Hello all,
I am analyzing a model using SEM, but my result seems hard to interpret the effect between non-formal learning experience(X), volition(Y), and expectancy(M) in the context of beginning farming.
1. indirect effect is positively significant (coef=0.172, pM(coef=0.155, pY(coef=1.106, pY) is negatively significant(coef=-0.116, p
@Seong Ji In your case, there seem to be the problem of multicollinearity among the variables as you've mentioned that the correlations among Non-Formal Learning (X1) Expectancy (M), and Volition (Y) are positive when checked but when you run SEM model with Informal Learning (X1), the correlation between Non-Formal Learning (X1) and Volition becomes negative. Can you please check whether there are any cross loadings of items from X1, X1, and Y variables. Moreover, your path coefficient between M and Y is also above 1 (Heywood Case) and indicating towards multicollinearity among the variables.
Refer to this article to resolve the case.
Deegan, J.Jr. (1978). On the Occurrence of Standardized Regression Coefficients Greater Than One. Educational and Psychological Measurement, 38, No. 4, 873-888.
Imran Anwar Thank you. In terms of path coefficient, It refers to regression coefficient, not beta(standardized regression coefficients=.949 in this case). Can it also be a problem?
When I implemented EFA and CFA of each variable it appears adequate but I will double-check with the cross roadings. Thank you!
@Seong Ji If standardized coefficient is .949 then it seems to fine but on a bit higher side given the correlation between M and Y is .668. Moreover, look for VIF for multicollinearity among your IV, M, and DV to check if multicollinearity exist there as you have already taken care of cross loadings during EFA and CFA.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Structural Equation Modeling