Interview Methods

This section provides the following information on interview methods:

• Interview procedures for Main Youth survey (Current interview procedures and round 1 procedures and instruments)
• Content of the Main Youth survey questionnaire
• Parent-provided data (Parent Questionnaire and Household Income Update)
• Fielding periods
• Respondent incentives
• Special survey collections

Interview procedures for Main Youth survey

Many users may be most interested in current (through round 20) interview procedures; round 1 procedures (which involved many steps not needed in subsequent rounds) are also provided here as background information.

Current interview procedures (rounds 2 through present)

The interviews are conducted each round using a CAPI (computer-assisted personal interview) instrument, administered by an interviewer with a laptop computer. Computer software automatically guides interviewers through an electronic questionnaire, selecting the next question based on a respondent's answers. The program also prevents interviewers from entering invalid values and warns interviewers about implausible answers. A set of checks within the CAPI system lowers the probability of inconsistent data both during an interview and over time. The preferred mode of interview is in person. When an interview is conducted in person, during sensitive portions of the interview, the respondents enter their answers directly into the laptop rather than interacting with the interviewer. This self-administered portion, called ACASI, includes an audio option so that the respondents can listen to the questions and answers being read via headphones if they prefer. The audio component theoretically improves response quality when the respondent's literacy is in question. In some cases, due to the location of the respondent or the respondents' reluctance to be interviewed in person, interviews are conducted by phone. In this case the interviewer must administer the SAQ sections. Table 1 shows the number of in-person and telephone interviews for each round.

Round 1 procedures and instruments

Screener, Household Roster, and Nonresident Roster Questionnaire. This instrument was initially administered to a member of each household selected for sampling in the NLSY97 survey areas. It was completed by a household resident age 18 or older, referred to as the household informant. This questionnaire was used to identify youths potentially eligible for the NLSY97 survey and/or the administration of the computer adaptive version of the Armed Services Vocational Aptitude Battery (CAT-ASVAB), a military enlistment test (see Administration of the CAT-ASVAB for more information). In general, eligibility for either group was dependent on the youth's age and, in some sample areas, on the youth's race or ethnicity. Sample Design & Screening Process provides more detailed information about the precise age and race/ethnicity requirements. This instrument comprised four sections: the simple screener, extended screener, household roster, and nonresident roster, administered in that order and described below. Question names for the simple and extended screeners begin with "SE," household roster questions begin with "SH," and nonresident roster questions begin with "SN."

The simple screener section was first administered at each household in the sample areas to identify potentially eligible sample members. The simple screener collected the name and birth date or age of each person who could be linked to the household (see Sample Design & Screening Process for more information). If the household contained a youth potentially eligible for the NLSY97 or the administration of the CAT-ASVAB, the household informant completed the extended screener. This section collected the gender and race/ethnicity of each person in the household and the year in school the potentially eligible youth was currently attending, if any. Race and ethnicity were used in supplemental sample areas as further criteria for NLSY97 eligibility; in certain cases, grade in school affected eligibility for the administration of the CAT-ASVAB. If the household contained a youth eligible only for the administration of the CAT-ASVAB, the interviewer solicited his or her participation. If the household included one or more youths potentially eligible for the NLSY97, the interview continued with an extensive two-part roster.

The household informant first answered the questions in the household roster section of the Screener, Household Roster, and Nonresident Roster Questionnaire. This section established the relationships between household occupants and collected basic demographic information (e.g., marital status, highest grade of schooling completed, employment status) for all household members, including any NLSY97-eligible youths. After the roster was created, one parent of the sample youth was chosen from the list of household occupants and asked to participate in the Parent Questionnaire. Finally, the household informant was administered the nonresident roster section, which gathered data on members of the youth's immediate family (e.g., biological, adoptive, or stepparents; biological or adoptive siblings; spouse; biological children; parent of the youth's biological children) who lived elsewhere at the survey date.

