Research variables were based on diagnostic-rehabilitation, prognostic-public defense, organizational, and conflict models. The diagnostic-rehabilitative model assumes that criminals are ill persons suffering from psychosocial deficiencies. Thus, in this model, the probation officer's recommendation should be based mainly on the offender's personal, family, and social characteristics. The prognotic-public defense model assumes that punishment must be meted out proportionately to the offender's degree of danger to the public, a conclusion requiring a prognosis for the offender relating to recidivism. According to this model, the probation officer's recommendation should consider personal, family, and social variables and especially criminal variables, the most prominent of which is the offender's criminal record. The organizational model implies that probation officer decisionmaking is influenced by general legal system principles. The conflict model presumes there are constant power struggles between social groups. The probation officer's recommendation in this model is based on the offender's social inferiority. These models were applied to Israeli offenders between 18-21 years of age in cases where a jail sentence or probation was handed down during the 1978-1981 period. Based on research variables associated with each model and use of the Multiple Classification Analysis technique, the findings did not verify the diagnostic-rehabilitation model. The conflict model was verified to a limited degree. Results indicated the predominance of the prognostic-public defense model and the organizational model. 49 references, 2 tables. (Author abstract modified)