Reply
Thu 5 Mar, 2020 03:36 am
Hello!
For my thesis I want to compare 3 training groups (2 different types of training and 1 control group) to see the influence on mindfulness-skills and exhaustion level (so 1 IV- type of training- and 2 DV). All the participants get a pre and post test.
I'm doing the analysis in SPSS, but I am very confused about the different options. I thought I had to use MANOVA, because we want to look at 2 dependent variables. But I need to look at the whitin- subject and between-subject effects (so in each group I want to know if the training had an effect per group but also between the different types of groups- keeping in mind that time between the tests can also play a factor).
Because we have two measures, I was thinking of using repeated measures, but I'm not sure.
The assumptions of normality and homoscedasticity are oke, but there are some outliers which may cause issues when doing a MANOVA? Do I need to work with contrasts or post-hoc tests?
We also want to control for variables like gender, age, education, ... but can they be covariates when there is barely any correlation between these and the dependent variabeles? Is it worth using them as covariates?
Please help! What should I do?
@jotu,
Start with the basics. You say you have two independent variables? From what you describe, it sounds like you are testing one variable (training type) that has three possible values, T1, T2 and Control. Your other variables (gender, age, education) would also be independent variables, so we're up to four just with those.
You can use MANOVA or two separate ANOVA analyses. I would not assume any co variance unless you see evidence that your test data set is not very random.
Stumble It!
Tweet This
Bookmark on Delicious
Share on Facebook
Share on MySpace