Choice of Household Informant. To identify youths potentially eligible for the NLSY97, the screener collected data from selected households within a sample area. A single member of the household, designated as the household informant, was asked to provide certain information on persons who usually resided in the household. To ensure more accurate reporting of these data, the NLSY97 required the household informant to be age 18 or older and to consider the selected household his or her usual place of residence.

Bilingual Interviewing. To ensure that accurate data were collected from Spanish-speaking respondents, CHRR prepared both English and Spanish versions of all survey instruments, and NORC employed bilingual Spanish-speaking interviewers to administer the Spanish version to those requesting it. During the initial round, the Spanish version of the questionnaire was requested by 297 responding parents and 96 NLSY97 youths.

Screen and Go. In round 1, use of the computer-assisted personal interviewing system (CAPI) allowed for a screen and go method of screening households. When an NLSY97-eligible youth was identified in the simple screener portion of the interview, information from the remainder of the Screener, Household Roster, and Nonresident Roster Questionnaire was collected. Selected data (e.g., basic demographic information, a roster of household members) were then transferred automatically into the Parent and Youth Questionnaires for verification and use during the interview. Therefore, the interviewer could administer the parent or the youth portion of the NLSY97 immediately. It was expected that this would increase the likelihood that eligible youths would participate in the survey since the number of visits interviewers had to make to a household decreased.

However, in some cases, the respondents (parent and youth) were not available to participate in the parent and youth interviews immediately after screening. In these cases, a screen and come back method was utilized, in which the interviewer made an appointment to return to the household to administer the Youth and Parent Questionnaires at a convenient time.

Paper Screener. During round 1, the interviewers had the option of using a paper screener to perform the initial screening of the household. The paper screener collected the same basic information as the initial CAPI screener. This was useful in cases where the simple screener information could not be collected using CAPI (e.g., weather conditions, computer battery life, dangerous neighborhood) and also gave the interviewer an alternative medium for collecting the initial screener data. Like the screen and go model, the paper screener was designed to determine if anyone residing in the housing unit was eligible for either the NLSY97 or the administration of the CAT-ASVAB. If a youth was identified as being potentially eligible for the NLSY97, the information from the paper screener was entered into CAPI. The interviewer could then continue in CAPI with the Screener, Household Roster, and Nonresident Roster Questionnaire and the Youth and Parent Questionnaires. Approximately 28,000 paper screeners were administered, including those used for the screen and come back method described above.

Proxy Screener. In cases where a round 1 interviewer made several visits to a household and still could not contact household members to administer the initial screener, a proxy screener was administered to an adult living either next door to or directly across from the selected housing unit. Before the interviewer could administer a proxy screener, at least three attempts were made by the interviewer, on different days and at different times, to contact anyone in the selected housing unit.

The purpose of the proxy screener, a paper questionnaire, was to assess whether a person eligible for the NLSY97 resided in the household. In particular, the proxy screener was designed to determine the best time to establish contact with a household member, whether or not a person between the ages of 8 and 28 currently lived in the household, and the steps required to contact a household member. The broad 8-28 age range was intended to ensure that youths close to the endpoints of the actual age range were not missed due to inaccurate reporting. If the proxy screener indicated that none of the household members were in the age range of 8 to 28, the screener was coded as a proxy screener and no more attempts were made to contact the household. However, if the proxy informant was unable to definitively deny the presence of residents ages 8-28, the interviewer was instructed to return as many times as reasonable and necessary to administer the simple screener and, if appropriate, the remainder of the survey instruments. A total of 5,175 proxy screeners determined that no one between ages 8 and 28 lived in the household.

Gatekeepers. The gatekeeper disposition code was used in cases where the interviewer could not gain direct access to the sample household, such as a high-rise building with a locked door where access was denied by a building manager or a gated housing community where the entry guard refused entrance. In these cases, the interviewer asked the gatekeeper or other community official whether anyone between the ages of 8 and 28 lived in the sample households. If the gatekeeper was unable to definitively deny the presence of household members ages 8-28, the interviewer then attempted to gain access to the household in order to complete the Screener, Household Roster, and Nonresident Roster Questionnaire and was not permitted to use this disposition code. A total of 4,055 cases were closed with a gatekeeper disposition code after the interviewer determined that no one between ages 8 and 28 lived in the household. This code was mainly used in gated housing communities for senior citizens.

