Discrepancies Between the Rorschach Inkblot Method and Self-Report Measures of Personality: Methodological and Theoretical Reflections

Because the quasi-absence of correlation between Rorschach variables and the dimensions described in academic psychology of personality is now a well-established fact, this paper reviews the main reasons which could account for this puzzling discrepancies. Some of these reasons are methodological or statistical ones, and they are linked to formal properties of variables to be correlated: raw scores, ratios, or percentages, or broader constructs operationalized by specific pattern of scores and ratios. It is emphasized that some difficulties are related to the excessive number of variables addressed in studies lacking theoretical hypotheses, as to the categorical nature of many Rorschach variables, which opposes the dimensional nature of the constructs of academic personality psychology. More fundamentally, it is suggested that, as exemplified by introversion and introversiveness, psychological realities assessed by the Rorschach differ from those conceptualized by personality psychology not only by their content but also by their structure. Finally, the authors discuss the hypothesis suggested by Ganellen and Meyer: the correlation between the Rorschach and self-report measures would be moderated by the patients' response style or test-taking attitude; the absence of correlation in the studied samples would conceal strong positive correlations among subjects with the same attitude toward both tests, and strong negative correlations among subjects with test-taking attitude different toward each kind of instrument. It is suggested that this hypothesis is promising, but it will be illuminating only when a reliable and specific marker of attitude will be identified. To conclude, it is remembered that while these discrepancies between the two kinds of tests raise some problems at the theoretical level, at a clinical level they convey much valuable information about patient's self-representation and defence mechanisms. Partant du fait maintenant bien établi de la quasi-absence de corrélations entre les variables du Rorschach et les dimensions décrites en psychologie différentielle de la personnalité, les auteurs passent en revue les principales raisons susceptibles de rendre compte de ce décalage surprenant. Certaines de ces raisons sont d'ordre méthodologique, voire statistique et tiennent aux propriétés formelles des variables mises en relation: scores isolés, proportions ou pourcentages, ou dimensions plus vastes opérationnalisées par un ensemble de plusieurs