Hi all
I have conducted a study looking at different teaching styles used by teachers in two different studies.
I have two Hypothesis:
1. Teaching styles will differ across the WERID (the United Kingdom) and non-WEIRD countries (Zambia)
2. The motivations behind teaching styles will differ across the WERID (the United Kingdom) and non-WEIRD countries (Zambia)
I sent out a survey and with two different questionnaires requiring responses from a 5 point Likert scale. the first was on teaching styles with a total of 5 different teaching styles and they had to rate each one from 1-5 and could use the same rating more than once. the second included 9 statements on motivations for teaching styles that they were asked to rate from 1-5 and again could use the same rating more than once.
Any suggestions on parametric tests to run?
TIA
There are so many ways in which your data can be approached, but without more information (e.g., sample sizes, data distribution, intercorrelations), it is difficult to specify which tests would be best. One that comes to mind is a cluster analysis.
If you want only to test the hypotheses as stated, two different MANOVAS would work fine, with country being the IV & the different styles (and then the motivations) being the DV. Of course, you need to check at least the distributions of the responses first to make sure that you do have an acceptable distribution of ratings, since a MANOVA is an analysis of variance.
BTW, it really helps to have a statistical technique in mind before one collects data. Then you can just make sure that your data satisfy the assumptions before proceeding.
If you have a decent sample size, you can just use a a t-test for differences in means.
José-Ignacio Antón That would work, but the alpha value would need to be adjusted for family-wise error.
- Badges
- Science topic
- Similar topics
- Psychology
- Mental Processes
- Learning