Telephone Screener. In rare cases, the simple screener was conducted by telephone at the conclusion of the field period. A total of 931 telephone screeners were administered. Instances in which the housing unit was contacted by telephone include:

1. The proxy screener revealed a person between the ages of 8 and 28 living in the household and the interviewer was unable to contact anyone in the housing unit on three subsequent in-person visits; or
2. The interviewer made three in-person visits but was unable to find a neighbor to whom he or she could administer the proxy screener.

The full Screener, Household Roster, and Nonresident Roster Questionnaire was also administered by telephone in rare instances. Situations in which the full instrument was conducted by telephone include:

1. After completing the paper screener, the interviewer was unable to contact anyone in the housing unit to complete the full extended screener. At least three in-person contacts must have been attempted before the telephone contact was approved.
2. The sample housing unit was inside a residential community to which the interviewer was barred access by the community (e.g., housing board authority). Prior to the telephone interview, the correct person must have been contacted about gaining access at least three times (in person, by telephone, or by letter).

Youth Questionnaire and Parent Questionnaire. When the Screener, Household Roster, and Nonresident Roster Questionnaire was complete, any NLSY97-eligible youth(s) and one of the youth's parents (the responding parent) were interviewed using CAPI (See Parent Questionnaire section for more information on parent data). Prior to these interviews, selected data (e.g., basic demographic information, a roster of household members) were automatically transferred into the Parent Questionnaire and the Youth Questionnaire for verification and use during the interviews. Consequently, the interviewer was able to administer the parent or the youth portion of the NLSY97 immediately. CAPI interviews were conducted in either English or Spanish; parent and youth respondents could choose either version.

In round 1, the NLSY97 youth respondent(s) and responding parent(s) in the household are listed on the household roster, but they are referred to as "Household Member #" in the same way as noninterviewed household members. The youth respondent's position on the household roster can be identified by using the variable YOUTH_HHID.01. The responding parent's position on the roster is provided in PARYOUTH_PARENTID. See Household Composition for further discussion of the structure and use of the household roster.

Content of the Main Youth survey questionnaire

In the initial round, 8,984 youth completed the Youth Questionnaire, the main component of the NLSY97 data set. In each survey round after round 1, the Youth Questionnaire is administered to every eligible respondent who can be located and who agrees to be interviewed. Table 2 shows the Youth content.

Interviewer Remarks: Each NLSY97 questionnaire also includes an interviewer remarks section, which interviewers complete after finishing the interview with the respondent. This section records objective information about the interview, such as the presence of another person during the survey, where the interview took place, and the language in which the questionnaire was administered. Interviewers are also asked to provide an overall assessment of the interview. See Interviewer Remarks, Characteristics & Contacts for more details.

Parent-provided data

One parent of the youth respondent was selected to provide information via a Parent Questionnaire in the first round. In rounds 2 through 5, a parent answered questions via a Household Income Update. After round 5, parents did not answer questions in the NLSY97.

Parent Questionnaire

The Parent Questionnaire (administered in round 1 only) collected extensive background information from one of the youth's biological parents. If no biological parent lived in the respondent's primary household, another adult household member was selected for the Parent Questionnaire according to predetermined criteria (see Choice of Parent below for a detailed description of this process). Table 3 shows the content found in the Parent Questionnaire.

Choice of Parent. One parent of each respondent was asked to participate in the parent interview. This parent was identified during the household roster portion of the survey. The responding parent (or guardian) was asked for extensive background information, including marital and employment histories. He or she was also asked to answer questions about the family in general, as well as to provide information about aspects of his or her (NLSY97-eligible) children's lives.

