MANCOVA and MANOVA Discussion > Within-subject effect vs. [independent variable] effect
Hi Nina!
This is a great question. The "within subject effect" indicates that individuals in your sample tend to vary on their score for your DV across the measures (as the name indicates, it examines "within-person/subject" variability).
The other table (which sounds like it is the between-subjects effect) indicates that there may be variability across the three measurement occasions (not taking into account the within-person effect). This is somewhat similar to just grouping the results into the three measurement occasions and seeing if there are mean differences across the occasions, but this may not be a great idea, because that ignores the fact that the measurements are not independent.
I hope that helps!
May 6, 2013 |
Jeremy Taylor
Jeremy Taylor
Hello!
I'm hoping maybe someone can help me interpret two different tables in my SPSS output. I ran a single-factor, repeated measures (3x), multivariate (2 dependent variables) experiment. The repeated measures are not points in time, but rather levels of the factor. N = 37
I ran a MANOVA to see if the levels of the factor have an effect on my two variables. SPSS has given me two tables, both with significant results, but different error degrees of freedom.
One of the tables is titled "within subject effect" and has an error df of 142. And the other table is titled according to the name of my factor and has an error df of 33.
What is the difference between these two types of output? Since I only ran the experiment with repeated measures to reduce the number of subjects I needed, I'm assuming the "within subjects effect" is less interesting for me. Please correct me if I'm wrong!