0
$\begingroup$

I am analyzing data where my experimental design is as follows: DNA and RNA samples were taken from 3 mesocosms at seven time points across the duration of the study to see DNA and RNA decay across time.

For example: At time point 1, 1 DNA and 1 RNA sample was taken from each mesocosm (total 3 samples of DNA, 3 samples RNA for each time point).

My understanding is that when I do a pairwise t-test for DNA and RNA at each time point, I say that DNA and RNA are NOT paired: data.type = DNA or RNA count = copies of DNA or RNA recovered

dataset %>%
  group_by(timepoint) %>%
  pairwise_t_test(
    count ~ data.type, paired = FALSE,
    p.adjust.method = "BH") 

However, since I am doing repeated measures in the same mesocosms over time, when I do pairwise comparisons of DNA/RNA counts over time, my samples are paired:

dataset %>%
  group_by(data.type) %>%
  pairwise_t_test(
    count ~ timepoint, paired = TRUE,
    p.adjust.method = "BH")

I just want to make sure I am doing this correctly and not specifying if something is paired when it's not and vice versa.

Thanks!!

$\endgroup$

1 Answer 1

0
$\begingroup$

Each time period $t$ of measurement during the experiment can be considered as an experimental unit. Given this, and the fact that each test unit (mesocosms) has two associated measurements (DNA and RNA), then this would make DNA and RNA a paired t-test.

Disclaimer: I do not have a applied professional experience in experimental study designs. Rather, just what I have learned through textbooks / papers.

$\endgroup$
1
  • $\begingroup$ Criticism to my thinking here is appreciated—thanks $\endgroup$ May 5, 2022 at 19:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.