Abstract
Several studies suggest that ‘don’t know’ (DK) responses to risk perception items may represent meaningful expressions of uncertainty about disease risk. However, researchers are often discouraged from including a DK response option in survey items due to concerns about respondents overusing it to minimize cognitive effort—a phenomenon often referred to as satisficing. Our objective was to investigate whether patterns of DK responses to risk perception survey items were consistent with satisficing behavior. We conducted a secondary analysis of survey data from 814 parents and guardians (hereafter caregivers) of children with asthma. Caregivers answered 18 items assessing their perceived risk of their child experiencing two types of poor asthma outcomes: asthma exacerbation, and low asthma control. We examined differences in the frequency and distribution of DK responses across all 18 items and by type of risk perception item (i.e. 2 vs. 5 response options, absolute vs. comparative risk). We found that 32% (n = 548) of respondents marked DK at least once. Of the 266 caregivers who provided any DK response, most did so for only 1 or 2 items (51.9%, n = 138), and only 6% (n = 15) answered DK to more than half of the items. Using random coefficient Poisson models, we found more DK responding for dichotomous absolute (30.1%) than ordinal absolute items (5.3%), b = 1.72, p<.001. We also found fewer DK responses to the ordinal absolute items than the comparative items (8.2%), b=-0.49, p<.001. Using Chi-square tests, we found that inattentive responding was not associated with responding DK. Our findings suggest that satisficing is unlikely to completely explain DK responding to perceived risk survey items. Researchers who exclude DK response options from risk perception survey items may obtain an incomplete understanding of their study sample’s beliefs about risk.
Ethical approval
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Consent to participate
Informed consent was obtained from all participants prior to engaging in the study.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Data availability statement
The design and variables for the parent study that provided data for the analyses described herein were pre-registered prior to data collection. Those pre-registration documents and the de-identified data, the analytic code used to conduct the analyses presented in this manuscript, and all materials used to conduct the study are freely available in the Open Science Framework public archive: (Masked for review).
Notes
1 In exploratory analyses, we observed that lower numeracy and less formal education were associated with: (a) more DK responding (i.e. no DK response vs. any DK response; rNumeracy = .13, rEducation = .19, p’s < .001) and (b) a greater number of DK responses (i.e. range 0-18; rNumeracy = .10, rEducation = .18, p’s < .003). The association between numeracy and DK responding appeared limited to the comparative items (rNumeracy*Comparative = .14, p < .001); the associations between numeracy and absolute ordinal and absolute dichotomous items were not statistically significant. In contrast, the association between education and DK responding were significant for each of the three indices (rEducation*Comparative = .11, rEducation*AbsoluteOrdinal = .09, rEducation*AbsoluteDichotomous = .18, all p’s < .006).