Professor of Psychology at University College London
Traditional ways of assessing people have not changed significantly for many years. They include ability and personality tests, 360 degree ratings, interviews and references. The limitations of each method is well known for particular problems with self-report and observer bias. Since the millennium, there has been a great interest in harnessing new technologies to evaluate people more accurately and in more subtle ways. These new ways include gamification, social network analysis, crowd-sourcing, voice profiling, wearables, etc.
This presentation will look the many problems of demonstrating the predictive validity of both old and new methods of people assessment as well as what we know about emerging technologies, including:
Professor and Associate Chair of Psychology at University of South Florida
Personality measures have been shown to predict important educational and workplace outcomes without the adverse impact of cognitive ability tests. Until recently, however, high-stakes applications have been quite limited due to concerns about faking, random responding, and other forms of rater bias on test scores. To address these concerns, research over the last decade has avidly explored the viability of item response theory based forced-choice methods designed to reduce response biases, while improving efficiency through computer adaptive item selection. Research has also explored the benefits of forming composites of narrow-factor measures to improve validity for selection and classification decisions. This presentation will trace the development and implementation of a computerized adaptive multidimensional forced-choice personality assessment system, called TAPAS, that was designed for research and operational decision making in military and civilian personnel screening contexts. It will review the psychometric models, computer adaptive testing algorithm, methods to detect aberrant responding, and validity evidence from research and organizational applications. It will conclude with a discussion of the implications, challenges, and research underway to further improve efficiency and robustness for expanding high-stakes uses.
Check out the conference topics