National Survey of American Life Self-Administered Questionnaire (NSAL-SAQ), February 2001-June 2003 (ICPSR 27121)

Version Date: Apr 13, 2010 View help for published

Principal Investigator(s): View help for Principal Investigator(s)
James S. (James Sidney) Jackson, University of Michigan. Institute for Social Research; Cleopatra H. Caldwell, University of Michigan. Research Center for Group Dynamics; David R. Williams, University of Michigan. Survey Research Center; Harold W. Neighbors, University of Michigan, Research Center for Group Dynamics; Randolph M. Nesse, University of Michigan. Research Center for Group Dynamics; Robert Joseph Taylor, University of Michigan. Research Center for Group Dynamics; Steven J. Trierweiler, University of Michigan. Research Center for Group Dynamics

Version V1

Slide tabs to view more

The National Survey of American Life, 2001-2003 (NSAL) was followed up by a self-administered interview (NSAL SAQ) as a way to reduce respondent burden following the 2 1/2 hour NSAL survey. The SAQ includes additional questions about social, group, and individual characteristics: psychological resources (i.e., John Henryism), group and personal identity (racial awareness and identity), as well as ideology and racial relations (i.e., social dominance; stratification beliefs; egalitarianism; national pride; work ethic; authoritarian, interracial contact; and exposure to Black social contexts); political attitudes (i.e., Race-conscious Policy Index, Race-blind Policy Index, Non-Electoral Participation Index); care of elderly values; job and financial stressors; and wealth. Demographic variables include age, race, and sex.

Jackson, James S. (James Sidney), Caldwell, Cleopatra H., Williams, David R., Neighbors, Harold W., Nesse, Randolph M., Taylor, Robert Joseph, and Trierweiler, Steven J. National Survey of American Life Self-Administered Questionnaire (NSAL-SAQ), February 2001-June 2003. Inter-university Consortium for Political and Social Research [distributor], 2010-04-13.

Export Citation:

  • RIS (generic format for RefWorks, EndNote, etc.)
  • EndNote

census-tract; county

To protect respondent privacy, the data are restricted from general dissemination. Users interested in obtaining these data must complete an Agreement for the Use of Confidential Data, specify the reasons for the request, and obtain IRB approval or notice of exemption for their research. Apply for access to these data through the ICPSR Restricted Data Contract Portal, which can be accessed via the study home page.

Inter-university Consortium for Political and Social Research

2001-02 -- 2003-06
  1. The variable CPESCASE may be used to merge these data with the NSAL data.

  2. For more information on applying for restricted data please go to the Applying for Restricted Data page on the ICPSR Web site.

Complex survey data sample that has a random component to it.

The NSAL SAQ is a subset of the NSAL. All NSAL adult respondents had an opportunity to complete a supplemental self-administered mail survey (NSAL SAQ) following their NSAL interview. The SAQ 40-page questionnaire contains key questions initially included in the NSAL that were removed from the initial interview in order to reduce respondent burden. The respondent was asked if they wanted to participate in a mailback survey. If they agreed, the questionnaire was sent by mail and once received, an incentive payment was sent. The overall response rate was 56.5 percent (n = 3,438); the response rate was 59.9 percent for African Americans (n = 2,137), 42.9 percent for Caribbean Blacks (n = 695), and 68.0 percent for Whites (n = 606). Overall, the demographic distributions of re-interview respondents did not differ greatly from those of the main NSAL sample. The largest differences were among response rates for the different race groups. Being female, being unemployed, having higher levels of education, and participating in the original NSAL interview post-September 11, 2001, were also associated with higher response rates on the mailback survey relative to other NSAL respondents. The mailback survey weights take these differences into account.


Of the 6,082 NSAL respondents, 3,438 completed the RIW. The overall response rate was 56.5 percent. Response rates differed for each race group: 68 percent of Whites and 60 percent of African Americans responded compared to 43 percent for Caribbean Blacks.




2018-02-15 The citation of this study may have changed due to the new version control system that has been implemented. The previous citation was:
  • Jackson, James S. (James Sidney), Cleopatra H. Caldwell, David R. Williams, Harold W. Neighbors, Randolph M. Nesse, Robert Joseph Taylor, and Steven J. Trierweiler. National Survey of American Life Self-Administered Questionnaire (NSAL-SAQ), February 2001-June 2003. ICPSR27121-v1. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2010-04-08.

2010-04-13 Previously, released restricted files without making documentation public. The documentation is now available.


Modified versions of the NSAL weights were created for use in analyses involving variables from the NSAL re-interview. These weights were created by multiplying the original NSAL weights by an additional factor adjusting for nonresponse in the self-administered mail questionnaire. A logistic regression to predict likelihood of mail survey response was conducted using all 6,082 NSAL respondents. Predictors in the final model included race; sex; education; work status; age; home ownership; foreign-born versus US-born; whether the respondents' main NSAL interview took place after September 11, 2001; participation in the NSAL Clinical Reappraisal interview; count of chronic health conditions; length of main NSAL interview; respondent prayer and religiosity; respondents' assessment of his/her weight; financial hardship, respondents' birth order; welfare receipt; respondent impatience during the main NSAL interview; and respondent suspicion about main NSAL interview. Several interactions were also included in the model. Predicted probabilities of responding to the mail survey were calculated and, for respondents, the inverse of this probability was multiplied by the original NSAL weight, and then divided by the mean of these new weights. All mail survey nonrespondents received a weight of zero on the mail survey weights. Mail survey weights should be used for all analyses where any variables from the mail survey are included, regardless of their status as independent variables or dependent variables.