0
$\begingroup$

I have some data comparing training participants who experienced the same training course either face-to-face or virtually (via Zoom). I have two survey items I am comparing. Sample A is for virtual delivery and has 186 participants, and sample B is for face-to-face delivery and has 100 participants. I have the raw data and can compute means, standard deviation, etc. What is the appropriate text of statistical difference? Student's T-Test or Wilcoxin/Mann-Whitney?

Complicating the matter is that the samples were taken at different time periods, roughly pre-COVID and during COVID. I.e, July to December 2019 versus July to December 2020. I appreciate that my samples are more convenience, there is no randomisation to condition, etc. But this is a real-world opportunity to compare the differences. Any thoughts on this?

Thanks for your guidance!

Gene Johnson

Improve this question
$\endgroup$
7
  • $\begingroup$ Welcome to Cross Validated! What do you want to compare? Both tests you mentioned examine differences in means while not being concerned about much else (it's more complicated than this, but both have solid properties when they are used to detect differences in means). $\endgroup$
    – Dave
    Commented Jan 19, 2022 at 23:04
  • $\begingroup$ Two survey items. Unfortunately we changed specific survey items over that time period. We do have two identical items: one on clarity of learning objectives, and one on facilitator's enhancement of experience. Fortunately they reflect two important aspects of the training experience. $\endgroup$
    – Gene Johnson
    Commented Jan 19, 2022 at 23:06
  • $\begingroup$ What about those survey responses? $\endgroup$
    – Dave
    Commented Jan 19, 2022 at 23:07
  • $\begingroup$ They are Likert-scaled, agree to disagree. So I can compare mean responses for each group. $\endgroup$
    – Gene Johnson
    Commented Jan 19, 2022 at 23:07
  • $\begingroup$ What are you trying to answer, whether there is a statistical difference in test scores between the two groups? $\endgroup$
    – Jay Ekosanmi
    Commented Jan 19, 2022 at 23:17

0

Reset to default