• Often management and executive managers have control of and create the organizational climate, thus whether the climate is restrictive (i.e., Theory X) or not (i.e., Theory Y). New CEOs and executive management is often replaced after 3-5 years, except family owned businesses so much. Would this not have an effect on the implications this article has on whether the climate is restrictive or not? It would seem that getting a new CEO who possesses a theory Y mentality of employee performance may provide better validity measures within selection.
It seems that a major challenge of validity generalization and situational specificity is to correctly partition out the variance into that which is due to statistical artifacts and that which is due to the situation. The authors imply that the future of validity generalization requires more of an emphasis on accounting for situational influences; however, they don’t provide much guidance, other than restrictive climate, about what these situational influences are/might look like. How can we identify and then account for potential situational artifacts when creating personnel selection systems?
James et al. are skeptical of the VG analysis assumption that artifactual error variance and error variance due to situational variables are uncorrelated. They propose a large scale study to examine if this assumption is generally reasonable. In the meantime, should practitioners do as Murphy (2000) advises and examine the validity studies on which the VG analysis is based for contextual similarity?
The only situational variable that the authors discuss is the restrictive climate. What other situational variables might influence how generalizable a test is?
I think these authors make a good case that situational variables should be included in a generalizability analysis. They convinced me that the climate differences could have real impact on restricting variations in performance apart from differences in statistical artifacts like sample size or range restriction or even reliabilities of criterion or predictors. Where are we now with this?
I thought the authors made good points about the need for situational/contextual variables to fully understand validity generalization. These author's only suggest restrictiveness as a situational variable. However, I believe there are many situational variables one could use here, some would be more relevant depending upon the industry. Would it be difficult to capture all of the various relevant situational variables that could be included in these models? What are some others besides restrictiveness? How would multilevel modeling help here? Would that help capture the context of the organization?
How might the opposite of restrictive climate (a “non-restrictive climate”) influence the future of VG analysis? Such a climate might be seen in more modern, google-like climates, which emphasize employee creativity. This also might be the case since the Dot Com era of workplaces (this article is from 1992 right before the age of internet).
• Often management and executive managers have control of and create the organizational climate, thus whether the climate is restrictive (i.e., Theory X) or not (i.e., Theory Y). New CEOs and executive management is often replaced after 3-5 years, except family owned businesses so much. Would this not have an effect on the implications this article has on whether the climate is restrictive or not? It would seem that getting a new CEO who possesses a theory Y mentality of employee performance may provide better validity measures within selection.
ReplyDeleteIt seems that a major challenge of validity generalization and situational specificity is to correctly partition out the variance into that which is due to statistical artifacts and that which is due to the situation. The authors imply that the future of validity generalization requires more of an emphasis on accounting for situational influences; however, they don’t provide much guidance, other than restrictive climate, about what these situational influences are/might look like. How can we identify and then account for potential situational artifacts when creating personnel selection systems?
ReplyDeleteJames et al. are skeptical of the VG analysis assumption that artifactual error variance and error variance due to situational variables are uncorrelated. They propose a large scale study to examine if this assumption is generally reasonable. In the meantime, should practitioners do as Murphy (2000) advises and examine the validity studies on which the VG analysis is based for contextual similarity?
ReplyDeleteThe only situational variable that the authors discuss is the restrictive climate. What other situational variables might influence how generalizable a test is?
ReplyDeleteI think these authors make a good case that situational variables should be included in a generalizability analysis. They convinced me that the climate differences could have real impact on restricting variations in performance apart from differences in statistical artifacts like sample size or range restriction or even reliabilities of criterion or predictors.
ReplyDeleteWhere are we now with this?
I thought the authors made good points about the need for situational/contextual variables to fully understand validity generalization. These author's only suggest restrictiveness as a situational variable. However, I believe there are many situational variables one could use here, some would be more relevant depending upon the industry. Would it be difficult to capture all of the various relevant situational variables that could be included in these models? What are some others besides restrictiveness? How would multilevel modeling help here? Would that help capture the context of the organization?
ReplyDeleteHow might the opposite of restrictive climate (a “non-restrictive climate”) influence the future of VG analysis? Such a climate might be seen in more modern, google-like climates, which emphasize employee creativity. This also might be the case since the Dot Com era of workplaces (this article is from 1992 right before the age of internet).
ReplyDelete