Data Collection
The Survey Questionnaires
The questionnaires in each of the component surveys were largely the same -- roughly 90 percent of the questions were identical. The remaining questions were survey-specific; that is, they collected information relevant only to that survey's population. Each year, the NSCG and SDR surveys used a mixed-mode approach, beginning with a self-administered mail questionnaire. These questionnaires were carefully designed to be as "mode-neutral" as possible to ensure that the mode (self-administered, telephone, or in-person) did not influence a person's responses.
The draft 1993, 1995 and 1997 mail questionnaires were pretested in focus groups. Questionnaires were distributed at the start of the focus group, and the participants were asked to complete the questionnaire as if it had just arrived in the mail. Once the participants had completed their questionnaire, the focus group moderator used a retrospective "Think Aloud" approach to probe for any problems the participants experienced while completing the questionnaire.
See SESTAT Questionnaires to view facsimiles of the survey questionnaires.
Mode of Administration/Response Rates
Mode of administration refers to how a survey is conducted -- by mail, by telephone, or in person. The National Survey of College Graduates (NSCG) and Survey of Doctorate Recipients (SDR) are mixed-mode surveys, while the National Survey of Recent College Graduates (NSRCG) is primarily conducted as a telephone survey. More specifically:
- The 1993 NSCG was a mail survey with telephone and in-person follow-up of sample members who failed to respond by mail. The telephone follow-up was conducted as a computer-assisted telephone interview (CATI). Sample members who did not respond by mail and were not available by telephone were targeted for an in-person interview using a computer-assisted personal interview (CAPI) approach. These efforts achieved an overall 1993 unweighted response rate of 80 percent; 73 percent of the responses were by mail, 15 percent were by telephone (CATI) and 13 percent were in-person.
- The 1993 NSRCG was conducted primarily by telephone (CATI). A handful of sample members, inaccessible by telephone, were mailed a questionnaire. The 1993 NSRCG achieved a unweighted response rate of 85 percent.
- The 1993 SDR was a mail survey with telephone follow-up (CATI) of sample members who did not respond by mail. The unweighted response rate was 87 percent; 80 percent of the responses were by mail, and the remaining 20 percent were by CATI.
- The 1995 NSCG, like the 1993 NSCG, was a mixed-mode survey, but it was restricted to responding, eligible members of the 1993 NSCG and 1993 NSRCG. Nonrespondents to the 1995 mail questionnaire were subsampled for follow-up by CATI and CAPI. An unweighted response rate of 95 percent was achieved; 77 percent of the responses were by mail, 17 percent were by CATI, and the remaining 6 percent were in person.
- The 1995 NSRCG, like its 1993 predecessor, was primarily a CATI survey, with some responses collected by mail. In 1995, cases requiring location searches were subsampled. An unweighted response rate of 83 percent was achieved; fewer than 5 percent of the responses were by mail.
- The 1995 SDR data collection procedures were similar to those for 1993 except that nonrespondents to the mail questionnaire were subsampled for CATI follow-up. An unweighted response rate of 77 percent was achieved; 86 percent of the responses were by mail, 14 percent were by CATI.
- The 1997 NSCG was conducted by two seperate survey organizations. The 1993 NSCG portion of the 1997 NSCG followed the same mixed-mode data collection strategy as the earlier NSCG surveys. An unweighted response rate (complete responses divided by eligible sample) of 94 percent was achieved; 79 percent of completed responses were by mail, 17 percent by CATI, and 5 percent by CAPI. The 1993 and 1995 NSRCG Panels were administered by a CATI survey with some responses collected by mail. The unweighted response rate was 85 percent; 96 percent of the responses were CATI, the remaining 4 percent were mail responses.
- The 1997 NSRCG, like the 1993 and 1995 NSRCG, was primarily a CATI survey with some responses collected by mail. The unweighted response rate was 82 percent; 97 percent of the responses were CATI, the remaining 3 percent were mail responses.
- The 1997 SDR, like the 1995 SDR was a mail survey with CATI follow-up of a sample of mail nonrespondents. An unweighted response rate of 85 percent was achieved (eligible responses divided by eligible sample); 75 percent of the responses were by mail, 25 percent were by CATI.
- The 1999 NSCG was conducted by two separate survey organizations. The 1993 NSCG and 1993 NSRCG Panel portions of the 1999 NSCG followed the same mixed-mode data collection strategy as the earlier NSCG surveys. An unweighted response rate (complete responses divided by eligible sample) of 90 percent was achieved; 80 percent of completed responses were by mail and 20 percent by CATI. The 1995 and 1997 NSRCG Panels were administered a CATI survey with some responses collected by mail and web. The web responses were part of an experiment of the feasibility of offering a web option to respondents. The unweighted response rate was 80 percent; 87 percent of the
responses were CATI, 4% by mail and 8% were received through the web.
- The 1999 NSRCG, like the 1993, 1995 and 1997 NSRCG, was primarily a CATI survey with some responses collected by mail. The unweighted response rate was 79 percent; 95 percent of the responses were CATI, and 5 percent were mail responses.
- The 1999 SDR was a mail survey with CATI follow-up of mail nonrespondents. An unweighted response rate of 82 percent was achieved (eligible responses divided by eligible sample); 84 percent of the responses were by mail, 16 percent were by CATI.
- The 2001 NSRCG, like the 1993, 1995, 1997 and 1999 NSRCG, was primarily a CATI survey with some responses collected by mail. The unweighted response rate was 80 percent; 93 percent of the responses were CATI, and 7 percent were mail responses.
- The 2001 SDR was a mail survey with CATI follow-up of mail nonrespondents. An unweighted response rate of 82 percent was achieved (eligible responses divided by eligible sample); 82 percent of the responses were by mail, 18 percent were by CATI.
