I have a time series analysis, in which I have determined a mean per each phase. When I run an ANOVA Test with repeated measures comparing the phases, I receive a non significant result, despite a visible difference between means per phase. I would assume that this is due to high variability amongst the phases. If I run the same analysis, albeit with a T-Test, then I receive a significant difference between 2 phases. Question is, is this discrepancy due to variability, or another factor? Can one just use T-test on a phase per phase basis, or does a multiple phase analysis require an ANOVA test?
I am not sure that I understood the question completely. In short, a t-test and an ANOVA should give the same result for a pre-test post-test comparison on two groups. However, if you have more than 2 timepoints a repeated measure ANOVA is appropiate, and if you have more than two groups an ANOVA would also be appropiate. Maybe there is also some help for your question in this discussion:
https://www.researchgate.net/post/What_is_the_difference_between_T-test_and_ANOVA2#:~:text=The%20t%2Dtest%20is%20a,statistically%20different%20from%20each%20other.
Thank you for your answer and reference to the previous thread. In short, why is it that a post hoc test via ANOVA shows no significance between Phase 1 and Phase 2 for example, but a T-Test does reveal a significance between Phase 1 and Phase 2? It is mots likely due to variance factors, but just wanted to see what other people had to say.
Hello Michael Nordine,
I can't answer your question without the data, but take into account several things:
I hope this information will help you.
Best regards,
JD
Hello Juan, and thank you for the informative answer. This makes alot more sense too me, as I have read up on the differences between RM ANOVA, ANOVA, T-test, etc. I left the significance at 0.05 for now, but this can be adjusted using JASP (program which I am using).
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running