So I did an experiment for 1 month where I investigated the change of pH of 3 different types of milk in the refrigerator and room temperature. I'd now like to compare the results of each type of milk in the fridge with those at room temperature, which leaves me with a pair of results to compare. Therefore, I came up with the paired t-test. However, I do not understand whether I will have to do a t test for the results of each day or whether I can do one single t test for all the results of one group?
$\begingroup$
$\endgroup$
1
-
1$\begingroup$ I would consider a repeated (paired) measures ANOVA... $\endgroup$– Giuseppe Biondi-ZoccaiCommented May 2, 2016 at 8:53
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
If for each sample of each type of milk, you divided that sample in two and gave the treatment (one in refrigerator, the other room temperature) to each, then you can use a paired samples t test, or maybe, as mentioned by @Joe_74, a repeated measures anova, which would permit also taking into account other factors (such as type of milk).
You should not do a t-test for each day, one common test. However, if the two samples used in refrigerator and at room temperature did not arise arrived from subsampling one single sample, a paired test would not be appropriate.
answered Sep 1, 2017 at 20:17