I don’t mean this to sound cynical, but with all of the research on faking I wonder how we can really determine if someone is being honest in responding to questions on a measure? I can think of things we could do to examine honesty - like test-retest reliability, compare self-report with other-report assessments, etc. But if there are discrepancies between these times and sources, does that really mean that a person is not honest? I think in terms of faking and honesty, we are relying very heavily on the idea that an individual can accurately introspect, reflect, and self-report about his or her own personality. The concerns arise with the use of outside sources. Even behavioral assessments rely on some sort of outside judgment. With all this being said, and beside the problem of construct-related validity, should we really be so concerned with faking? Or just be happy that some individuals may be able to self-monitor or socially adjust enough to provide the responses we are looking for – and then hope they will do the same when on the job?
I was surprised at how little the authors discussed stereotypes in this study. It seems to me (and as the point out very briefly in the discussion) that the people in the "fake accountant manipulation" were assessing stereotypes about accountants. For this reason, I believe it was especially problematic using college students in this study. If a job posting for an accountant was posted we would assume that most of the applicants for the job had some experience with being an accountant (e.g., real work experience, majored in business-accounting in college) so they may consider themselves to be part of the "accountant group" and may not need to rely on stereotypes to determine what a "good" way to answer the questions would be (or at least they would have a more accurate idea of the stereoypes of there group). However, when you use college students, they may need to completely rely on this and end up worse at faking than real accountants would be. Did you think this was an issue as well?
•The SIOP 2004 article had a panel that came to the decision that faking was inevitable and not as big of a deal as expected in harming the validity of the personality tests. It is interesting how this study mentions that their results show the difference between field and lab faking results. How does this benefit us in the selection field? Should we continue to investigate the consequences of faking after the current research shows no strong negative results? I agree with the SIOP panel and Vicki. Also, that people can make mistakes, misunderstand, and overestimate on our own values and behaviors. A more valid and reliable measure would be to include measures of our own personality from close friends , family, and co-workers and then to compare.
I like their conclusion that it is possible that people who know how to fake the demands of the job might actually perform better on the job than those who don’t know how to fake. It made me think of fake it, until you make it approach to work. I think people will find what work is being monitored and only perform those tasks; meanwhile, they neglect their other duties of the job because they are under qualified. Anecdotal evidence to this point, the woman who had my HR job after me, was filing nothing and keeping track of zero paperwork. She had stuffed employee requests in her desk and did not send anything to corporate. Could faking of personality measures lead to under qualified employees?
I was really surprised that self-monitoring did not contribute to our understanding of whether or not people fake on assessments of personality. Do you think this is actually so, or just an issue specific to the way the operationalized the construct/design of their study?
This article begins by stating, “Personality variables have been shown to predict a variety of work behaviors and outcomes with modest validity and low levels of adverse impact (e.g., Barrick, Mount, & Judge, 2001).” Yet, Murphy and Dzieweczynski (2005) described this same study as finding that the highest mean validity for any personality trait was .12 for conscientiousness and that “after highly generous corrections for range restriction and unreliability, the estimated population validity for four of the Big Five factors was smaller than .20.” Is anyone else puzzled and concerned by Raymark and Tafero’s assertion of “modest validity” without reporting the actual validities found by Barrick et al.?”
I don’t mean this to sound cynical, but with all of the research on faking I wonder how we can really determine if someone is being honest in responding to questions on a measure? I can think of things we could do to examine honesty - like test-retest reliability, compare self-report with other-report assessments, etc. But if there are discrepancies between these times and sources, does that really mean that a person is not honest? I think in terms of faking and honesty, we are relying very heavily on the idea that an individual can accurately introspect, reflect, and self-report about his or her own personality. The concerns arise with the use of outside sources. Even behavioral assessments rely on some sort of outside judgment. With all this being said, and beside the problem of construct-related validity, should we really be so concerned with faking? Or just be happy that some individuals may be able to self-monitor or socially adjust enough to provide the responses we are looking for – and then hope they will do the same when on the job?
ReplyDeleteI was surprised at how little the authors discussed stereotypes in this study. It seems to me (and as the point out very briefly in the discussion) that the people in the "fake accountant manipulation" were assessing stereotypes about accountants. For this reason, I believe it was especially problematic using college students in this study. If a job posting for an accountant was posted we would assume that most of the applicants for the job had some experience with being an accountant (e.g., real work experience, majored in business-accounting in college) so they may consider themselves to be part of the "accountant group" and may not need to rely on stereotypes to determine what a "good" way to answer the questions would be (or at least they would have a more accurate idea of the stereoypes of there group). However, when you use college students, they may need to completely rely on this and end up worse at faking than real accountants would be. Did you think this was an issue as well?
ReplyDelete•The SIOP 2004 article had a panel that came to the decision that faking was inevitable and not as big of a deal as expected in harming the validity of the personality tests. It is interesting how this study mentions that their results show the difference between field and lab faking results. How does this benefit us in the selection field? Should we continue to investigate the consequences of faking after the current research shows no strong negative results? I agree with the SIOP panel and Vicki. Also, that people can make mistakes, misunderstand, and overestimate on our own values and behaviors. A more valid and reliable measure would be to include measures of our own personality from close friends , family, and co-workers and then to compare.
ReplyDeleteI like their conclusion that it is possible that people who know how to fake the demands of the job might actually perform better on the job than those who don’t know how to fake. It made me think of fake it, until you make it approach to work. I think people will find what work is being monitored and only perform those tasks; meanwhile, they neglect their other duties of the job because they are under qualified. Anecdotal evidence to this point, the woman who had my HR job after me, was filing nothing and keeping track of zero paperwork. She had stuffed employee requests in her desk and did not send anything to corporate. Could faking of personality measures lead to under qualified employees?
ReplyDeleteI was really surprised that self-monitoring did not contribute to our understanding of whether or not people fake on assessments of personality. Do you think this is actually so, or just an issue specific to the way the operationalized the construct/design of their study?
ReplyDeleteThis article begins by stating, “Personality variables have been shown to predict a variety of work behaviors and outcomes with modest validity and low levels of adverse impact (e.g., Barrick, Mount, & Judge, 2001).” Yet, Murphy and Dzieweczynski (2005) described this same study as finding that the highest mean validity for any personality trait was .12 for conscientiousness and that “after highly generous corrections for range restriction and unreliability, the estimated population validity for four of the Big Five factors was smaller than .20.” Is anyone else puzzled and concerned by Raymark and Tafero’s assertion of “modest validity” without reporting the actual validities found by Barrick et al.?”
ReplyDelete