mac101
06-22-2010, 12:53 PM
Hi,
I am doing the analysis for a study during which we had 2 groups of therapists (new and experienced). Each group performed 5 thrusts and we took some measurements such as peak force. When I graphed the mean and SD. I noticed novices had a much greater SD than experienced therapists which was expected. The means are very close so a t-test is not significant. What I want to do is run an analysis that would tell me that novices where less consistent in their 5 trials than experienced therapists.
I did my own research and found a method in which I get the variance for each group and then get the ratio between the variances and I can use the F distribution tables with both degrees of freedom being 4 (5 trials -1).
I have a few questions about this:
1. First of all is this approach correct? If not, what would be a better way of doing this?
2. If the approach is correct, something doesn't seem right. The examples I saw were for 5 trials vs 5 trials and this was easily done. However, in my case we have more than one therapist. Each therapist has 5 trials. So, each will have a different variance for each therapist. Can I take the average of variances for each group, if not what can I do?
3. Last but not least, I have no problem doing this in excel but is there any easier way of doing this analysis in SPSS, does anyone have a link to a tutorial?
Thank you very much for your time and hope someone could help me out.
Mac
I am doing the analysis for a study during which we had 2 groups of therapists (new and experienced). Each group performed 5 thrusts and we took some measurements such as peak force. When I graphed the mean and SD. I noticed novices had a much greater SD than experienced therapists which was expected. The means are very close so a t-test is not significant. What I want to do is run an analysis that would tell me that novices where less consistent in their 5 trials than experienced therapists.
I did my own research and found a method in which I get the variance for each group and then get the ratio between the variances and I can use the F distribution tables with both degrees of freedom being 4 (5 trials -1).
I have a few questions about this:
1. First of all is this approach correct? If not, what would be a better way of doing this?
2. If the approach is correct, something doesn't seem right. The examples I saw were for 5 trials vs 5 trials and this was easily done. However, in my case we have more than one therapist. Each therapist has 5 trials. So, each will have a different variance for each therapist. Can I take the average of variances for each group, if not what can I do?
3. Last but not least, I have no problem doing this in excel but is there any easier way of doing this analysis in SPSS, does anyone have a link to a tutorial?
Thank you very much for your time and hope someone could help me out.
Mac