This report attempts to profile students who participated in a computer-assisted language learning (CALL) research and development project. The researchers hoped to find some pattern or common denominator within each group i.e., the successes, the failures, and the dropouts, that would identify group membership and distinguish one group from another in order to further understand the characteristics of language learners and to be in a position to rationally design alternative methods of computer-assisted language learning for failures and dropouts. To accomplish this objective, three groups of factors were examined, namely, individual characteristics, strategy use, and course information, to try to account for students' differential performance. The subjects for the study were 158 students enrolled in freshman composition classes during the spring semester, 1990, at Northern Arizona University in Flagstaff. Participants undertook four CALL computerized reading and note-taking lessons; two achievement tests were part of each lesson, so that altogether there were eight measures of achievement. After finishing all four lessons, students turned in four types of data from each of the lessons--notes, recall, recognition, and attitude. Scores were then computed for the recall and recognition parts and these scores formed the basis of the grouping criteria. Methods used to assess the three types of factors that might have contributed to group membership are explained next, and finally the Discriminant Function Analysis out of which a profile for the subjects in each group emerged, is discussed. It is concluded that with more information about what students are like; i.e., what kind of profile describes a successful or unsuccessful student, aids can be incorporated into computer assisted lessons that will help students who experience difficulty to overcome their problems rather than leaving them with no alternative but to be forced into the mold of the "ideal" successful student. (KM)