Truthfulness of personality-descriptive statements and the Barnum effect
Iennarella, Ralph S.
MetadataShow full item record
Base-rate refers to the prevalence of some characteristic within a population. Two traditions have arisen in studies involving this concept. In studies investigating Bayesian information processes, base-rate refers to some imputed level of incidence of a characteristic, preset by experimental instructions. In studies examining the acceptance of generalized personality interpretations, base-rate refers to an empirical or actual level of incidence of a characteristic. Research in both traditions has shown that subjects do not make optimal use of base-rate information. Bayesian studies have shown that base-rate has less impact on subjective probability judgements than expected from Bayes' theorem, while personality interpretation studies have revealed subjects' ready acceptance of high base-rate interpretations as being uniquely true of them. The latter phenomenon has been dubbed the "Barnum effect." Research has focused primarily on situational factors which increase acceptance of generalized personality interpretations. The theoretical processes that may underlie the Barnum effect have received relatively little attention. The present study investigated the effects of truthfulness and base-rate on subjects* acceptance of personality interpretations. The design was a 2 x 2 factorial with within-subject independent variables of population base-rate (high vs. low) and truthfulness for the individual subject (true vs. false) of interpretations. Each subject received four types of interpretations: personally true/high base-rate, personally false/high base-rate, personally true/low base-rate, and personally false/low base-rate. Subjects were asked to make a number of assessments of each of the interpretations, constituting the dependent variables. Results showed that personally true interpretations yielded higher ratings than personally false interpretations for accuracy, uniqueness, and scoring system adequacy measures. These findings were explained in terms of subjects' ability to act in a rational manner. High base-rate interpretations produced higher ratings than low base-rate interpretations for accuracy, uniqueness, and scoring system adequacy measures. The findings for the accuracy and adequacy ratings were discussed in terms of the greater likelihood of an interpretation being true for a given individual, the truer it is for the general population. The findings for the uniqueness ratings were discussed in reference to self-enhancement motivations. No interactions between personal truth and base-rate were found for any of the measures. In spite of the initial equation of the four types of interpretations for social desirability, an analysis of desirability ratings revealed main effects for truthfulness and base-rate. It was suggested that this reflected the operation of a general halo. An analysis of covariance, with social desirability as a covariate, revealed results essentially unchanged from the univariate analyses for accuracy, uniqueness, and scoring system adequacy. Correlations between the ratings of each of the four types of interpretations and the rating of the interpretations as a set were obtained, separately for each dependent measure. For each assessment measure, evaluations of the interpretation set were most related to the subjective judgement of its poorest single interpretation (i.e., the least accurate, or least unique, or worst scoring system, or most undesirable interpretation). The clinical implications of the findings of this study were also discussed. In general, in order to maximize acceptance, an interpretation should be perceived by a client as being personally true and true of most others. Interpretive statements which are untrue for most people and for the particular recipient carry the risk of affecting believability of the entire set of statements; one bad apple spoils the barrel. Good apples, however, bear little relationship to evaluation of the total set.