What statistical method would you apply to determine if treatments are statistically different. For one treatment I have 26 samples all with 4 measurements. Those measurements are then averaged for each sample and then added for a total value. This total value is then compared to the other treatments' total values. Is there any way to statically confirm that these total values for each treatment are significantly different from one another?
Asked
Active
Viewed 130 times
1 Answers
0
Once the results are averaged in the way you describe, you could use a t-test (with some assumptions) or a Mann Whitney U test with very few assumptions.
I am not sure that averaging is a good idea here, it depends on context, but it probably loses some information.
Peter Flom
- 94,055
- 35
- 143
- 276