Hello!

A basic question from a not-very-statistically-savvy person:

I have data from 2 groups of patients who underwent knee replacement operations. One group was pre-conditioned (treatment) the other was not (control).

Blood samples were taken from both groups at 6 time-points (1 pre-op, 2 intra-op and 3 post-op).

From those samples the levels of various cytokines with funny names like IL-12p70 and IFN-y were measured.

I am considering each variable (cytokine) separately and thinking that a repeated measures ANOVA would be the appropriate test to do in order to compare the two groups, but I'm not sure...

Any thoughts would be much appreciated!

Hi okennedy,

Are the cytokine levels the dependent variables?
And is treatment the only independent variable (other than time)?

If so, yes, you want repeated measures anova. In fact, it's a mixed model because time is within subjects (repeated) and treatment is between subjects.

Karen

Hi Karen,

Yes, the cytokines are the dependant variables, and the treatment is the only independant variable other than time.

So, I'll give the mixed model repeated measures anova a go!

Thanks a million for your help, it is much appreciated.

Also, I've just signed up for Statwise - it is a great resource...I'm looking forward to the first newsletter!

Thanks again,

Oran

Hi Oran,

Thanks for signing up! Hope you find it useful. The next one goes out next week.

Karen

Hi Karen,

Thanks, I'll look forward to that.

A quick follow-up question...if I may?

Resources over here are limited so I tried doing my analysis (two-factor ANOVA with repeated measures on one factor) using an online calculator...and I'm not sure if I'm using it correctly.

I have entered my data in a 2 row x 4 column structure. Where data in row 1 is control and row 2 is treatment. Each column is a different timepoint.

The result sheet has various values (SS, df, MS) among them are three F-values of 0.09 for A(between subjects) 2.68 for B(repeated measures variable) and 1.13 for AxB

So I think I'm either using this tool incorrectly or this experiment shows nothing of interest!

Here's hoping for the former...

Thanks again!

Oran

Hi Oran,

I really can't tell you if you're doing it correctly. There are so many ways to set up the data--it differs for each program.

Usually each row in a data spreadsheet is a subject and each column is a variable, but even this can vary.

Karen