Repeated measures design - correct analyses and GLM assumptions

#1
Hi all,

I have collected data for a study but I am not sure now if I am carrying out the correct analyses. I have tried for the last few days to figure out whether my analyses are correct, but I keep getting a lingering doubt that they are not and so I thought I will ask this question because I have tried to figure this out but with no good degree of certainty.

Background of study

The study looked at whether loading a person with a bit of information (low load, which is first IV) versus a lot of information (high load, which is second IV) led to different numbers of intrusions (i.e., the DV). This design was repeated measures with people getting both low and high load stimuli. Low load has been organised into one column and high load into another column.

Are these analyses for repeated measures correct?

The issue is that I have carried out some analyses but I feel they could be wrong.

1) To look at significant difference I have simply used a paired t-test with these two variables

2) To find out if other demographic variables, for example, were related to levels of intrusions, I carried out a repeated GLM analysis with one factor with two levels (low & high load) and then put the various demographic variables in all at once as covariates.

Which analyses for repeated measures to solve the problem below?


1) To make sure GLM assumptions are met, I removed outliers but I have not been sure how to carry out other assumption analyses for repeated measures. In particular, checking for (a) normality of data, (b) unbiasedness, (c) homoscedasticity and (d) independence.

Thanks and I hope this question will also benefit other people that read about it that are doing a repeated measures design but have not had much experience of analysing repeated measures data like myself.