Substantive Meanings of Missing Data in Family Research: Does “Don't Know” Matter?

Abstract

In this article we analyze “don't know” responses from three sources of longitudinal data: the National Longitudinal Study of Adolescent Health (n = 14,528), the National Survey of Families and Households (n = 5,488), and the National Health Interview Survey Second Longitudinal Study of Aging (n = 1,131). We asked whether these responses are meaningful in family research, and, if so, how evaluating these responses can contribute to the development of theory, the discovery of novel findings, and identification of sensible methods for analyzing these nebulous responses. We found that “don't know” responses to questions about family members predicted less educational attainment, poor marital quality, and earlier mortality. Results suggest that this response category may have substantive meanings rather than indicating neutral responses or being missing data.

KEYWORDS:

• measurement
• missing data
• nonresponse
• regression
• relationship ambivalence
• survey research

ACKNOWLEDGMENTS

The authors thank Justin Dyer and Sarah O. Meadows for their insightful comments as discussants during the 2012 TCRM workshop. This study uses data from the National Survey of Families and Households (Waves 1 [1987–88] and 2 [1992–1994]) public use data set. NFSH public use data set is distributed by the University of Wisconsin-Madison (Madison, WI) with funding from the National Institutes of Health. This study also uses data from Add Health, a program project directed by Kathleen Mullan Harris and designed by J. Richard Udry, Peter S. Bearman, and Kathleen Mullan Harris at the University of North Carolina at Chapel Hill. Add Health was funded by grant P01-HD31921 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, with cooperative funding from 23 other federal agencies and foundations. Special acknowledgment is due to Ronald R. Rindfuss and Barbara Entwisle for assistance in the original design. To obtain the Add Health data files, contact the Add Health website (http://www.cpc.unc.edu/addhealth). No direct support was received from grant P01-HD31921 for this analysis.

Notes

^a White is the reference category.

^b Parent received a bachelor's degree or more is the reference category. This model uses data from the Add Health survey. OR = Odds ratio.

*p < .05, **p < .01, ***p < .001.

^a White is the reference category.

*p < .05, **p < .01, ***p < .001.

^a White is the reference category.

*p < .05, **p < .01, ***p < .001.

In supplementary analyses we examined whether DKs were significantly different from adolescents who said at least one parent received a bachelor's degree, adolescents who did not identify any parent(s) in receiving bachelor's degrees, and adolescents who refused to answer the question, on adolescent receipt of a bachelor's degree. These analyses were conducted using the same four models in Table 2. Adolescents who did not know their parents' educational attainment were less likely to receive a bachelor's degree than those who said at least one parent received a bachelor's (significant in Models 1 through 4) and compared with those who did not identify any parent(s) as receiving bachelor's degrees (significant in Models 1 and 2). This supplementary table is available upon request.

In supplementary analyses we examined whether DKs were significantly different from respondents who said their partner desires no to two children, respondents who said their partner desires more than two children, and respondents who refused/gave no answer on marital quality. These analyses were conducted using the same four models in Table 3. Those who reported DK on their spouse's desired number of children had lower marital quality compared with those who reported their spouse desiring more than two children (significant in Models 2 through 4) and compared with those who reported their spouse desiring no to two children (significant in Model 3). This supplementary table is available upon request.

In supplementary analyses we examined whether DKs were significantly different from refusals and the following reported ages at mother's death: 18–49 years, 50–59 years, 60–69 years, 70–79 years, 80–89 years, 90 years and over. People who did not know their mother's age at death were different from those who did know, no matter the actual age of mother's death. This supplementary table is available upon request.