Personnel Selection
& Analytics Lab

Scott B. Morris, Ph.D.
Industrial-Organizational Psychology
Illinois Institute of Technology

The selection and analytics lab explores the application of quantitative methodology in employee selection. Topics include:

  • Adverse impact analysis

  • Meta-analysis methodology

  • Computer adaptive testing

  • Selection system design

  • Bias in subjective hiring decisions

  • Application reactions to artificial intelligence in hiring

Adverse Impact Analysis

Organizational efforts to promote equity and diversity rely on accurate assessment of the disparities produced by employment practices. Adverse impact analyses play an important role in Equal Employment Opportunity (EEO) litigation and compliance, and provide critical information to identify the sources of workplace inequity. The lab explores alternate statistical methods to optimize the utility of adverse impact statistics.

Meta-analysis

One of the biggest challenges in meta-analysis is obtaining effect sizes from diverse research designs in a way that is comparable across studies. The lab explores methods that make full use of the information available in primary studies, while maintaining a common metric despite differences in the original research designs. For example, properly accounting for the dependency of scores in longitudinal designs can enhance the accuracy and precision meta-analytic results.

Item Response Theory

The lab explores applications of psychometric models to improve the precision and efficiency of attitude and trait measures.

Multidimensional computer adaptive tests have the potential to increase measurement precision while reducing the burden on examinees created by lengthy questionnaires.

Differential item functioning (DIF) refers to group differences in how people interpret and respond to test and questionnaire items. By ensuring that measures are free of DIF, researchers can be confident that observed group differences reflect true differenes on the underlying trait and not an artifact of the measure. Group project have examined and extended Raju's Differential Functioning of Items and Tests framework for conducting DIF analysis. We developed software to support DFIT analysis of Likert-type items.