Hi there,
I'm conducting research testing the UTAUT2 technology acceptance model (Venkatesh et al. 2012) in a smartwatch context. However, I want to use age as a moderator.
The UTAUT2 model is large and I would probably be splitting the sample into 3 different age ranges/groups.
My question is, based on this information, which approach should I use: Partial Least Squares or AMOS? My understanding is that if I am to use AMOS + age as a moderator I need a very large sample size.
I am expecting valid 200 - 300 respondents, so this restriction needs to be considered.
Thanks in advanced for any help!
If you plan to split data into three, it is better to use PLS-PM. However, you will still need to consult with power analysis.
By the way, splitting data is considered as conditioning data. It is not to test a moderator. For this concept, please refer to causality theory written by Judea Pearl.
Ideally, when you test an established model (in your case, UTAUT2 technology acceptance model (Venkatesh et al. 2012)) in a new context, you are to use CB-SEM, not the PLS-SEM. Secondly, choosing the software package is up to you and AMOS would be fine. Thirdly, "very large sample" is a subjective issue. That is to say, how large is "very large"? In fact, the issue of sample size is not an independent decision, rather it all depends on the complexity of the model, number of parameters, distribution of the sample and so on. In some cases, a sample of 300 is large enough specially when your model is simple one. However, when you do multi-group analysis (moderator variable), your total sample will be split into small groups that may require a relatively large sample, even if you have a simple model.
Hope this helps.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Materials
- Inorganic Materials
- Ceramics