodicity (or a whole new approach) for gathering data on hours should a redesign of this series be undertaken in the future. For example, the first question in the series might ask: “During how many different months in the past twelve months did you volunteer for the New World elementary school?” Then: “In those months when you volunteered, about how many hours per month did you volunteer, on average, for the New World elementary school?

E. Volunteer Activities: S8 and S8A

E1. Observations: In contrast to English-only interviewers, Spanish-speaking interviewers identified S8 as one of the most problematic items on the supplement. The key to understanding this difference is this: Most bilingual interviewers, when conducting interviews with Spanish-only respondents, typically refer to a paper version of the supplement in which the translation from English to Spanish is something less than optimal. For example, one interviewer mentioned that for option 4, the word “usher” appears as “portero” (phonetic spelling), which actually refers to a “goalie” in soccer; this interviewer claims that the appropriate Spanish equivalent is “comadore” (again, phonetic spelling”). Apparently, there are other words in S8 (and elsewhere) that are either not translated well from English to Spanish or for which a good/direct word-for-word translation is not possible.

On the positive side, most interviews were quite successful at reading this long-winded question (13 precodes) in its entirety. Respondents, for the most part, were patient and attentive, with a good proportion selecting multiple sets of volunteer activities. For those respondents who did mention multiple activities, very few appeared to have difficulty identifying the activity performed most frequently.

E2. Recommendations: In an effort to standardize the manner in which interviews are conducted in Spanish, it is recommended that the Census Bureau solicit the views of bilingual interviewers at all three telephone centers for suggestions as to how the translation of S8 (and other supplement items) might be improved. Where there is a reasonable disagreement as to specific word-to-word translations, interviewers should be provided with several alternatives from which to choose.

F. Foreign Volunteering: S15, S15.5 and S16

F1. Observations: Based on observations made during supplement administration (September 2006), no manifest problems with S15 were detected; and very few respondents answered “yes” to this question. As a consequence of the latter, S16 was not asked frequently enough to gather reliable observational evidence. Items S15 and S15.5 were asked during the course of our cognitive interviews, both during a simulated CPS-supplement interview and also during a second simulated interview with the research participant serving as proxy for a hypothetical household member (“Joe”). In the latter simulation, a pre-scripted scenario was used to evaluate both items (see Attachment B). For the most part, S15 appeared to work as intended; but we did observe two problems during the proxy simulation for S15.5. First, in the process of generating a response to S15.5, some research participants included informal volunteer activities in their “computations” (e.g., included both formal and informal volunteer work in the denominator); of the four unpaid activities described in the scenario, only two represented formal volunteer activities. And secondly, some research participants appeared to struggle with the response task (i.e., computing a ratio).

F2. Recommendations: If memory serves, the wording and format of S15.5 is already a compromise agreed upon in order to avoid gathering detailed data on hours for all organization-specific volunteer activities—local/domestic, long-distance/domestic and foreign. Knowing what we know now, it seems more reasonable to simply ask persons who have done volunteering abroad to report how many hours they had done so; this sort of response task is much simpler and one that should produce fairly accurate data, relatively speaking. But the sponsor seems willing to try the ratio approach—and our simulations could be misleading, given that very few of our research participants had done any formal volunteering abroad. So, it may be best to ignore my misgivings, leave question wording as is, and see how S15.5 works in September 2007.

G. Long-Distance [Domestic] Volunteering: S16.3, S16.4 and S16.5

G1. Observations: Items S16.3, S16.4 and S16.5 were not asked in September 2006; they did not exist at that time. But these questions were asked during the course of our cognitive interviews, both during a simulated CPS-supplement interview and also during a second simulated interview with the research participant serving as proxy for a hypothetical household member (“Joe”). In the latter simulation, a pre-scripted scenario was used to evaluate both items (see Attachment C). For the most part, S16.3 and S16.5 appeared to work as intended; but we did observe two problems during the proxy simulation for S16.4. First, in the process of generating a response to S16.4, some research participants included informal volunteer activities in their “computations” (e.g., included both formal and informal volunteer work in the denominator); of the four unpaid activities described in the scenario, only two represented formal volunteer activities. And secondly, some research participants appeared to struggle with the response task (i.e., computing a ratio).

