THE FOUR TECHNIQUES WERE SELF-REPORT, BEHAVIOR AS REPORTED BY CORRECTIONS OFFICERS, ATTITUDE VARIABLES, AND BACKGROUND VARIABLES. SELF-REPORT AND ATTITUDE QUESTIONNAIRES WERE FILLED OUT BY 364 INMATES IN A MEDIUM SECURITY INSTITUTION. BEHAVIOR CLASSIFICATIONS (NOMINATIONS) WERE MADE BY 13 GUARDS. BACKGROUND DATA CAME FROM THE INSTITUTION'S FILES. THE THREE COMMON INMATE SOCIAL TYPES EXAMINED WERE 'RIGHT GUYS,' 'OUTLAWS,' AND 'WHEELER-DEALER-POLITICIANS.' THERE WERE 63 RIGHT GUYS IDENTIFIED THROUGH BACKGROUND VARIABLES, 77 THROUGH ATTITUDE FACTORS, 140 FROM SELF-REPORTS, AND 4 FROM BEHAVIOR NOMINATIONS. THERE WERE 102 OUTLAWS IDENTIFIED FROM BACKGROUND DATA, 25 FROM ATTITUDES, 17 FROM SELF-REPORTS, AND 21 FROM BEHAVIOR NOMINATIONS. WHEELER-DEALERS HAD SIMILAR DISPARITY: 106 FROM BACKGROUND REPORTS, 125 FROM ATTITUDE REPORTS, 34 FROM SELF-REPORTS, AND 19 FROM BEHAVIOR NOMINATIONS BY GUARDS. A SERIES OF TABLES SHOW THAT THE PERSONS NOMINATED BY EACH METHOD WERE GENERALLY DIFFERENT INDIVIDUALS. RESEARCH METHODS REPORTED IN THE LITERATURE SEEM TO FOCUS ON THE FOLLOWING IDENTIFICATION PROCEDURES: AN 'IDEAL TYPE' WHICH IS BASED ON ATTITUDES AND BEHAVIOR, AN 'EMPIRICAL TYPE' WHICH FOCUSES ON BACKGROUND, AND A 'SOCIAL TYPE' WHICH IS A BEHAVIORAL CONSTRUCT. IT WOULD BE USEFUL TO DISTINGUISH AMONG THESE CONCEPTS WHEN DISCUSSING INMATE CLASSIFICATION. THIS ARTICLE CONTAINS NOTES, REFERENCES AND A CHART. (GLR)