Hi all,
I hope someone can help me with a query I have about the results of a multivariate repeated measures ANOVA I have conducted.
The analysis involves participants' scores on various measures on three occasions (e.g. Anxiety at baseline, Anxiety at 6 months, Anxiety at 9 months; Depression at baseline, 6 and 9 months).
The univariate test outcomes table shows a significant main effect for Anxiety and Depression over time, and then the within subjects contrasts table indicates that there is a BOTH linear and quadratic significant effects for both - what does this mean?
Looking at the means there certainly appears to be a linear trend in that anxiety (and depression) levels reduce from baseline to 6 months (which is significant in the pairwise comparisons later), but the means and the graph in the anxiety scores (although not for the depression scores) also suggests that anxiety increases slightly at the third and final timepoint.
I'm a bit confused as to what I should report - the linear or the quadratic trend?? Is there anything more that I should do to further explore these trends?
Thanks in advance for any help anyone can give.
Fiona
|