G2. Recommendations: As noted, the wording and format of S16.4 is already a compromise agreed upon in order to avoid gathering detailed data on hours for all organization-specific volunteer activities—local/domestic, long-distance/domestic and foreign. Knowing what we know now, it seems more reasonable to simply ask persons who have done long-distance, domestic volunteering to report how many hours they had done so; this sort of response task is much simpler and one that should produce fairly accurate data, relatively speaking. But the sponsor seems willing to try the ratio approach—and our simulations could be misleading, given that very few of our research participants had done any long-distance, domestic volunteering. So, it may be best to ignore my misgivings, leave question wording as is, and see how S16.4 works in September 2007.

H. Civic Engagement: S17 and S17A

G1. Observations: Bilingual interviewers report that Spanish-speaking respondents experience more difficulties with S17 and S17A than do English-speaking respondents. Other than translation (or possibly cultural) issues, the reason for this disparity is not readily apparent. The wording of S17A (i.e., “About how many times …?”) may inadvertently communicate to all respondents that it is acceptable to provide an approximate answer to this question (e.g., three or four meetings) or a vague answer (e.g., “a couple of times”). Some interviewers will probe to identify the best answer, which is good practice, while others will simply select one of the two numbers provided by the respondent, and move on.

G2. Recommendations: Other than ensuring that the Spanish translation of S17 seems reasonable to bilingual interviewers (see subsection E.2), I have no recommendations to offer at this time. Given the length of the reference period and the low salience of the subject matter (i.e., attending public meetings), it is not immediately apparent how item S17A might be improved. Revising the wording of this item to emphasize accuracy (e.g., by deleting the word “about”), may simply add burden and increase response time, and probably would not improve estimates substantially.

J. Informal Volunteering: S18 and S18A

J1. Observations: Bilingual interviewers report that Spanish-speaking respondents experience more difficulties with S18 and S18A than do English-speaking respondents. Other than translation (or possibly cultural) issues, the reason for this disparity is not readily apparent. And like S17A, the wording of S18A (i.e., “About how many times …?”) may inadvertently communicate to all respondents that it is acceptable to provide an approximate or a vague answer to this question (e.g., “Oh, maybe ten times”).

J2. Recommendations: Other than ensuring that the Spanish translation of S18 seems reasonable to bilingual interviewers (see subsection E.2), I have no recommendations to offer at this time; but see suggestions below. Given the length of the reference period and the broad range of activities that might qualify as “fixing or improving something”, it is not immediately apparent how item S18A might be improved. Revising the wording of this item to emphasize accuracy (e.g., by deleting the word “about”), may simply add burden and increase response time, and probably would not improve estimates substantially.

J3. Suggestions: Notwithstanding the fact that the interviewer manual provides very helpful/useful guidance as to what counts as “fixing or improving something”, it is my intuition that item S18 underestimates the number of persons who work with neighbors to fix or improve something. [Interviewer instructions, no matter how informative and helpful, will have little effect on response/measurement accuracy if respondents do not request clarification when the wording of a questionnaire item is ambiguous, or when they fail to perceive such ambiguity and interpret the question (or a specific term) in a manner unintended by the survey design team.] One design change that might improve the accuracy of the estimate would be to generate a list of specific sets of activities (response precodes) that reflect the sponsor’s intent, have interviewers read the list of precodes, and ask respondents to answer “yes” or “no” to each set of activities. The specificity of such an approach would greatly clarify the intent of the question. [Note: The wording of S18A would have to be modified accordingly.] Another issue with S18 is the ambiguity of the phrase “in your neighborhood”, which modifies the word “others” in the predicate “worked with others” but which also leads some respondents and interviewers to think that the volunteer work is restricted geographically (i.e., to work done “in your neighborhood”). It may be possible to reduce the level of ambiguity somewhat by rephrasing the question: for example: “Since September 2007, have you worked with others in your neighborhood to fix some problem or improve some condition in your community or elsewhere?”

