Key points are not available for this paper at this time.
Each of 320 male subjects was paired with a fictitious female partner with whom they were led to believe they either would or would not interact. Three other variables were manipulated: the partner's attitude similarity, the partner's physical attractiveness, and whether or not the partner would be evaluating the subject. A significant interaction was found between attitude similarity and anticipated interaction. Contrary to predictions from reward-cost theory, the pattern of means in the interaction indicated that attitude similarity had a greater effect upon attraction when no future interaction with the partner was anticipated than when future interaction was expected. The results were interpreted in terms of the amount of information available to the subject at the time the evaluation ratings were administered.
Layton et al. (Sat,) studied this question.