MIMIC Models

[spoeko]

G´day everyone,

I'm trying to measure a construct (good stadium atmosphere) that is by definition conceptualized as being a specific combination of certain environmental stimuli (as formative indicators) and specific sensations of people (as reflective indicators).

Only the consideration of BOTH things the stimuli (as causes) AND the sensations (as effects) captures what is defined as the constructs content. High values of the construct demand the appearance of the stimuli and the sensations.

Thus, the isolated usage of formative or reflective indicators is not sufficient to cover the construct.
My question is, if a MIMIC-Model accounts for that?

Thanks very much
Sebastian

[schroer]

Hello,

yes, your scenario sounds very much so. As you probably know, you can model MIMIC models by a) "splitting up" your latent and then b) let your formative latent predict the reflective one. This is mathematically equivalent to using a MIMIC model, a good reference on this is Wold (1982).

Best wishes and good luck,

Joachim

[jlengler]

Dear Joachim,

I have the very same situation. My dependent variable in the structural model has 3 reflective indicators and 2 formative. Thus, I splitted up this latent and specified 1 latent with the formative indicators predicting the latent with the reflective indicators, just as you mentioned on your previous message. My question is whether this is enough to satisfy the condition of identification argued by Diamantopoulos, Riefler and Roth (2008) or not.

Cheers,

Dr. Jorge Lengler