A comparison of hypothetical versus real attributions for social situations : examination of a clinic population

Abstract

Nineteen male adolescent and 19 female adolescent patients, undergoing short-term residential treatment at a psychiatric facility, were asked a series of questions about 13 different life events. One of the events (actual) was the basis for the client's admittance into the inpatient treatment program at the hospital. The remaining 12 events were one-sentence descriptions of incidents (hypothetical) that patients may or may not have experienced. Hypothetical events were grouped into mild and severe categories (6 events in each category). Subjects were asked (a) to explain what caused success or failure for each event- and (b) to judge how important each event was to them. Questions were designed to highlight attributions ( i.e., internality, stability, and globality) that people typically use to explain beliefs about what causes what. Initial analysis of the data was targeted at examining differences in how male and female adolescents responded to questions pertaining to actual and hypothetical situations. For each of the 12 hypothetical events and the one actual event, adolescents were queried about importance and cause. Analysis of the data based on aggregating events into hypothetical-mild and hypothetical- severe categories and comparing mean levels of responding across causal dimensions resulted in few interpretable differences. Male and female adolescent ratings tended to vary no more than a few tenths of a rating point. Moreover, differences among dimension ratings were found to be small and not statistically significant. Including importance as a covariate in the analysis failed to produce a consistent finding. When composite scores, summing across events and causal dimensions, were used, controlling for importance eliminated any effect due to Event Type. When composite scores, summed across events but nested within causal dimensions were used, the no difference for Event Type finding was not maintained. Finally when specific event mean ratings within causal dimension were compared two events in the mild category and three events in the severe category differed from the actual event on at least one dimension. In summary, no evidence from this study could be interpreted to support the use of composite scores to identify attributional styles of responding for hospitalized adolescents.