3

In my experimental design, mice are given 2 small wounds on either side of their bodies. The first wound is treated with an experimental drug A, and the second with a control drug B. Over the course of 14 days, wound size is recorded every other day as a percentage of that mouse's original wound size on day 0. Averages for each day are taken and the final data set looks like this:

Drug A         Day      Avg. Wound area %
               0        100
               2        72
               4        47
               6        35
               8        30
               10       27
               12       24
               14       22


Drug B         Day      Avg. Wound area %
               0        100
               2        80
               4        70
               6        65
               8        63
               10       54
               12       48
               14       45

If I perform a t-test, I would have to perform multiple unpaired t-tests for each time point between the two groups, and that would leave me with multiple p-values. Is there a specific statistical test that can tell me if two groups are significantly different over a given timeframe? Is a two way ANOVA appropriate? Thanks in advance.

Arvene Golbazi
  • 31
  • 1
  • for t.tests on time series check out [this thread](https://stats.stackexchange.com/questions/3616/what-is-the-best-statistical-test-for-a-time-series) – HoofarLotusX Sep 17 '18 at 13:46
  • This problem has the structure to be tackled by a repeated measures anova test. My only concern is that the underlying assumptions for it to work may not hold (normality and spheric variance). – MauOlivares Sep 17 '18 at 16:10

0 Answers0