I have conducted a study comparing an experimental group to a control group. Both groups were given a pre-test. The experimental group was then treated with a math gaming system. After a seven week period a post test was administered to both groups. Math scores were compared using ANOVA to determine if the scores were significantly different. ANOVA assumes homogeneity of variance. My study failed. Laerd said that this violation was not a serious one when the groups are equal sized as mine were. However, they make post hoc suggestions to remedy the violation but the post hocs require at least three categories (pre,mid,post) to determine where the differences lie. I have reported that the violation is not serious. How do I report why and explain in dissertation format that I can't remedy through post hoc. Or is homogeneity even applicable with a prepost test?

  • 1
    $\begingroup$ What does "eqaul group sas mine" mean? Perhaps you should be using a Wilcoxon signed-rank test for paired data? Why do you want to use ANOVA when a test for paired differences of location is likely all you need to do? $\endgroup$ – Carl Mar 16 '18 at 7:44
  • $\begingroup$ as mine. (my apologies) $\endgroup$ – LaDonna Martin Mar 16 '18 at 9:06
  • $\begingroup$ Your study failed what? Leard is a person or a they; more than one person? The post hocs require at least three categories? What? For ANOVA perhaps, otherwise, no. What can you not remedy, heterogeneity? $\endgroup$ – Carl Mar 16 '18 at 9:43
  • $\begingroup$ I am using ANOVA as required by a college committee. Laerd is the statistical resource guide I was refered to. Yes is there a way to remedy heterogeneity? $\endgroup$ – LaDonna Martin Mar 16 '18 at 14:34
  • $\begingroup$ Your study failed what, how? There are many ways to process appropriately process data despite heterogeneity, e.g. Wilcoxon, nonparametric regression, etc.etc., which to use depends on data and goals of study. Need more information. $\endgroup$ – Carl Mar 16 '18 at 19:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.