Derivation Of A Response Inconsistency Scale for the Matrix-Predictive Uniform Law Enforcement Selection Evaluation Inventory

Abstract

The Matrix- Predictive Uniform Law Enforcement Selection Evaluation Inventory (M-PULSE) is an actuarial instrument used to predict job performance and liability risk of police officer candidates. Leark and colleagues created the INC Scale to detect inconsistent responding on the M-PULSE. This study examined the ability of the INC Scale to detect inconsistent responding to items on the M-PULSE. This study utilized archival data provided by MHS, Inc., publishers of the M-PULSE. A random sample of 3,392 from the M-PULSE's normative full sample was used as the control group and the experimental group consisted of 500 randomly assigned subjects. An independent t-test was conducted to determine if there were statistically significant INC Scale mean score differences between the control group and the experimental group. Post-hoc analyses were conducted to determine internal reliability and validity of items on the INC Scale. Mean scores for the INC Scale were statistically significantly higher than mean scores for the normative data set indicating that the INC Scale was able to correctly identify inconsistent responses (t= 86.967, df =3890, p < .01). This study indicates that the INC Scale is able to detect inconsistent responding to items on the M-PULSE, thereby improving the validity of the M-PULSE.