To study the role of parens patriae and "police powers" considerations in an individual judge's civil commitment decisions, the judge's reports of the impact of various characteristics of the patient were analyzed. The validity of this methodology was tested by comparing it to an alternative technique based on objective statistical analysis of the dependence of the judge's decisions upon patient characteristics. A probate court judge filled out a questionnaire after each civil commitment hearing over which he presided during a seven-month study. For each of 26 decisions, the judge rated the patient on 26 features and indicated the impact of each feature on the decision. The judge's responses were analyzed to measure the role of various statutory and nonstatutory considerations (expressed as patient characteristics) in the judge's decisions. Results using self-reported impacts are compared with an objective, statistical characterization of the judge's decision-making policy. As in previous studies, the parens patriae model more closely described the individual judge's decision process than the "police powers" model. Contextual variables (e.g., the patient's family favoring commitment) also were influential. Results with the two methods were similar. The methodology developed here can be used not only in further research on judicial commitment decisions but also to educate judges and other decision-makers individually faced with potentially tragic choices as to their personal implicit decision-making strategies.