The choice of the preferred responding parent was based on the pre-ordered list in Table 4. For example, a biological mother was chosen before a biological father, and so forth. However, in some cases a parent figure lower on the list was chosen if a parent higher on the list was in the household but was not available at the time of the interview. If the youth did not live with a parent-type figure, or lived with a guardian or parent not listed, no parent was interviewed; the youth's record will not contain any data from the Parent Questionnaire. Users should note that the records of some youths who do live with a listed parent or parent-figure do not contain any data from the Parent Questionnaire due to nonresponse.

Interviews are available with 6,124 parents; 7,942 youth respondents have information available from a parent interview. Table 5 shows the number of respondents by age who had a parent participate in the round 1 survey.

In multiple respondent households, more than one parent may have been interviewed during round 1 if the selection criteria above indicated different parents for different NLSY97-eligible youths in the household. For example, if a couple residing in a sample household each had an NLSY97-eligible youth from a previous marriage, the biological parent of each youth would be interviewed. The survey first collected parent-specific information from each parent and then asked for information about the NLSY97-eligible youth matched to that parent. In this example, each parent would be asked to provide youth-specific information for his or her NLSY97-eligible biological child.

Due to a computer programming error, however, both parents in some multiple respondent households were asked to provide youth-specific information only for the oldest NLSY97-eligible youth(s) living in the household. In the example above, both parents would be asked to give information about the older of the two children. In these infrequent instances, the correct parent-specific information is matched to each youth, but one or more youths in the household do not have any youth-specific information. This programming error was corrected during the survey period and affected only 33 youth cases.

Household Income Update

In rounds 2-5, the brief Household Income Update questionnaire collected basic income information from one of the respondent's parents (usually the parent who signed the youth's interview consent form). All respondents who lived with a parent were eligible for this questionnaire, regardless of age or other criteria for independence. If the youth respondent lived with a parent, this questionnaire collected the parent's total pre-tax income from wages, salaries, commissions, and tips during the past calendar year; the same data for the parent's spouse or partner; and the total pre-tax amount of any other income (i.e., farm or business income, inheritances, child support, government programs) received by the parent and spouse/partner together. These data were collected for the respondent's current household regardless of his or her residence during the past calendar year. Question names for the variables based on the Household Income Update begin with "HIU."

The parent answered these questions on a self-administered paper instrument. Interviewers then entered the data into a computer-assisted questionnaire on their laptops and attached the information to the records of all NLSY97 youths in the household. Additional quality control checks were performed in the central office, where hard copy questionnaires were reviewed against the coded data. In round 2, parents of 7,601 respondents answered at least one question from the Household Income Update. Parents of 5,488 respondents answered at least one question in round 3, and 5,225 parents of respondents answered at least one question in round 4. Parents of 4,090 respondents answered at least one question in round 5.

Beginning in round 6, all respondents were at least 18 years old, so the Household Income Update is no longer administered.

Fielding periods

With the exception of round 1, the NLSY97 fielding periods began in the fall of the survey year and continued through the spring of the following year. Surveys were conducted annually through round 15, with biennial interviews beginning in round 16.

Locating respondents is a coordinated effort of NORC's central office, locating shop, and local-level field staff. Prior to fielding, NORC's central office sends a short, informative "locator letter" to each respondent reminding him or her of the upcoming interview and confirming the respondent's current address and phone number. During the field period, field interviewers use contact information to track down hard-to-find respondents, while central office staff assist with database searches and other centralized locating methods.

Round 1 fielding period

Most round 1 NLSY97 interviews were conducted between January and early October 1997. Due to concerns about the number of eligible youths found during the initial field period, investigators decided to conduct a refielding between March and May 1998. During this second part of the initial survey round, 395 additional respondents were interviewed. These respondents (identified in the dataset as CV_REFIELD_YOUTH) were administered the same instrument as those initially interviewed in 1997.

Researchers analyzing topics where time periods are critical should carefully examine the reference period of the questions and the actual interview date for individual respondents. In particular, the round 1 fielding period has implications for questions on education; see Educational Status & Attainment for more information.

