|
|
 |
| Higher level constructs and MFA |
 
|
| Message/Author |
|
|
|
|
I have a question regarding constructs that are conceptualized at a level higher than the individual.
Klein and Kozlowski (ORM, 2000) identify three kinds of constructs, namely global, shared and configural. I am particularly interested in the implications of shared and configural constructs given that they are conceptualized to generate at the individual level, but exist at a higher level.
Certain higher-level constructs, like organizational climate, do have an individual level counterpart (i.e. psychological climate). In this case the two constructs are homologous, and the theoretical structure of the construct should be the same at both level. In a MFA setting this would imply imposing the same restrictions on both the within-level and between-level matrix.
However other constructs, like for instance “relational social capital” or “team efficacy”, may only exist at a higher-level, while still originating at an individual level. In other words, relational social capital may not be the sum of individual social capital, much like team-efficacy may not be the sum of individual self-efficacy.
In these cases, how should one analyse the theoretical structure of the constructs in a MFA framework? Specifically, how should one model the within-level matrix?
So far I haven’t found any definite answer and would love to hear your thoughts on this issue.
Thanks u! |
|
|
|
|
Perhaps this paper is of interest to you:
Lüdtke, O., Marsh, H.W., Robitzsch, A., Trautwein, U., Asparouhov, T., & Muthén, B. (2008). The multilevel latent covariate model: A new, more reliable approach to group-level effects in contextual studies. Psychological Methods, 13, 203-229. |
|
|
|
|
| Thank you Bengt. The paper you suggested was extremely useful! |
|
|
| Back to top |
|
|
