2
$\begingroup$

I am trying to compare two matched samples. In total I have a sample of 34 people. Each patient receives two treatments, a C1 treatment and a C2 treatment. So each patient will be compared to himself according to different parameters. Some of my variables are normally distributed while others are not (or even not unimodal). Can a parametric test based on the central limit theory be used (as my sample size >30)?

See below the distribution of each variable.

enter image description here

Improve this question
$\endgroup$
4
  • 6
    $\begingroup$ (Andrew Gelman) As you know from teaching introductory statistics, 30 is infinity. $\endgroup$
    – user2974951
    Commented Oct 12, 2022 at 11:50
  • 1
    $\begingroup$ But seriously, what is your hypothesis and how will you test it? $\endgroup$
    – user2974951
    Commented Oct 12, 2022 at 11:50
  • $\begingroup$ It is not a confirmatory study. It is just exploratory. The objectif is to compare the means between the two groups et prove that their means are different $\endgroup$
    – Seydou GORO
    Commented Oct 12, 2022 at 12:17
  • $\begingroup$ In any case, if this is exploratory a simple t-test will do unless you have skewed data or something. $\endgroup$
    – user2974951
    Commented Oct 25, 2022 at 9:06

0

Reset to default