Self-reported health depends on the internal frame of reference and on response styles. One way of studying this dependency is using anchoring vignettes. Response shift effects are assumed to induce a negative correlation between self-reported health and the health assessments attributed to the vignettes. Method: A representative sample of the German adult population (N = 2,409) was selected. Participants were asked to rate their health state and the health states of two rather complex vignettes representing patients with several health complaints on a 0-100 scale. Results: The mean score of self-assessed health was M = 76.20 (SD = 20.6). There was a very small positive correlation between the assessment of the vignettes and the self-assessed health state (r = .12). After controlling for a proxy of objective health, measured in terms of chronic conditions, the relationship remained slightly positive. Chronic conditions were only marginally associated with the assessments of the vignettes (0 conditions: M = 44.8; ≥ 2 conditions: M = 42.2). Conclusions: The lack of the postulated association between self-reported health and vignettes’ ratings means that we cannot derive tools to correct the subjective ratings for differential use of frames of reference.

Self-reported health depends on the internal frame of reference and on response styles. One way of studying this dependency is using anchoring vignettes. Response shift effects are assumed to induce a negative correlation between self-reported health and the health assessments attributed to the vignettes. Method: A representative sample of the German adult population (N = 2,409) was selected. Participants were asked to rate their health state and the health states of two rather complex vignettes representing patients with several health complaints on a 0-100 scale. Results: The mean score of self-assessed health was M = 76.20 (SD = 20.6). There was a very small positive correlation between the assessment of the vignettes and the self-assessed health state (r = .12). After controlling for a proxy of objective health, measured in terms of chronic conditions, the relationship remained slightly positive. Chronic conditions were only marginally associated with the assessments of the vignettes (0 conditions: M = 44.8; ≥ 2 conditions: M = 42.2). Conclusions: The lack of the postulated association between self-reported health and vignettes’ ratings means that we cannot derive tools to correct the subjective ratings for differential use of frames of reference.