Using Shapiro-wilk test, test the normality of scores (pretest and post test) If scores are normally distributed use paired samples t-test, (this test enable to see you changes over time)

If you data not normally distrubuted ((pretest and post test scores)  use Wilkinson signed rank test.  (this test enable to see you changes over time)

dont forget to compare pre-test and post test scores (use independent sample t-tests or manN- Whitney (If normality assumption violated) this demonstrates your intervention effectiveness, ıf there were no difference in pre-test scores between groups but experimental group significantly higher scores than control group) . if pre-test scores significantly differ across groups  use ANCOVA and add co-variate as pre-test scores. group as fixed factor ( experimental and control) and post test score as dependent variable.

https://www.researchgate.net/figure/Pre-Test-mean-comparisons-between-the-Experimental-and-Control-Group_fig3_275993174

http://ietd.inflibnet.ac.in/bitstream/10603/1548/14/14_chapter%204.pdf

Hello David,

If you're content using number right scores as the outcome of interest, then a one between, one-within anova may be appropriate for your analysis. The between-subjects factor is group membership, and the within-subjects factor (the repeated measures factor) is occasion (pre, post). Of the three principal tests available (main effect of group across combined occasions; main effect of occasions across combined groups; and group x occasions interaction), the third test is usually the one of greatest interest, as it would help you deduce whether pre-post changes are equal across groups. As Cristian's comment implies, there are assumptions that need to be met (normality, homogeneity, independence).

Should you have concern that assumptions are not viable, then exact/permutation or bootstrap/resampling methods would offer a viable alternative to the traditional result. SPSS and other statistical software offer these options in more analytic methods (and the R system has multiple libraries that offer one or more of these as well).

Here are some links that can help explain the alternative methods:

1. https://statweb.stanford.edu/~tibs/stat315a/Supplements/bootstrap.pdf

2. https://machinelearningmastery.com/a-gentle-introduction-to-the-bootstrap-method/

3, https://www.burns-stat.com/documents/tutorials/the-statistical-bootstrap-and-other-resampling-methods-2/

Good luck with your work.

I agree with Ajit kumar Roy