Hi there, I am working on my dissertation and am using SEM for the first time. I collected data from two different sources (a college student sample & a sample collected from people on substance use related subreddits). I ran some t-tests to check for group differences between the two samples and there are several significant differences both demographically and on dependent variables. Would I be able to combine these two groups into one dataset for SEM?
My idea behind collecting from two different sample sources was that I wanted greater variance within my dataset (especially for my substance use variable) and so, ideally, it would be great to combine them. However, I don't know if that's allowed in the world of SEM.
Thanks so much for your help!
If those samples represent distinct populations and/or if they differ significantly with respect to certain variables, then they are better analyzed separately in SEM, or -- even better -- by means of multigroup SEM. In multigroup SEM, the samples are analyzed simultaneously in one analysis, but parameters are estimated separately for each sample/group. You can then compare the parameter estimates across groups/samples and even test formally whether some or all of the parameter estimates differ significantly across groups using model comparisons and nested model chi-square difference tests. We call this testing of measurement invariance/equivalence and structural invariance/equivalence across groups.
When a sufficient level of invariance/equivalence can be established across groups, the samples could be combined into one big sample (because in that case, they do not differ in relevant ways). Otherwise, the unique parameter values/results should be reported separately for each group.
I recommend that you check out the literature on multigroup CFA/SEM as well as the literature on measurement equivalence/invariance testing for further information.
Christian Geiser Thank you so much- this is very helpful! I'll definitely check out that literature.
See these links, please:
https://www.researchgate.net/post/What-is-the-minimum-sample-size-needed-to-conduct-the-SEM-analysis-utilizing-AMOS-software
https://www.sciencedirect.com/topics/biochemistry-genetics-and-molecular-biology/structural-equation-modeling
Article “How Well Does Your Structural Equation Model Fit Your Data?...
- Badges
- Science topic
- Similar topics
- Psychology
- Emotion