Under commonly observed conditions, happy subjects appear to process information in a relatively passive or nonsystematic, less detailed manner and rely on peripheral cues and heuristics in judgement, whereas sad subjects appear to process in a more active or systematic, more detailed manner. Happy subjects should therefore display less accuracy on judgements that have a relatively objective accuracy criterion. Three studies were conducted to test this hypothesis. In study 1, subjects who had training in statistics were exposed to a happy, neutral or sad mood induction procedure. Subjects then judged the magnitude and direction of correlation coefficients associated with each of nine scatterplots. Happy subjects were least accurate and used fewest digits in their correlation estimates: sad subjects were most accurate and used most digits. In study 2, subjects exposed to orthogonal affect and arousal mood inductions, completed the correlation estimation task. To address process further, subjects provided ratings of their concentration on the task, their strategies for estimating the correlations, and explained the concept of a correlation. Happy subjects were least accurate, used fewest digits in their estimates, reported least concentration, provided least detail in their explanations of correlations, wrote least comprehensible explanations; however, they wrote most creative descriptions. Sad subjects displayed the opposite pattern. Arousal had minimal effects on all measures. In study 3, processing strategy was directly manipulated by asking half of the subjects to think in detail while completing their correlation estimates digit effects shown in studies 1 and 2 were replicated. Implications are discussed. (A)
Abstract