Response Entropy and Emotional Indicators in the Kent–Rosanoff Word Association Test: Implications for Psychological Assessment

Abstract

The Word Association Test (WAT) has long been used in psychological assessment to infer emotional disturbance, psychopathological, and creative processes, yet its diagnostic utility has yielded inconsistent empirical findings. This paper reviews theoretical and empirical evidence addressing a key methodological source of these inconsistencies: stimulus-related variability in response entropy and associative difficulty. Drawing on classic and contemporary studies using the Kent-Rosanoff Word Association Test, the review examines how response entropy systematically influences commonly used emotional indicators, including long reaction time, unique and infrequent responses, response repetition, misremembering, and forgetting. Evidence suggests that several indicators traditionally interpreted as reflecting emotional disturbance are substantially moderated by the response hierarchy characteristics of stimulus words. Indicators such as response repetition and response preservation appear less confounded by entropy effects, whereas emotional indicators, long reaction time, repetition of the stimulus before responding, reproduction failure, and response rarity require cautious interpretation. These findings underscore the importance of incorporating stimulus-level psychometric properties into the interpretation of WAT performance. Implications for assessment practice and test construction are discussed, emphasizing the need for controlled stimulus words selection when using word association test measures in clinical and applied assessment contexts.