There is some evidence that peer assessments are biased by friendship (that is, employees evaluate their friends most favorably), but friendships may be formed on the basis of performance. Also, many work group members do not like to evaluate one another, so part of the method’s success hinges on convincing participants of its value. Indeed, Cederblom and Lounsbury (1980) showed that lack of user acceptance may be a serious obstacle to this otherwise promising method. They found that a sample of college professors thought peer assessments were heavily biased by friendship. They thought peers would rate and be rated by their friends more favorably than would be justified. Problems with knowing the people to be rated and fostering a “mutual admiration society” caused the professors to question the value of peer assessment. They also felt that the method should be used for feedback, not for raises and promotions. Despite reluctance to use peer assessments for administrative decisions, research continues to support their predictive accuracy. Shore, Shore, and Thornton (1992) found peer assessments to be superior to self-assessments in predicting advancement