A PROTOTYPE INFORMATION FEEDBACK SYSTEM IS DEVELOPED AND APPLIED TO PROBATION TREATMENT SELECTION AND EVALUATION. CLIENTS ARE CLASSIFIED USING THE INTERPERSONAL MATURITY LEVEL CLASSIFICATION PROCEDURE, AND THE VALUE OF TREATMENTS IS DETERMINED FROM RESULTS OF THE CALIFORNIA TEST OF PERSONALITY (CTP). THE MATHEMATICAL MODEL USED TO ANALYZE THE CHARACTERISTICS AND PERFORMANCE OF THE INFORMATION FEEDBACK SYSTEM IS A STOCHASTIC DISCRETE EVENT SIMULATION. THE UNCONTROLLABLE VARIABLES TREATED INCLUDE THE SIZE AND CHARACTERISTICS OF THE CLIENT POPULATION, DURATION OF TREATMENTS, AND THE MEASUREMENT ERROR OF THE CTP. DECISION VARIABLES INCLUDE THE NUMBER OF CLIENTS EVALUATED EACH PERIOD USING THE CTP, THE DELAY IN FEEDBACK OF THE EVALUATIONS TO DECISIONMAKERS, AND THE CERTAINTY WITH WHICH THE TREATMENT RECOMMENDED BY THE INFORMATION SYSTEM IS SELECTED BY THE DECISIONMAKER. EXPERIMENTS WITH THE MODEL DEMONSTRATE THAT A RANDOM SELECTION OF TREATMENTS CAN EVOLVE INTO THE SELECTION OF THE BEST TREATMENT FOR EACH CLIENT CLASS OVER A RELATIVELY SHORT PERIOD OF TIME. THE SENSITIVITY OF THE RESULTS TO THE SAMPLE SIZE OF CLIENTS, FEEDBACK DELAY, AND CTP MEASUREMENT ERROR IS ALSO EXPLORED. TABLES OF DATA AND GRAPHS ARE INCLUDED. APPENDIXES SUMMARIZE THE CTP SOCIAL ADJUSTMENT SUBSCALE AND THE INFORMATION FEEDBACK MODEL EQUATIONS. (AUTHOR ABSTRACT MODIFIED--RCB)