This article builds nice support for the use of personality predictors in selection. I thought it raised some good questions about the conclusions in the Morgeson et al. critique of personality predictors’ proposed psychometric property problems (that’s an awesome alliteration by the way!!! Haha). Although, I don’t understand how these authors can conclude that faking does not ruin criterion-related or construct validity in an applied setting?
I think the main issue between opinions in the Morgeson et al. (2007) and Ones et al. (2007) is the definition of a “substantial validity.” That is, what level does a validity coefficient need to reach to be considered substantial? Must the validity coefficient be high for personality to be considered a useful predictor of job performance? Or could a small validity coefficient suffice, given that it can explain some percentage of variance in job performance? Might we consider using personality, with smaller validity coefficients, as useful when used in combination with other predictors?
•Personality measures seem to help find an applicant who would have a good fit with the organization and job. The article mentioned personality measures’ ability to predict significant work attitudes and organizational behaviors. In addition, observer ratings were discussed as a way to improve the validity. This seems doable for applicants who currently work in the company vs. those outside of the company. How can we apply this approach for external applicants?
Even though personality variables may only account for a small amount of variance in job performance, given that personality factors are relatively stable across a person's life, does this make up for that fact? Might personality tests help to identify glaring issues with personality that might surface later?
Ones mentions a challenge for future research is to find ways of recovering normative scores from traditional forced choice inventories. Is the one of the problems with the Meyer’s Briggs? He recommends use of multiple measures to assess the same trait and possible other ratings (e.g. peer ratings). This seems very costly. I look forward to hearing Roni on this topic!
In a footnote on p. 996, Ones et al. respond to the criticism of Campion in Morgenson et al. (2007) that meta-analysis conducted on low-quality unpublished articles does not change the limitations of those studies by stating that, “Appropriately conducted meat-analyses are not necessarily adversely affected by low quality studies.” This seems to be a contradiction of Murphy’s (2000) admonishment that any VG (a type of meat-analysis) be based on studies of high integrity. Why might Ones et al. be making this assertion?
This article builds nice support for the use of personality predictors in selection. I thought it raised some good questions about the conclusions in the Morgeson et al. critique of personality predictors’ proposed psychometric property problems (that’s an awesome alliteration by the way!!! Haha). Although, I don’t understand how these authors can conclude that faking does not ruin criterion-related or construct validity in an applied setting?
ReplyDeleteI think the main issue between opinions in the Morgeson et al. (2007) and Ones et al. (2007) is the definition of a “substantial validity.” That is, what level does a validity coefficient need to reach to be considered substantial? Must the validity coefficient be high for personality to be considered a useful predictor of job performance? Or could a small validity coefficient suffice, given that it can explain some percentage of variance in job performance? Might we consider using personality, with smaller validity coefficients, as useful when used in combination with other predictors?
ReplyDelete•Personality measures seem to help find an applicant who would have a good fit with the organization and job. The article mentioned personality measures’ ability to predict significant work attitudes and organizational behaviors. In addition, observer ratings were discussed as a way to improve the validity. This seems doable for applicants who currently work in the company vs. those outside of the company. How can we apply this approach for external applicants?
ReplyDeleteEven though personality variables may only account for a small amount of variance in job performance, given that personality factors are relatively stable across a person's life, does this make up for that fact? Might personality tests help to identify glaring issues with personality that might surface later?
ReplyDeleteOnes mentions a challenge for future research is to find ways of recovering normative scores from traditional forced choice inventories. Is the one of the problems with the Meyer’s Briggs? He recommends use of multiple measures to assess the same trait and possible other ratings (e.g. peer ratings). This seems very costly. I look forward to hearing Roni on this topic!
ReplyDeleteIn a footnote on p. 996, Ones et al. respond to the criticism of Campion in Morgenson et al. (2007) that meta-analysis conducted on low-quality unpublished articles does not change the limitations of those studies by stating that, “Appropriately conducted meat-analyses are not necessarily adversely affected by low quality studies.” This seems to be a contradiction of Murphy’s (2000) admonishment that any VG (a type of meat-analysis) be based on studies of high integrity. Why might Ones et al. be making this assertion?
ReplyDeleteAre you conceptualizing conditional reasoning tests to be similar to computer adaptive tests, such as the GRE?
ReplyDelete