Measuring Attitudes toward the Insanity Defense in Venirepersons: Refining the IDA-R in the Evaluation of Juror Bias¹

1. Portions of this paper are based on research presented at the 2007 convention of the American Psychological Association in San Francisco, California.

Abstract

The negative attitude against the insanity defense held by venirepersons has a direct effect on individuals with mental illness who are in contact with the criminal justice system. Identifying factors in jurors that support negative attitudes is critical in trying to ensure that the insanity defense is given full consideration when pleaded by people with mental illnesses. This article used the Insanity Defense Attitude scale-Revised (IDA-R; Skeem, Louden, & Evans, 2004) to evaluate attitudes and factors supporting bias in 239 venirepersons. Through the use of confirmatory factor analyses a refined factor model was developed for the IDA-R, which was cross-validated on a sample of 567 undergraduates. Greater support for the death penalty and higher perception of insanity defense use were both related to a more negative attitude toward the insanity defense. Implications for jury selection and the fair treatment of individuals with mental illness are discussed.

The authors are grateful to the helpful suggestions of two anonymous reviewers whose comments improved the quality of the manuscript.

Notes

1. Portions of this paper are based on research presented at the 2007 convention of the American Psychological Association in San Francisco, California.

2. The M'Naghten test or standard is rooted in nineteenth century England and is considered a cognitive test of insanity whereas the defendant must “know their actions are wrong and the immediate (i.e., nature and quality) consequences.

3. For these analyses we only report total score on the IDA-R. Notably, as with the demographic characteristics, there were no differences based on political affiliation for either scale. Please contact the corresponding author for these analyses.

4. We tested for multicolinearity prior to running the regression. Multicolinearity was low and therefore we found no need to control for additional variables in the regression model.