The central hypothesis of the first study is that young, inexperienced officers make arrests that are not as likely to produce convictions as compared to arrests made by older, experienced officers. While this hypothesis was not confirmed, results indicated that convictability is a function of offense severity, the particular judge hearing the case, the type of pretrial release from jail, and the number of days elapsed between the defendant's arrest and disposition. The second study, which involved a comparison of characteristics of prosecuted vs. nonprosecuted cases, was prompted by the vast expenditures allocated by the criminal justice system to the processing of cases that do not result in conviction of the defendant and are dismissed by the prosecutor. Prosecuted and nonprosecuted defendants were compared on 42 factors and analyzed for statistical significance. The two categories of defendants were significantly different on the factors of verification of information, jail population on day of arrest, retention of attorney at time of booking, charge level, violent offense, use of weapon during crime, possession of telephone by defendant, type of defendant residence, and type of pretrial release from jail. The final research analysis focused on the various factors associated with the sentencing options of amount of jail time and amount of the fine. Findings indicated that the amount of jail time a defendant receives is significantly influenced by severity of charge, time between arrest and disposition, prior arrests, and specific sentencing judges. The amount of fine was significantly related to the amount of bail, specific sentencing judges, violent offenses, charge severity, and marital status. There was no evidence that the amount of jail time or fine amount was related to the race, sex, or economic status of the defendant. The appendixes contain the offenses included in the data set, ranges of possible punishment by levels of offenses, information contained in the data set, and codes and the range of values for variables in the data set. A selected bibliography contains 145 listings.