I have 3 treatments groups and it has a p value < 0.05, in ANOVA. However, when post hoc pairwise comparison is done, there is no significant difference between group 2 and 3. Can I still argue that my result is good?
So, all the groups are significantly different from one another in the post-hoc, except for 2 and 3 ?
Statistics don't allow you to say a result is good or bad. As Sal Mangiafico is saying, you need to add more information to your question if you expect useful answers.
You might read section 7.2 here, on the hypotheses for a one-way anova. https://www.stat.cmu.edu/~hseltman/309/Book/chapter7.pdf .
https://statsandr.com/blog/how-to-do-a-t-test-or-anova-for-many-variables-at-once-in-r-and-communicate-the-results-in-a-better-way/
https://www.iuj.ac.jp/faculty/kucc625/method/anova.html
https://www.statology.org/what-is-the-difference-between-a-t-test-and-an-anova/
https://www.wallstreetmojo.com/anova-vs-t-test/
https://askanydifference.com/difference-between-t-test-and-anova/
Regards
You may have a significance between 1 and 2 or 1 and 3 or both but not between 2 and 3. Therefore you have a p of .05 instead of e.g. .01.
I suppose certainly it's possibile, because ANOVA and It probabilità 'p' value among 3 groups ( a ' composed variance') cannot be valid for group 2.vs 3 (a more simple 'direct' variance) But some details should be useful for a better judgment.
- Badges
- Science topic
- Similar topics
- Computer Science
- Program