K. Other Issues

K1. Proxy Responding: During monitoring and coding of supplement interviews on the first day of CPS interviewing (September 17, 2006), I was surprised to hear several proxy interviews being conducted. Later that week, during subsequent focus groups, several interviewers described the conditions under which they feel a proxy interview is justified (e.g., when a parent insists on answering for a teenage child; when a respondent asserts that her spouse will not answer the survey or notes that he/she is away on business). Given uncertainty regarding the accuracy of proxy responses in all but the most clear-cut of cases (e.g., a disabled spouse), gathering data from proxy respondents should be avoided or minimized. Do CPS interviewers receive explicit guidelines as to when proxy interviews can be conducted? Are there data available on the number of proxy interviews that have been conducted each day during CPS survey week?

K2. Reference Period: Based on feedback from interviewers during the focus groups, relatively few respondents commented on their ability/inability to recall behaviors and volunteer activities that had taken place over the course of the twelve-month reference period—this in direct contrast to a substantial cognitive literature which demonstrates that memories for episodic behaviors/activities drops off rapidly with the passage of time. During the cognitive interviews, several research participants admitted to having some difficulty remembering what they had done over the past twelve months (item S1 and elsewhere) and in trying to separate volunteer from non-volunteer activities. [One participant, who reported not having experienced recall difficulties when answering S1, actually had forgotten to report two volunteer activities when he initially completed the volunteer supplement for himself.] While forgetting sporadic volunteer activities may not be a serious measurement issue for individuals who volunteer regularly—they are still likely to remember the volunteer work they do consistently—it does represent a serious measurement issue for the episodic volunteer, the person who does volunteer work infrequently and who may have not done so recently. For the episodic volunteer, general recall questions like S1 and S18 provide few of the specific recall cues that would be helpful in recalling infrequent and temporally distant volunteer activities; and the rapid pace of most CPS telephone interviews is not likely to facilitate the determined efforts that would be needed for respondents to bring such memories to mind.

I hope these observations, recommendations, suggestions and comments prove helpful in making improvements/modifications to the volunteer supplement and to the interviewer manual. If I can clarify or elaborate upon any of the information provided herein, please do not hesitate to call (202-691-6368).

CPS VS/ Memos 060507

Attachment A: CPS Volunteer Supplement:

