Hello All,
I'm planning to Use a Multiple Cause Multiple Indicator Model in my study of Urban Informal Economy Employment of Ethiopia. To this end, what special considerations does this model have that I must aware of like the assumptions, data requirement, sample size, etc.?
MIMIC models are a version of Structural Equation Modeling, which is based on regression analysis, so the most basic answer is that you need to meet the assumptions for regression. With regard to sample size, one common standard is that you should have at least 5 (preferably 10) observations for every parameter in your model. You can count "parameters" by using a diagram of your model, where every arrow is an effect that you are estimating.
David L Morgan Thank you So Much and I'm glad if U have briefed materials on the Model and attach it for me.
Hi Mebtu
There are technical aspects of the analysis that you will need to consider, as well as theoretical ones. The temporal ordering of the 'causes' and the 'indicators' is a particularly important aspect of the model.
many of these issues are discussed in Diamantopoulos, A. & Temme, D. AMS Rev (2013) 3: 160. https://doi.org/10.1007/s13162-013-0050-0
Mark
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size