The Effect of Context on the Correspondence Between Questionnaire and Observational Data

• Activity Number: 327
• Type: Contributed
• Date/Time: Wednesday, August 6, 2003 : 8:30 AM to 10:20 AM
• Sponsor: Section on Survey Research Methods
• Author(s): Frederick O. Lorenz*+
• Companies: Iowa State University
• Keywords: family studies ; observational data ; questionnaire design ; hostility ; support

Abstract

Family researchers are interested in the relationship between observed reports and questionnaire reports of couples' behavior. These correlations are often low, in part because they include different contexts: observed behavior is based on the limited videotapled session, whereas questionnaire items often include preambles such as, "During the past month, how often have you. . ." In this study, we videtaped young couples (N = 300) inteacting with each other (A), and then had them complete questionnaires that contained items asking them about their behavior toward each other both "during the past month" (C) and "during the videotaping session" (B). For both hostile and supportive behaviors, the correlations between A and C were roughly a product of the correlations between A and B and between B and C [i.e., r(ac) = r(ab)*r(bc)]. Although there are other sources of variation (e.g., observed hostility is not identical to the hostility measured in questionnaires), these results do show that the much of the lack of correspondence between questionnaire and observational data is due to the lack of a common context.