I’m not sure I totally understand what the authors were getting at in this article, but it seems as if they were trying to statically support the notion that current meta-analytic techniques may be underestimating the value of personality as a predictor of job performance. While I never thought at the idea of a negative validity, it makes conceptual sense that some predictors may be negatively related to job performance in some contexts. Do other theorists support the idea of bidirectionality in personality research? How might we account for this in personality predictors?
This was an insightful article where personality measures are described as underestimated in many studies. How has this article influenced the statistical analyses used to measure personality for job performance?
What if during validation of a selection procedure we discover that conscientiousness is negatively related to performance. Do you think an organization would be comfortable using this as a predictor in that they would want to selected people low in conscientiousness?
Wow, what a complicated article! I understood about 30% of what they were talking about in terms of the meta-analysis. However, I did understand the part about bidirectionality, which seems to make intuitive sense. I think it would also be interesting to do a within subjects design and measure how workers’ personalities change depending upon the job they are given. I was thinking in terms of the Standford Prison Experiment, when randomly giving the participants “guard” jobs changed their behavior dramatically. Would this prove personality predictors useless, because many people can adapt their personality to fit the job?
I’m starting to think that personality meta analyses are in appropriate because of criteria issues and sloppy use of personality in studies (that is, the dust bowl approach of throwing variables into an analysis just to see what turns out to be significant). How might we deal with issue in research? Do we need to be more selective in the studies that we use in meta analyses on personality rather than just transform values into absolute values?
The concern that meta-analysis techniques underestimate predictive validities of personality variables for job performance due to the bi-directionality of the correlations between personality and specific types of jobs in particular contexts seems valid. I wonder, however, if doubling even some of the highest correlations substantially increases the amount of variance accounted for by personality in most jobs?
I’m not sure I totally understand what the authors were getting at in this article, but it seems as if they were trying to statically support the notion that current meta-analytic techniques may be underestimating the value of personality as a predictor of job performance. While I never thought at the idea of a negative validity, it makes conceptual sense that some predictors may be negatively related to job performance in some contexts. Do other theorists support the idea of bidirectionality in personality research? How might we account for this in personality predictors?
ReplyDeleteThis was an insightful article where personality measures are described as underestimated in many studies. How has this article influenced the statistical analyses used to measure personality for job performance?
ReplyDeleteWhat if during validation of a selection procedure we discover that conscientiousness is negatively related to performance. Do you think an organization would be comfortable using this as a predictor in that they would want to selected people low in conscientiousness?
ReplyDeleteWow, what a complicated article! I understood about 30% of what they were talking about in terms of the meta-analysis. However, I did understand the part about bidirectionality, which seems to make intuitive sense. I think it would also be interesting to do a within subjects design and measure how workers’ personalities change depending upon the job they are given. I was thinking in terms of the Standford Prison Experiment, when randomly giving the participants “guard” jobs changed their behavior dramatically. Would this prove personality predictors useless, because many people can adapt their personality to fit the job?
ReplyDeleteI’m starting to think that personality meta analyses are in appropriate because of criteria issues and sloppy use of personality in studies (that is, the dust bowl approach of throwing variables into an analysis just to see what turns out to be significant). How might we deal with issue in research? Do we need to be more selective in the studies that we use in meta analyses on personality rather than just transform values into absolute values?
ReplyDeleteThe concern that meta-analysis techniques underestimate predictive validities of personality variables for job performance due to the bi-directionality of the correlations between personality and specific types of jobs in particular contexts seems valid. I wonder, however, if doubling even some of the highest correlations substantially increases the amount of variance accounted for by personality in most jobs?
ReplyDelete