Judgments of behavioral frequencies: An information processing perspective

In marketing and consumer research, respondents are often asked questions about the frequency with which they engage in various behaviors (e.g., bank loans taken, purchases of durable goods, consumption of snack foods). The information processing demands that such questions placed on the respondent can be tremendous. For some behaviors, it may be easy to recall all specific instances of the target behavior and provide an exact count. For relatively frequent behaviors, however, this recall-and-count method may be difficult or impossible; instead consumers are likely to use heuristics to arrive at a behavioral frequency estimate.This research examines the process by which frequency judgments are generated for frequent behavior with the goals of (a) contributing to our knowledge of the cognitive processing of autobiographical information (organization and structure in memory, retrieval strategies), and (b) understanding how different questioning strategies may improve the accuracy of such behavioral frequency judgments.It is hypothesized that the regularity and the similarity of the target behavior determine the way the requisite information is stored in memory and therefore the process by which the judgment is generated. This, in turn, determines the effectiveness of alternative cuing strategies. These hypotheses are tested by obtaining judgments of the frequency of behaviors varying on regularity and similarity. The dependent measures are: (a) verbal protocols, concurrent and retrospective, (b) response time to generate a frequency judgment, and (c) the discrepancy between reported frequency and objective frequency (diary data).Data indicate that there is support for the model of autobiographical memory for frequent events tested in this research. However, the cuing strategies tested seem to interfere with the natural process of frequency judgment formulation in a way that they do not contribute to improving the accuracy of these reports.Theoretical implications of the results for autobiographical information processing and questionnaire and survey design are discussed, together with practical implications and proposed extensions for future research.U of I OnlyETDs are only available to UIUC Users without author permissio