MGA vs two different analysis
Posted: Wed Jan 22, 2020 9:48 pm
Hello,
I am new to the forum :-) Thank you for your efforts :)
I was working on a paper to test a model in two different context (different categories A and B). So I was comparing two categories with different observation A (n=141) B (n=247) and I did run them separately: Model A then Model B.
When I run the MGA with the two categories I found different coefficient then the ones I found previously and was surprised. I was thinking that I would find the same coefficients wether I put two different models or I run the MGA.
Any explanations please?
I am new to the forum :-) Thank you for your efforts :)
I was working on a paper to test a model in two different context (different categories A and B). So I was comparing two categories with different observation A (n=141) B (n=247) and I did run them separately: Model A then Model B.
When I run the MGA with the two categories I found different coefficient then the ones I found previously and was surprised. I was thinking that I would find the same coefficients wether I put two different models or I run the MGA.
Any explanations please?