0
$\begingroup$

How can I test the responses of students of a single question (Opinion) before (n=84) and after an intervention (n=81) with responses based on a 5 point Likert scale (SDA-1, DA-2, N-3, A-4. SA-5)?

For most of the students, before and after are same. But some students dropped out later, or didn't respond after the intervention, so its a different sample size.

$\endgroup$
2
  • $\begingroup$ Do you have the same people before and after? It looks from your $n$ that they are different. $\endgroup$
    – mdewey
    Jan 25 '17 at 17:16
  • $\begingroup$ Why not a paired t-test? See the duplicate flagged question above bearing in mind for repeated measures data things are different. 3 people lost to follow-up is not a substantial proportion. Just inspect their baseline values and report. $\endgroup$
    – AdamO
    Jan 25 '17 at 20:11
0
$\begingroup$

Most of the students before and after is same. But some students dropped out later not to respond after the intervention so its a different sample size

$\endgroup$
1
  • $\begingroup$ Welcome to CV! You shouldn't provide additional explanation as an answer. Instead, edit your original post, or add the explanation as a comment. $\endgroup$ Jan 25 '17 at 18:58

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