Suchergebnisse

Several major limitations of the ecological approach in the study of deterrence are traced & alternative res strategies are proposed which might facilitate the generation of more accurate & relevant data. 4 major problems in existing res are considered: (1) the failure to diff'iate between general & specific deterrence; (2) the failure to take into account problems of time lag; (3) the failure to select appropriate base pop's; & (4) the failure to utilize homogeneous units of analysis. The ecological approach to the study of general deterrence has the potential of making a signif contribution to the exploration & application of the concept, but--as presently used--it allows only for tentative conclusions at best. Studies should (a) insure that what is being examined is general deterrence. Eg, crime rates should be based only on the crimes of first offenders, rather than on those of both first offenders & recidivists. (b) To account for the fact that the effect of sanctions cannot be expected to be immediately visible after changes in official sanctions, res methodologies should be structured to allow for an appropriate time lag between cause & expected effect. (c) To avoid distortion in estimating the effectiveness of sanctions in changing behavior, the base line "pop at risk" should be defined as specifically as possible. (d) The unit of analysis being considered should be as homogeneous as possible. This requires abandoning the large & arbitrary units of analysis presently used, ie, whole states, cities or census tracts, in favor of small natural soc areas, such as those derived from isoplethic mapping. Finally, the specific methodological suggestions for each of the 4 problems should be incorporated into an inclusive design to provide cohort data for time series & distributed lag analyses. M. Maxfield.