0
$\begingroup$

My data is psychological test scores of teachers and students before and after an intensive training course. The group of students is more than twice as big as the group of teachers. Several teachers and students worked together in small groups during the training. Unfortunately, information on which teacher(s) have worked with which student(s) is not available anymore.

I want to test if the teacher scores before the training correlate with how much the scores of students improved during the training. How can I approach this?

I'm new to statistics (mostly self-taught) and really stuck because of the unequal group sizes, so any help is highly appreciated.

$\endgroup$

1 Answer 1

0
$\begingroup$

Without being able to match teachers to students in some way, you won't able to answer your question. You do have the data to answer questions such as "what is the effect of the training?" and "does the effect of the training differ between students and teachers?". Of course, in both cases it would be much better if you knew who worked together.

$\endgroup$
5
  • $\begingroup$ My research question was whether the teachers test scores (pre training) potentially affect how well they can teach the intervention to their students. I hypothesized a positive correlation between the teachers pre scores and the pre-post difference of students scores. The data wasn't collected in a regular school setting, so theoretically all teachers have worked with all students to some degree. (Of course, the small groups played a major role, but I won't be able to investigate that influence.) Is there a way to get at least a very rough answer to my reasearch question? $\endgroup$
    – Jonas
    Commented Feb 14, 2023 at 19:13
  • $\begingroup$ I thought about comparing the teachers average score (pre training) with the students average score (post training) but don't know how to take the students pre scores into consideration... $\endgroup$
    – Jonas
    Commented Feb 14, 2023 at 19:13
  • $\begingroup$ Your question makes sense, but it doesn't sound like you can run a correlation. Do you have any data you haven't mentioned that could match them? Such as, there were multiple classes and you know who was in which class? $\endgroup$
    – David B
    Commented Feb 14, 2023 at 20:07
  • $\begingroup$ Thanks for your help. Unfortunately, the only information I have (other than psychometric data) is participant ID and ROLE (teacher or student). Before deciding on my question, I was told data on matching was available but there has been a misunderstanding. Apparently, there were 1-3 teachers and 3-5 students in each small group but have no data to match the groups and it's to late to work out a completely new question :( $\endgroup$
    – Jonas
    Commented Feb 14, 2023 at 20:36
  • $\begingroup$ I mean... you probably should work out a new question, because there's no way to answer the one you posed. $\endgroup$
    – David B
    Commented Feb 14, 2023 at 20:49

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.