1
$\begingroup$

I ran a three-way ANOVA to find effect of insecticide (2 levels), species (2 levels) and location (2 levels) on body condition of turtles. I did not find any three-way interaction. However, I found an interaction between insecticide and species! I am not familiar with three-way ANOVA and I dont know how should I interpret my results and plot my data!! should I run a two-way ANOVA and focus on main effects?

$\endgroup$
1
$\begingroup$

You should be able to work with the model as is. Plot your data! What does it say? What about post-hoc tests? Keep going!

$\endgroup$
  • $\begingroup$ So, you are saying I should run a Two-way? sorry I did not get it $\endgroup$ – Non Mar 3 '17 at 19:48
  • $\begingroup$ Probably running a three way with that one significant two-way interaction would work. One could wonder whether some simple transformation of the response variable would eliminate the interaction, but we can't say much about that without the actual data. $\endgroup$ – Michael Hardy Mar 3 '17 at 22:25
  • $\begingroup$ Ok, so should I report two-way interaction between insecticide and species. and also report F value for each factor from the same table? or run two-way ANOVA and conclude on that? $\endgroup$ – Non Mar 4 '17 at 2:23
  • $\begingroup$ @MichaelHardy Ok, so should I report two-way interaction between insecticide and species. and also report F value for each factor from the same table? or run two-way ANOVA and conclude on that? $\endgroup$ – Non Mar 4 '17 at 3:05
  • 1
    $\begingroup$ I agree with @jebyrnes - plot your data, use the entire 3-way ANOVA table and post-hoc tests. A picture paints a thousand words so plot your data! You might be able to get more useful answers by posting the ANOVA table in your question, as well as any graphs you have. $\endgroup$ – Groovy_Worm Mar 4 '17 at 20:08

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.