Researchers should also pay close attention to the elapsed time between interviews for each respondent. While the time between the first and second interviews was about 18 months for most respondents, it may be less for those first interviewed during the refielding period.

Rounds 2 and up fielding periods

Most respondents were surveyed approximately 18 months after their first interview, although the elapsed time between interviews was substantially less for some respondents. Subsequent fielding periods also began in the fall. Respondents were surveyed annually through round 15, then biennially starting in round 16.

Table 6 provides fielding periods along with sample size and retention rates for each round.

Respondent incentives

Rounds 1-3. Respondents received $10 for their participation in rounds 1-3 of the Youth Questionnaire. Responding parents received $10 when they completed the Parent Questionnaire.

Round 4. Survey administrators offered different levels of incentives to respondents in an effort to study the effects of incentive level on survey participation. Three levels of compensation were offered: $10, $15, and $20. In addition, half of the respondents at each level were paid in advance and half were paid upon completion of the interview. Both the level and the timing of the compensation are included in the variable PAYINCENT, found in the round 4 data.

Rounds 5 and 6. All respondents received $20.

Round 7. Respondents who had not completed the Round 6 interview were eligible for an incentive experiment (R7_INCENTIVE), where respondents in the experimental group were offered an additional $5 for each consecutive round in which they had not participated (up to a maximum of an additional $15), while respondents in the control group were offered the standard $20 incentive.

Rounds 8 and up. Various incentive experiments have taken place. Users should use the Rx_INCENTIVE variables to determine the exact amount given to each respondent. Additionally, in round 10 questions were added about whether the respondent received a gift card or other in-kind payments in addition to the regular respondent payment.

Special survey collections

In addition to the NLSY97 main youth survey administered annually, special data collections took place in the early years of the survey that provided additional information about these young adults.

School surveys

School Survey (1996). Designed with an emphasis on the school-to-work transition, round 1 of the NLSY97 also included a mail survey of schools. Principals (or their proxies) were asked to complete a self-administered instrument that focused on institutional-level attributes such as school policies and management as well as student-level "experience" data. See School & Transcript Surveys for more detail about the content of the survey. Due to confidentiality restrictions, access to these data is limited. More information can be found on the BLS website's NLSY97 School Surveys page.

Schools in the NLSY97 sample areas that had a 12th grade comprised the sample for this survey. As described in Sample Design & Screening Process, the NLSY97 sample was drawn from 147 primary sampling units (PSUs). The PSUs were further divided into sample segments. All schools in any county with a segment selected for NLSY97 sampling were included in the survey. There were some counties in the PSUs from which no sample segments were selected. The 1996 survey did not include schools in these counties. Schools were identified using the Quality Education Data (QED) file, a proprietary national database of primary and secondary schools in the United States.

The original school survey form was mailed in September 1996; in-scope schools that did not respond by December 1996 were sent a shorter version of the survey, the "critical items" questionnaire. Of the 7,390 in-scope schools that received the survey, 5,295 responded to either the original school survey or the critical items questionnaire. The response rate by the end of the field period, April 5, 1997, was 71.6 percent.

Answer forms for the original school survey were electronically scanned by NORC. However, some hand editing was necessary. The majority of the edited questions were in decimal format. To ensure clean data, the answers were verified by randomly selecting cases, keying the data, and comparing the keyed data files against the scanned data files. The critical items questionnaire did not use a scannable format; the data were keyed using Computer Assisted Data Entry (CADE) and verified twice.

School Survey (2000). Round 3 of the NLSY97 also included a repeat survey of schools. Principals (or their proxies) were asked to complete a self-administered instrument similar to that used in 1996. To reduce the time burden, questionnaire items from the 1996 instrument were modified to encourage respondents to provide approximate values rather than requiring them to consult administrative records for exact figures. See School & Transcript Surveys for more details about the content of the survey. As is the case with round 1 school survey data, data access to round 3 school survey data is limited. See the school survey information at the BLS website for more details.

