Hi
I am trying to find significant differences in my experiment between day 1 and subsequent days (2-7) ie. 1 and 2, 1 and 3, 1 and 4 etc. I can't use ANOVA or T tests as they require factor levels to be independent of each other. My samples are in duplicate. Any suggestions on an appropriate stat. test and post-hoc test and also which program to use (SPSS or excel would be good!)
Thanks!
2007-01-17 21:46:39 · 2 answers · asked by Andrew H 1 in Science & Mathematics ➔ Mathematics
If you think there might have been a predictable tend (eg increasing results as the days go by) then you could plot results vs day and check if they form a relatively striaght flat line.
I think excel has a line fitting tool you can use. or you could just use a regression method to fit a flat line.
Then you could also plot the deviation from the line to show any trends.
I am not sure how you would quantify the significance, I would just find the standard deviation and mean and show that there is a large standard deviation is there is a sigificant difference.
2007-01-17 23:21:29 · answer #1 · answered by Mike 5 · 0⤊ 0⤋
If you have access to SAS you can run a repeated measures analysis in proc mixed. I'm not sure if SPSS has anything equivalent.
2007-01-18 14:04:32 · answer #2 · answered by statboy76 2 · 0⤊ 0⤋