Faking It.

Posted by Ricki Sharpe on May 14, 2007  
Filed Under Selection

Faking personality based employment inventories can be a difficult problem for the practitioner. It is not made any easier when some test publishers discount faking as nothing more than a normal personality trait, and that practitioners should expect job applicants to put their best foot forward. However, that does not explain what a practitioner should do when an applicant puts both feet forward.

In an attempt to compensate for faking, the more reputable test publishers have in-built lie detectors in their inventories called social desirability. Others have introduced a different format, the ipsative or forced-choice format, to overcome the problem. While some publishers flaunt the supremacy of ipsative tests over normative tests, its uptake is variable. Much controversy still overshadows the ipsative-normative debate.

And, if you believe that job applicants would not lie to get a job, then you must also believe in fairies at the bottom of the garden. If 85% of resumes contain untruths, you can believe that some applicants will try to fake their responses to a personality inventory. Books have even been written on how to Ace the Corporate Personality Test.

Not only does evidence exist that applicants fake their responses to personality based employment inventories, now there is evidence that applicant faking behaviour results in significant rank ordering changes that impact hiring decisions.

Researchers, Griffith, R.L., Chmielowski, T., & Yoshita, Y. (2007). Do applicants fake? An examination of the frequency of applicant faking. Personnel Review, 36, 341-355, used a within subjects design to assess whether applicants elevated their scores in an applicant condition. Subjects who applied for a job were later contacted and asked to complete the same personality measure under an honest instructional set. The within subjects design allowed the researcher to examine faking behaviour at the individual level of analysis rather than draw inferences between applicant and incumbent groups.

Results suggest that a significant number of applicants do fake personality based selection measures. Depending on the confidence interval used, between 30 and 50 percent of applicants elevated their scores when applying for a job. The results also show that applicant faking behaviour resulted in significant rank ordering changes that impacted hiring decisions.

Empirical research has previously supported the notion that respondents can fake when instructed, however, other research has suggested that applicants do not fake in applied settings. This study is the first to provide substantial evidence that faking does occur in applicant settings and that it disrupts rank ordering of applicants.

OK, so we know faking exists, but what does a practitioner do when faced with a candidate that has scored high on a social desirability scale? For the risk averse, the easy way out is to reject the candidate out of hand. After all, If you screen out a good candidate, you will never know about it. But, if you screen in a bad candidate, you will certainly know about it.

The practical implications of this paper are that personality measures should not be used alone. Rather they should be included in a set of measures that are less susceptible to faking behaviour, such as interviews, assessments centres and reference checking. In addition, applied researchers must continue research efforts to address the faking issue.


Print This Post Print This Post

Similar Posts

Comments

Leave a Reply