The effects of three methods of evaluation (norm-referenced, regulated criterion-referenced, and negotiated criterion-referenced) on academic self-concept (ASC), test anxiety, and aspiration of success were studied for 73 college students (71 females and 2 males) from a large midwestern research university. Students were given a limited amount of time to complete a series of difficult word analogies (subjects were given 3 minutes to complete each of 5 trials). Those in the norm-referenced group were told that their scores would be based on a comparison with other students. Subjects in the regulated criterion-referenced groups were told that a score of 10 out of 20 for each trial would be necessary in order to be considered proficient on the analogies test. Subjects in the negotiated criterion-referenced group were allowed to state the score that they thought would indicate proficiency. The time limit was imposed so that students' ASC, test anxiety, and aspiration of success could be studied under induced failure conditions. Data were analyzed using split plot analysis of variance; and it was found that the method of evaluation had no significant effect on subjects' ASC, test anxiety, and aspiration of success. Three tables contain the study data. A 26-item list of references is included. (SLD)