When random assignment fails: Some lessons from the Minneapolis Spouse Abuse Experiment

In this paper, we consider what may be done when researchers anticipate that in the implementation of field experiments, random assignment to experimental and control groups is likely to be flawed. We then reanalyze data from the Minneapolis Spouse Abuse Experiment in a manner that explicitly models violations of random assignment. As anticipated, we find far larger treatment effects than previously reported. The techniques developed should be useful in a wide variety of settings when random assignment is implemented imperfectly.     

Deterrence and the tipping effect

Previous research has focused on certaintyznty of punishment as one. factor. upon which any dzterrent effect of punishment might be contingent and Auggested that it must reach a critical tipping level before the deterrent effect becomes evident. Using data on index crimes and clearance. rates for. Virginia cities and counties, bupport was found for both the. deterrent effect and the exlstenee of a tipping level. Attention was directed towand a more precise specification of the form of the tipping effect, suggesting that there is a single critical tipping level and that it occurs at a relatively low level of certainty of punihment.     

Procedural fairness and work group responses to performance evaluation systems

In a variety of settings, procedures that permit predecision input by those affected by the decision in question have been found to have positive effects on fairness judgments, independent of the favorability of the decision. Two major models of the psychology of procedural justice make contrary predictions about whether repeated negative outcomes attenuate such input effects. If such attenuation occurs, it would lessen the applicability of procedural justice findings to some real-world settings, such as organizations, where procedures often provide repeated negative outcomes. The present laboratory investigation examined the procedural and distributive fairness justments produced by high- and low-input performance evaluation procedures under conditions of repeated negative outcomes. Thirty-five three-person groups of male undergraduates participated in a three-round competition. Groups either were or were not allowed to specify the relative weights to be given to two criteria used in evaluating their performance. All groups received negative outcomes on each of the three rounds. A second experimental factor varied whether or not the group learned after losing the second round that it could not possibly win the third and final round of the competition. Measures of procedural and distributive fairness showed that the high-input procedure led to judgments of greater procedural and distributive fairness across all three rounds. The input-based enhancement of fairness occurred regardless of whether reward was possible. The implications of these findings for theories of procedural justice and for applications of procedural justice to organizational settings are discussed.     

Is technology transfer more than a “mechanical” process?

Editorial Advisory Board