List of To-Be-Evaluated Questionnaire Items

Label	Supplement Questions
PRESUP	This month, we are interested in volunteer activities, that is, activities for which people are not paid, except perhaps expenses. We only want you to include volunteer activities that you did through or for an organization, even if you only did them once in a while.
S1	Since September 1st of last year, (have you/has NAME) done any volunteer activities through or for an organization? [Note: If “yes”, go to S3; if “no”, go to S2.]
S2	Sometimes people don’t think of activities they do infrequently or activities they do for children’s schools or youth organizations as volunteer activities. Since September 1st of last year, (have you/has NAME) done any of these types of volunteer activities?
S2A	Sometimes people don’t think of activities they do through religious organizations as volunteer activities. Since September 1st of last year, (have you/has NAME) done any of this type of volunteer activity? [Note: If “no”, go to S17.]
S3	How many different organizations (have you/has NAME) volunteered through or for in the last year, that is, since September 1, 2006?
S4A	What organization is it? [Enter response and then go to S4A1 and code response.] Note to Interviewer: ENTER NAME OF THE ORGANIZATION IF RESPONDENT DOESN’T KNOW NAME OF ORGANIZATION, ASK “WHAT TYPE OF ORGANIZATION IS IT?” AND RECORD VERBATIM RESPONSE
S4A1	Note to Interviewer: ASK IF NECESSARY. DO NOT READ CATEGORIES ALOUD. What type of organization is that?
	<1> Religious organization <2> Children’s educational, sports, or recreational group <3> Other educational group <4> Social and community service group <5> Civic organization <6> Cultural or arts organization <7> Environmental or animal care organization <8> Health research or health education organization, including public health <9> Hospital, clinic, or health care organization <10> Immigrant/refugee assistance <11> International organization <12> Labor unions, business, or professional organization <13> Political party or advocacy group <14> Public safety organization <15> Sports or hobby group <16> Youth services organization <17> Some other type of organization/Specify
S5A	During how many weeks in the last year did (you/NAME) do volunteer activities for (fill with response to S4A)?
S6A	In those (fill with response to S5A) weeks that you volunteered for (fill with response to S4A), how many hours per week did (you/NAME) do volunteer activities? [Note: A precoded response of “<V> Varies” is available.]
S7A	How many hours did (you/NAME) do volunteer activities for (fill with response to S4A) in the last year? [Note: This question is only asked of persons who answer “less than one week” or “don’t know” to S5A or “hours vary” to S6A.]
S8	Now I’m going to ask you about activities (you /NAME) might have done for (fill with response to S4A) in the last year. For each activity that I mention, please tell me—yes or no—whether (you/NAME) did that activity for that organization in the last year. Since September 1, 2006, did (you/NAME): READ AND MARK ALL THAT APPLY.
	<1> Coach, referee, or supervise sports teams <2> Tutor or teach <3> Mentor youth <4> Be an usher, greeter, or minister <5> Collect, prepare, distribute, or serve food <6> Collect, make or distribute clothing, crafts, or goods other than food <7> Fundraise or sell items to raise money <8> Provide counseling, medical care, fire/EMS, or protective services <9> Provide general office services <10> Provide professional or management assistance including serving on a board or committee <11> Engage in music, performance, or other artistic activities <12> Engage in general labor; supply transportation for people <13> Other (specify)
S8a	Which of the activities that (you/she/he) performed did (you/NAME) spend the most time doing for (fill with response to S4A) last year? [Note: This question is asked if the respondent identified two or more activities in question S8 above. Also, if deemed necessary, interviewers are permitted read back to the respondent those activities she/he identified in S8.]
S9	Now I’d like to ask you how you first became a volunteer for that organization. Did you approach the organization yourself, were you asked by someone, or did you become involved in some other way?
S10	Who asked you to become a volunteer for this organization?
	<1> Friend <2> Relative <3> Co-worker <4> Someone in the organization/school <5> Boss or employer <6> Someone else: Specify ___________
S11	READ IF NECESSARY: Please describe how you became involved with this organization?
S15	Did any of the volunteer work you’ve done since September 1st of last year take place in a foreign country, that is, outside the United States or any of its territories? [Note: If “yes”, go to S15.5; if “no”, go to S16.3]
S15.5	Considering all of the volunteer work (you have/NAME has) done since September 1st of last year, about how much of it was done abroad? Would you say:
	<1> All or almost all <2> More than half <3> About half <4> Less than half <5> Very little
S16.3	[Fill if S15=yes: Other than (your/NAME’s) volunteering abroad,] Did any of the volunteer work (you have/NAME has) done since September 1st of last year take place more than 120 miles from your home? [Note: If “yes”, go to S16.4; if “no”, go to S17.]
S16.4	Considering all of the volunteer work (you have/NAME has) done [fill if S15=yes: in the United States] since September 1st of last year, about how much of it was done more than 120 miles from your home?
	<1> All or almost all <2> More than half <3> About half <4> Less than half <5> Very little
S16.5	In what state or states did it take place?
	[Interviewer Note: Include the respondent’s home state if she/he traveled 120 miles or more to the work site and were still within their home state.]
S17	Now I’d like to ask about some of your involvement in your community. Since September 1st, 2006, have you attended any public meetings in which there was discussion of community affairs?
S17a	About how many times did you do this?
S18	Since September 1st, 2006, have you worked with other people in your neighborhood to fix or improve something?
S18a	About how many times did you do this?

