Citation

Ekblom P, Pease K. Crime Justice 1995; 19: 585-662.

Copyright

(Copyright © 1995, University of Chicago Press)

DOI

unavailable

PMID

unavailable

Abstract

Most evaluations of crime prevention are carried out with little regard for methodological probity. Of work that aspires to methodological adequacy, the standard designs are the before-after comparison group and the interrupted time series. The critical questions are whether a program has an effect (and if not, whether because of theory failure, implementation failure, or measurement failure), the extent of any effect achieved, and the means by which it was achieved. The reduction of crime is a measure of outcomes, but others are often used. Clarity is necessary in the use of noncrime measures. Too often, evaluators settle for the "consolation prizes" of reductions in fear and incivilities. Distinct evaluative requirements attach to different types of prevention. Improvements in standard methodology and innovation by nonstandard approaches are needed. The "scientific realist" tradition eschews conventional one-shot quasi experimentation in favor of repeated manipulations leading to the cataloging of possible mechanisms and consistency of outcome pattern with presumed mechanisms. Different actors (practitioners, evaluators, administrators) have different objectives and interests; evaluators must take account of those differences.

Language: en