As in 1996, schools in the NLSY97 PSUs that had a 12th grade were mailed survey instruments. However, the 2000 sample was expanded to include vocational schools. The sample also included schools in the counties that were in NLSY97 PSUs but did not include any sample segments. Schools in these counties had been omitted from the 1996 survey but were included for limited data collection in 2000. No telephone follow-up was done for schools in these "omitted counties." Finally, in addition to the geographically based sample, other schools were included if an NLSY97 respondent was enrolled during round 2 and that school met the grade and program requirements for eligibility. Schools were identified using the 1998 Quality Education Data (QED) file.

By January 2000, survey staff had secured cooperation from state school officers and local school districts. In February 2000, questionnaires were mailed to 9,632 sampled schools, including 8,925 schools in a longitudinal sample (comparable to the 1996 school survey), 492 in the omitted counties sample, and 215 eligible only due to round 2 youth enrollment. After mail and telephone follow-up, 5,955 schools (71.6 percent) in the longitudinal sample (comparable to the 1996 school survey) completed questionnaires. The overall response rate for all schools in the 2000 survey was 71 percent.

Due to "births" and "deaths" of schools between 1996 and 2000 and nonresponse in 1996, not all schools in the longitudinal sample are present in the 1996 data. The retention rate of 1996 schools into the 2000 survey was 74.2 percent (3,900 of 5,253).

Transcript surveys

High School Transcripts. At two separate points in time, the NLSY97 program sought high school transcripts for respondents who were no longer enrolled in high school and for whom field interviewers had secured parent and respondent consent for transcript release. Eligible respondents were those who either had graduated from high school or who were age 18 or older and no longer enrolled in high school. The first wave of transcripts was collected in 1999-2000, the second wave in 2004. Transcripts were received and processed for 1,417 respondents in Wave 1 and 4,815 respondents in Wave 2 for a combined total of 6,232 respondents. Using course catalogs, transcript data, and clarification calls to school administrators, survey staff constructed histories of courses taken and term enrollment calendars for each youth. Data files also include information on absences, standardized test scores, and indicators of special education, gifted/talented, and high school graduation status. Courses were coded into the Revised Secondary School Taxonomy (SST-R). A large set of transcript variables is included in the public use data set, as described in School & Transcript Surveys.

Post-Secondary Transcripts. College transcript data for NLSY97 respondents were collected in 2012-2013. These data include a youth-level file with selected transcript-related data, as well as course and term-level files with variables pertaining to course enrollment. For more detailed information, see School & Transcript Surveys and Appendix 12: Post-Secondary Transcript Study.

Comparison to Other NLS Surveys. School surveys have been conducted for the NLSY79, Children of the NLSY79, the Young Women, and the Young Men. Each of these surveys has recorded information on the school's total enrollment, number of books in the school library, qualifications of the staff, and ethnic/racial composition of the faculty and students. The NLSY79 and Children of the NLSY79 surveys also asked about the school's grading system and average daily attendance.

Transcript surveys have been conducted for the NLSY79 and for the Children of the NLSY79. These surveys have included information on course subject matter, enrollment dates, and grades earned. For more precise details about the content of each survey, users should consult the appropriate cohort's User's Guide.

CAT-ASVAB

The Department of Defense (DOD) used the NLSY97 sample as part of a larger effort to establish new norms for the computer adaptive form of the Armed Services Vocational Aptitude Battery (CAT-ASVAB), a military enlistment test. Administered from summer 1997 to spring 1998, the CAT-ASVAB collected baseline information on the aptitudes and vocational interests of NLSY97 respondents. Additionally, the CAT-ASVAB was administered to a group of participants age 17-23 who were sampled at the same time as NLSY97 respondents (see Administration of the CAT-ASVAB for more information).

COVID-19 survey

A supplemental COVID-19 survey was fielded in the spring of 2021 to document the effects of the pandemic on respondents’ jobs and health. Respondents received invitations to complete a short (about 12 minute) web questionnaire about their experiences during the pandemic. For more detailed information, see Appendix 14: NLSY97 COVID-19 Supplement.