Murphy identifies one important aspect of organizational climate: level of restrictiveness, a macro-level organizational variable, as an important element in the measurement (or the interpretation of the measurement results) of performance - a micro-level organizational variable. This interconnectedness of the various levels of analysis within organizations makes salient the necessity of moving between a ground level view of the construct(s) we are interested in and an aerial view of the construct(s) we are interested in so that we can better grasp the various relationships. What other macro-level variables might impact performance measurement and interpretation?
Murphy posits that range restriction in individual scores on performance measures contributes to the apparently high level of variability in VG for certain tests. What organizational factor may contribute to how much range restriction there is in performance measures? That is what organizational variables might increase eor decrese the amount of range resriction on performance measures?
To me, it seems that validity generalization would be most helpful in guiding an organization toward a wide variety of predictor measures that they could select from. However, if an organization lacks the ability to conduct a local validity study, they may use validity generalization evidence as support for the validity of their selection system. If challenged, can an organization legally support their selection system if the only evidence of validity is validity generalization?
Murphy (2000) posits validity generalization and situational specificity to be two separate ideas that are examined independently. James et al. (1992) posits that they should be examined together. Which view is more widely accepted/correct?
• It seems too simple of VG inferences where there is a lack of validity to be solely based on errors of artifacts (e.g., range restriction, etc.). Why are moderators and situational factors not strongly considered within the VG model. We may benefit from these models if we merge them. Social psychology has taught us that many times things are based on the situation and context at hand. Situational specificity is important to understanding behavior of employees, as it strongly does in the selection function.
Murphy seems to conclude that in spite of all the debate, validity generalization analyses empirically affirms the value and consistency of professionally developed selection tests and methods for practical hiring decisions. This was published in 2000. Is the debate still a heated topic today?
Isn't there a danger here in trying to make our selection procedures and measures in such a standardized fashion? I feel like selection programs could be turned into an assembly line where organizations get an off the shelf selection system based upon the positions in the organization. Would this be a detriment to our field? To the organization?
I am reluctant to believe that situation does not play as important a role in validity generalization as Murphy argues. I would have liked to have known more information about the variables of interest. I think that there are variables that generalize well across all types of situations such as IQ, but I would be more careful with use of personality, for example, when it comes to arguing that validity generalization is not influenced much by the situation.
Murphy identifies one important aspect of organizational climate: level of restrictiveness, a macro-level organizational variable, as an important element in the measurement (or the interpretation of the measurement results) of performance - a micro-level organizational variable. This interconnectedness of the various levels of analysis within organizations makes salient the necessity of moving between a ground level view of the construct(s) we are interested in and an aerial view of the construct(s) we are interested in so that we can better grasp the various relationships. What other macro-level variables might impact performance measurement and interpretation?
ReplyDeleteMurphy posits that range restriction in individual scores on performance measures contributes to the apparently high level of variability in VG for certain tests. What organizational factor may contribute to how much range restriction there is in performance measures? That is what organizational variables might increase eor decrese the amount of range resriction on performance measures?
ReplyDeleteTo me, it seems that validity generalization would be most helpful in guiding an organization toward a wide variety of predictor measures that they could select from. However, if an organization lacks the ability to conduct a local validity study, they may use validity generalization evidence as support for the validity of their selection system. If challenged, can an organization legally support their selection system if the only evidence of validity is validity generalization?
ReplyDeleteMurphy (2000) posits validity generalization and situational specificity to be two separate ideas that are examined independently. James et al. (1992) posits that they should be examined together. Which view is more widely accepted/correct?
ReplyDelete• It seems too simple of VG inferences where there is a lack of validity to be solely based on errors of artifacts (e.g., range restriction, etc.). Why are moderators and situational factors not strongly considered within the VG model. We may benefit from these models if we merge them. Social psychology has taught us that many times things are based on the situation and context at hand. Situational specificity is important to understanding behavior of employees, as it strongly does in the selection function.
ReplyDeleteMurphy seems to conclude that in spite of all the debate, validity generalization analyses empirically affirms the value and consistency of professionally developed selection tests and methods for practical hiring decisions. This was published in 2000. Is the debate still a heated topic today?
ReplyDeleteIsn't there a danger here in trying to make our selection procedures and measures in such a standardized fashion? I feel like selection programs could be turned into an assembly line where organizations get an off the shelf selection system based upon the positions in the organization. Would this be a detriment to our field? To the organization?
ReplyDeleteI am reluctant to believe that situation does not play as important a role in validity generalization as Murphy argues. I would have liked to have known more information about the variables of interest. I think that there are variables that generalize well across all types of situations such as IQ, but I would be more careful with use of personality, for example, when it comes to arguing that validity generalization is not influenced much by the situation.
ReplyDelete