A universal contention in the psychological literature is that attitudinai similarity leads to attraction. I argue that attitudinai similarity does not lead to liking but that dissimilarity does indeed lead to repulsion. Primary attention is given to Byrne's experimental paradigm in which subjects are shown the attitude scale of a stranger that is similar or dissimilar to their own and who are then asked to indicate their attraction to the stranger. Consistently, Byrne and others have found a linear relation between similarity and attraction. Unfortunately, the Byrne paradigm has never included a control condition in which ratings are made in the absence of attitudinai information. Research that used the Byrne paradigm and other procedures that included an appropriate control group is reported, and support is found for a repulsion hypothesis. Byrne's reinforcement model of attraction is also shown not to be supported. Consideration is given to special conditions in which attitudinai similarity does lead to attraction, to the origins of the hypothesis that similarity leads to attraction, and to the theoretical basis for the repulsion effect.At the end of a chapter in their excellent book Interpersonal Attraction, Berscheid and Walster (1978) offered a summary statement:The answer to the question "Does attitudinal similarity generate liking?" is a resounding "yes." When we discover that others share our beliefs and attitudes, it is satisfying; we like them. When we discover that others disagree with us, it is unsettling; it's hard to like such persons, (p. 88)

Two experiments were conducted in which triads participated in multitrial block-stacking tasks that allowed for objective measures of productivity and process. In Experiment 1, the task was executed either interdependently in the form of a single tower or individualistically in the construction of three separate towers. Reward points were distributed equally (cooperative), in relation to contribution (independent), or only to the most productive group member (competitive). Results indicated that cooperative and independent systems were associated with greater productivity than competitive systems only under conditions of high task interdependence, but there was no relation between reward system and productivity for the individualistic task. Competitive systems also impaired facilitative process events (e.g., turn taking) and led to less efficient, poorer quality products than cooperative systems. In Experiment 2, independent and competitive allocations were each combined in varied proportions with cooperative allocations to examine the effects of mixed reward systems on productivity and process. The results of this second study showed that even a modicum of competitive reward led to lowered efficiency and productivity.

Byrne, Clore, and Smeaton attempted to account for my failure to find differences between similar attitude and nonattitude control conditions primarily by invoking the concept of assumed similarity. Their own statements in previous work argue against bringing to bear mediating processes that are not theoretically specified or empirically supported. In fact, the Byrne group has provided the only data relating assumed similarity and attraction, and the results show a lack of support for their argument. In addition, the quantitative aspects of the attraction model in the form of the Byrne-Nelson formula would not seem to have the power to reveal differences between critical predicted and obtained values associated with differential assumed similarity. The proposed two-stage process of relationship development is welcomed.

No abstract

No abstract