Predicting Job Performance

Predicting Job Performance

There are two common criticisms of the use of profiling tools for recruitment. The first is that ‘personality’ measures are poor predictors of job performance. The second criticism is that job applicants distort (‘fake’) their scores for reasons of ‘social desirability’.

There are two common criticisms of the use of profiling tools for recruitment. The first is that ‘personality’ measures are poor predictors of job performance. This criticism still persists in some areas despite the overwhelming body of evidence showing that well-constructed and validated profiling tools do reliably predict job performance. 

For example, research has demonstrated that behavioural assessment contributes unique information to the prediction of job performance, over and above that offered by methods such as cognitive ability testing and managerial assessment centres (Sackett, Schmitt, Ellingson & Kabin, 2001; Goffin, Rothstein, & Johnston, 1996). 

The second criticism is that job applicants distort (‘fake’) their scores for reasons of ‘social desirability’. While some ‘faking’ is devious and intentional, most ‘faking’ is motivated by a desire to look good “in front of others”. This means that, during social interaction, most people behave in ways that are intended to convey a positive impression of themselves. They do this whether reacting to questions in an employment interview or to items on a profiling inventory. 

PRISM does not take sides in the debate between the relative merits of ‘ipsative’ and ‘normative’ instrument: it provides outputs based on both measurements.

The belief that candidates can ‘fake’ the results of profiling tools is probably the single greatest concern employers have about using such tools during the recruitment process. These concerns are understandable. Job applicants are motivated to ‘pass the test’ and come closer to a job offer. This should not come as a surprise to anyone.

However, most applicants do not know precisely which characteristics are desirable for the position, and respondents who intentionally distort their responses can end up pushing themselves outside of the desirable range for the role. The best policy for respondents is to respond honestly, and most good measures clearly indicate this in the instrument’s instructions. 

Even so, for those who continue to distort their responses, good instruments have a built in ‘faking’ scale or response distortion scale (SD scale). Typically, these scales identify individuals who respond in a way that is designed to make them look very favourable. Response distortion scales can help users determine whether an accurate judgment can be made based on the candidate’s responses.

Leave a Comment

Your email address will not be published. Required fields are marked *