Salience and Order in Impression-Formation

Abstract

Due to the pervasive confounding of effects of order in impression formation, an attempt was made to reduce these effects. The addition of a storage and recall task to the standard impression-formation procedure led to the prediction that increased attention would be paid to all of the stimulus words within a given list of traits. Consequently, it was expected that this increased attention would be reflected in a reduction of effects of order between subjects and a more neutral evaluation of persons who are described by equal amounts of isovalent traits (positive and negative traits which are affectively equidistant from neutrality). This hypothesis was derived from finding no effects of order in studies of recall (cf. Bartz. Lewis, & Swinton, 1972), reduced effects when recall followed evaluation (Anderson & Hubert, 1963). and more accurate impression formation when inconsistency is made salient (Hendrick, 1972). Sixty-eight subjects were randomly assigned to one of 4 equal-sized groups: (1) recall and evaluation, (2) reverse-order recall and evaluation, (3) evaluation, and (4) reverse-order evaluation. The stimuli were 11 lists, each containing 6 isovalent traits (3 positive and 3 equally negative) which were presented as a descriprion of a person. Subjects in Groups 1 and 2 were required to store and recall each list and then make an evaluation of the person (likeability-rating, - 5 to + 5) while Groups 3 and 4 subjects simply read each list and then made likeability-ratings. No reduction in effects of order was found in rhe nonsignificant interaction of task X order (F = 1.00, df = 1/763). An over-all effect of recency was obtained (F = 4.60, df = 1/763, p < .05) in that descriptions which contained 3 positive traits followed by 3 negative traits were consistently rated more negatively (M = -.55) than were acg3rive-positive descriptions (M = -.24). However, the hypothesis that salience of the neutrality of descriptions would be increased with the inclusion of a storage and recall task and that this increased salience would be reflected in more neutral ratings was confirmed (F = 15.80, df = 1/736, fi < ,001). Noc only were likeability-ratings closer to neutrality (0) in the recall-and-evaluadon groups (M = -.lo) than in evaluarion-only groups (M = -.70). but the incidence of neutral ratings was also significantly greater (X3 = 19.5, df = 1, p < ,001) in recall-and-evaluation groups (77/340 = 23%) than in evaluation-only groups (30/340 = 95% ). The failure to reduce effects of order was perhaps a function of the inadequacy of the task. However, the data suggest that increased familiarity with descriptions of persons (via storage and recall) facilitates the formation of more accurate impressions because the affective meaning of the descriptions is made salient.