I'm trying to analyze the results of my Master Thesis with SPSS.

I've 3 experimental conditions (factor: Group) and every subject in each condition has 2 expressions (factor: Hemisphere). I want to test the effects of these 2 factors on my dependent variable A. What I would like to test is:

- effects of Group on A

- effects of Group x Hemisphere on A

- effects of Hemisphere on A between my 3 Groups

- effects of Hemisphere on A whithin my 3 Groups

If I understood correctly, this corresponds to a repeated measures ANOVA. Is that correct?

However I'm not sure whether I should use a parametric test. My data appears to be normally distributed, but my sample size is really small (n=8 per group). Do you think I should rather use a non-parametric test? What would be the equivalent of a repeated measures ANOVA for non-parametric statistics?

Thank you very much for your help! ]]>

I am conducting a study that compares rates of CIF (Client Intake Forms) completion collected among via self-administered paper-based interviews (SAPI) as compared to electronic collection via iPads. By completion I mean CIF that have no missing values, that is all questions were answered in the CIF. I am using data of 26 centers. These centers administered the CIF in the two modalities - paper and ipad. The CIF is administer to every single client that appear at the centers. (8000 clients). In total, 70% of CIFs were administered via paper and 30% via iPAd. I have two questions:

1. Should I include in my manuscript p-values? I believe that my data is not a sample because all clients were included. I believe it is more like a census type of data. when I run test of significances, everything is significant even small differences since my sample is huge.

2. Although all center have the two option modalities (paper CIF or ipad), some centers only use on modality. For example center1 have 100 CIFs via the paper and 0 CIFs via iPad. Is this a problem? will this affect my results and the validity of my study?

2.1 Also some centers, have very low client volume- i.e. only 3 CIFs. Should I exclude center with low CIF volume like my example. If so, what would be the rationale?

I think that the problem is that were not randomly assigned to one of the other modality.

Any thoughts?

Best,

Marvin ]]>

Outcome1 Outcome2

Group 1 100 (low) 80 (high)

Group 2 77 (low) 23 (high)

which gives me a p value of 0.0004. Am I doing it correctly, also, under what circumstances would I use a chi-square test instead?

Thanks a lot!!! ]]>