Attachment B: Scenario Used in Cognitive Interviews to Simulate

Formal Volunteering in a Foreign Country

Instructions to Research Participant: We’d like you to imagine that you have a retired relative living in your household—let’s say his name is Joe—who did a considerable amount of volunteer work during the past year (specifics provided below). Please take a few minutes to read the information provided below. When you have finished, we’d like to run through our questionnaire again to see how well our questions are working with regard to this scenario.

Description of Household Member: Joe is a healthy man, 58 years of age, who worked as a carpenter for 30 years before retiring in 2003. He has lived with you and your family for the past six years.

Volunteer Work (March 2006 through March 2007):

• Joe helped to repair the walls and ceilings of several classrooms in the local elementary school during the month of August. He worked about ten hours per week for four weeks (40 hours total). The local school system sponsored his work and paid for all construction materials, but Joe was NOT PAID for any of his services.

• Joe also traveled to Guatemala in October with two of his friends to help repair storm-ravaged homes in the village of San Pietro on the Pacific coast. Joe worked about forty hours a week for three consecutive weeks (120 hours total) and accepted NO PAYMENT for his services. The work performed by Joe and his friends was sponsored by San Pietro social-services organization and all construction materials and travel expenses were paid for by that organization.

• Joe also got together periodically with several close friends in the neighborhood to provide free, home-related services to low-income, elderly residents. He did mostly carpentry work—repaired porches, fences, garage doors and cabinets—and managed to do this work about eight different times a year (about 40 hours total). Joe would not accept payment from any of the elderly residents that he had helped—and the work he had done was not funded or sponsored by any group or organization.

Other Civic Engagement (March 2006 through March 2007):

• Joe attended about 10 community meetings to discuss the needs of elderly residents and to develop strategies for helping the most needy community residents. Each meeting lasted about two hours (20 hours total for ten meetings).

Attachment C: Scenario Used in Cognitive Interviews to Simulate

Formal Volunteering Done Domestically but at a Distance Greater than 150 Miles

Instructions to Research Participant: We’d like you to imagine that you have a retired relative living in your household—let’s say his name is Joe—who did a considerable amount of volunteer work during the past year (specifics provided below). Please take a few minutes to read the information provided below. When you have finished, we’d like to run through our questionnaire again to see how well our questions are working with regard to this scenario.

Description of Household Member: Joe is a healthy man, 58 years of age, who worked as a carpenter for 30 years before retiring in 2003. He has lived with you and your family for the past six years.

Volunteer Work (March 2006 through March 2007):

• Joe helped to repair the walls and ceilings of several classrooms in the local elementary school during the month of August. He worked about ten hours per week for four weeks (40 hours total). The local school system sponsored his work and paid for all construction materials, but Joe was NOT PAID for any of his services.

• Joe also traveled to West Virginia, Georgia and South Carolina during October and November with two of his friends to help repair homes that had been damaged by severe flooding in those states. Joe worked about 40 hours per week for eight weeks (320 hours total) and accepted NO PAYMENT of any kind for his services. All of the work performed by Joe and his friends was sponsored by local social-service organizations and all construction materials and travel expenses were paid for by the communities in which they worked.

• Joe also got together periodically with several close friends in the neighborhood to provide free, home-related services to low-income, elderly residents. He did mostly carpentry work—repaired porches, fences, garage doors and cabinets—and managed to do this work about eight different times a year (about 40 hours total). Joe would not accept payment from any of the elderly residents that he had helped—and the work he had done was not funded or sponsored by any group or organization.

Other Civic Engagement (March 2006 through March 2007):

• Joe attended about 10 community meetings to discuss the needs of elderly residents and to develop strategies for helping the most needy community residents. Each meeting lasted about two hours (20 hours total for ten meetings).