In a structural modeling, there is one independent variable (X) and two dependent variables(Y1 & Y2). While doing correlation analysis in SPSS, X is found to be significantly correlated with Y1, but it has no significant correlation with Y2. While performing SEM in AMOS taking the same variables, the results show significant causality at 5% (i.e., X significantly predicts both Y1 & Y2 ). Kindly clarify how can be insignificantly correlated variables lead to significant causality in SEM (Sample Size=120)
Using different types of software should not give a different results. Therefore, it is most likely you are testing two different models that you might not even realize it.
Please double check to ensure that the models in SPSS and AMOS are the same.
If the problem still can't be solved, please give us more details.
Thank you so much, Stanley..
I am sharing more details ,particularly wrt to the one showing insignificant correlation.
The variable Y2 is a dependent variable in the model. It consists of five items
{ Y2(i), Y2(ii)), Y2 (iii), Y2(iv) and Y2 (v)}.
The variable X is an independent variable comprising of three factors (Xa=4 items, Xb=4 items, and Xc=2 items )
In SPSS, on correlating a total items composite score of Y2 with Xa, Xb and Xc, none of the correlation are significant. While on correlating separate items of Y2 with Xa, Xb and Xc; two correlations are found to be significant at 5% (Y2(ii) Xa and Y2(iii) Xb).
On running the causality relationship in AMOS (taking the latent structure into account), the causal relationship between X and Y2 is significant at 5%.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size