Two-way ANOVA or one-way ANOVA for time-course experiment with different drugs?

[Biochemist]

Hi,

I am not sure which statistical procedure can and has to be applied here.

First, I will try to explain the experimental setup resulting in the data to be analyzed statistically. The data are levels of the cytokine IL-6 in the blood of mice treated with three different drugs and of untreated control mice. The mice were given one of the drugs at day 0 (or not given any drug in the case of the control mice). Then, blood was taken for determining IL-6 levels at day 5, day 7, day 10 and day 14.

The question to be answered is: Are there any statistically significant differences in the IL-6 levels between treated mice and control mice at the individual time points?
The time course itself is not of interest here, i.e. when IL-6 levels increase significantly.

So, would it be OK to use a one-way ANOVA + Dunnet's post-hoc test for the data from each individual time point (d5, d7, d10 and d14) to find out whether there are statistically significant differences between treatments and control? Or do I have to use a two-way ANOVA because there are two factors here, time and drug treatment?

 

[Buckeye]

To me, it seems the two-way design is the way to go here.

 

[Biochemist]

Can you explain to me why you think so?

I mean testing for factor interactions is not of interest here, which is one of the questions to be answered by a two-way ANOVA, and neither is the time course.

 

[GretaGarbo]

To me it looks like one way anova with repeated measures. Search for it. Here is one example.

 

[Biochemist]

@GretaGarbo: Thanks for the link. Actually, it is a combination of the two examples given on the first page of your website link. There are different time points as in the first example and there are different treatments as in the second example. The experimental design is definitely a repeated measures design but there is a one-way repeated measures ANOVA and there is a two-way repeated measures ANOVA.

If I was interested in the influence of both factors (time and treatment), then I would clearly say two-way ANOVA is the way to go but since I am only interested in the effect of the treatment, I am uncertain about it.

Does the fact alone that two factors are involved in the experimental design require analysis by a two-way ANOVA?

 

[robertjohnson]

It was always a big issue for me to solve such a test! It requires a deep analysis and great understanding of the subject. That is why I decided to refer similar requests to professionals - follow http://customcollegeessays.com/blog/page/4 for further details.