Classical conditioning of human 'evaluative' responses.

Human behaviour includes animportant component which may conveniently be called evaluative. Using postcard reproductions as stimulus materials. 10 volunteer subjects selected the two pictures most liked and the two most disliked. These were then used as UCSs with appropriate controls from the neutral category. The conditioning hypothesis-that a neutral stimulus followed by a positively or negatively valued stimulus will acquire the evaluative weight of the second stimulus-was supported at a highly statistically significant level. The effect of negative evaluation was demonstrably stronger than that for positive evaluation, a result consistent with our knowledge of aversive conditioning. The possibility is discussed that evaluation of the UCS by the subject, shown to be a su~cient condition for learning, may also be the only necessary condition. This would imply a model of~onditioning based on affective evaluation rather than on response production.