Criminal Behaviour and Self-Report

Self-report scales play a large role in psychological measurement. Self-report scales are used because they are convenient and, for the most part, thought to be able to be developed in such a way that they produce satisfactorily reliable and valid interpretations of data. As a third year psychology student, I have been exposed to countless self-report scales. Want to know how extraverted you are? There’s a self-report for that. Your likeliness to sexually harass? There’s a self-report for that, too. How do we determine baseline motivation levels for clients entering therapy? Give them self-reports, of course!

Self-report scales have broad applications, and because psychology deals with intimate thought processes, emotions and behaviours, are often the go-to test type in the domain. Some applications of data obtained through self-report are obviously higher stake than others; it isn’t hard to imagine how a self-report scale administered as part of a job application process may have more important outcomes than a “How Slytherin Are You” BuzzFeed quiz (available here for anyone who’s interested). I do not doubt that self-report scales are generally good enough for the research purposes that they serve, but do worry about how useful this type of psychological test is when high cost decisions come into play.

Just in my own past experience, almost every, if not every study that I have participated in here at McGill includes some kind of self-report component. I have often found self report measures to be challenging; I always feel like I am able to justify more than on answer — do I like being the centre of attention sometimes? Rarely? Often? My answers depend on the situations that are accessible in my mind at the time of reporting. Because even I struggle to accurately represent myself on tasks whose outcomes have no effect on my life, I wonder how accurate self-report scales completed by those in high-stake situations really are.

In forensic psychology, for example, personality self-report questionnaires are frequently used to detect psychopathology and assess adaptability to life in prison. It is easy to understand why subjects may want themselves to appear a certain way in this context. Luckily, researchers have a solution — validity scales will detect deliberately faked personality tests! Or will they…? A study done by Medina, Cebolla, Banoes and Botella (2012) suggests that participants asked to “fake” their results on personality scales were able to, undetected. Validity scales in this experiment indicated that 90.1% of fake good participants were sincere, proving ineffectiveness of the validity scales in detecting malingered data.

This study does have some limitations. The results cannot readily be generalized to all populations (the sample was exclusively male and scales were given in Spanish), but these downfalls do not take away from profound implication of these findings: individuals can, (intentionally or unintentionally) fake self-report data. This study also suggests that validity scales may just not be enough for detecting liars based on data. What’s worse is that clinical populations often have cognitive deficiencies that reduce ability to self-monitor (e.g. compromised executive function, extreme scores on self-esteem measures, etc.). For all of these reasons, I fear that when it comes to high-cost decisions, self-report measures are simply not good enough.

Reference:

Medina, P., Cebolla, A., Botella, C., & Banos, R. (September 01, 2012). Are validity scales useful for detecting deliberately faked personality tests? A study in incarcerated populations. Journal of Forensic Psychiatry and Psychology, 23, 4, 452-465.