IN A SERIES OF 4 PRELIMINARY STUDIES, THE MODEL DESCRIBED HERE WAS EMPIRICALLY REFINED INTO A PERSON-TIMES-SITUATION-RESPONSE MODE-TIMES-REINFORCER (SANCTION) MODEL WITH THESE VARIABLES REPRESENTED BY THE NUMERALS 3, 2, 4, AND 2, RESPECTIVELY. THE UTILITY OF THE MODEL WAS ASSESSED USING QUESTIONNAIRE AND EX-POST-FACTO METHODOLOGY. SECOND-ORDER AND THIRD-ORDER INTERACTIONS WERE INCONSEQUENTIAL IN ACCOUNTING FOR BEHAVIORAL VARIANCE. MAIN EFFECTS OBTAINED FROM THE QUESTIONAIRE STUDY WERE THAT INMATES REPORTED A GREATER LIKELIHOOD OF COMMITTING INFRACTIONS IN FREE TIME THAN IN HIGHLY SUPERVISED SUBSETTINGS, OF COMMITTING INFRACTIONS WHEN PARDONED THAN WHEN PUNISHED, AND OF COMMITTING VERBAL AGGRESSION INFRACTIONS AND ENGAGING IN EVASIVE BEHAVIORS. THESE FINDINGS SUGGEST THAT INSTITUTIONAL DISCIPLINE IS MOST EFFECTIVE WHEN A STRICT AND SYSTEMATIC POLICY IS MAINTAINED SO THAT FEW INFRACTIONS GO UNDETECTED AND INMATES PERCEIVE THAT ALL INFRACTIONS WILL LIKELY LEAD TO PUNISHMENT. MANAGEMENT POLICIES SHOULD BE DESIGNED SO THAT INCARCERATION IN THE INSTITUTION IS LESS CONDUCIVE TO VERBAL AGGRESSION AND EVASION. IT IS CONCLUDED THAT A COMPLEX MODEL INVOLVING SPECIFICATION OF RESPONSE MODES AND REINFORCERS (SANCTIONS), AS WELL AS PERSON DIFFERENCES AND THE PSYCHOLOGICAL IMPACT OF SITUATIONS, MAY BE MEANINGFULLY OPERATIONALIZED IN A CORRECTIONAL SETTING AND MAY YIELD DATA THAT HAVE PRACTICAL IMPLICATIONS FOR THE MANAGEMENT OF INMATES RULE INFRACTIONS. FURTHERMORE, THE MODEL'S PRACTICAL UTILITY IN ACCOUNTING FOR VARIANCE IN A GIVEN DIMENSIONAL RESPONSE SYSTEM WILL PARTLY DEPEND ON AN INVESTIGATOR'S SKILL IN CHOOSING PSYCHOLOGICALLY DISTINCTIVE LEVELS OF EACH INPUT VARIABLE AND OPERATIONALIZING THEM TO ACCOUNT FOR MAXIMUM DEGREES OF RESPONSE VARIANCE. REFERENCES, FIGURES, AND TABLES ARE PROVIDED. (DEG)