I did MICOM and obtained partial invariance for both my demographics: teaching experience.
I split the teaching experience into three groups to do Multigroup analysis
However, I am not able to confirm or reject my hypothesis that teaching experience moderates the relationships of all the constructs with the endogenous variable due to different significance within a group.
The question is:
1. If there is significance for the third group of teaching experience in SI and none for the others, do I generalize that as the moderation being generally significant or not siginificant? (Refer image)
2. should I split the data for teaching experience into two groups instead?
What you propose is common and probably the norm but unethical. I take it you meaningfully formed groups (which has the possible downside of losing information, as raw data are transformed into categories). With that assumption, you should avoid HARKing (hypothesizing after results are known) and p hacking. Not proving your hypothesis is clearly your results. Publishers love positive findings, so researchers often run tests until that happens. Don't do it! Live with your results.
Hi there, i think there is misunderstanding. these are indeed findings. My hypothesis remains. The question is how do i interpret this data, to find if my data supports my hypothesis or not. If it doesn’t, I would state so. I need help interpreting it thats all. Three category is trickier then 2 categories. 2 categories in MGA is more straightforward in terms of hypothesis testing.
read this article Article Multigroup analysis of more than two groups in PLS-SEM: A re...
Tq but i have, it doesnt have criteria on accepting or rejecting a hypothesis of a moderator analysed using MGA with 3 groups. Most articles identify this as a literature gap in terms of MGA (esp NDT NPT)
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India