Greetings,
I’m testing the antagonistic effect of three bacteria (B1, B2, and B3) alone and in combination with another bacteria R (B1+R, B2+R, and B3+R) against Fusarium. I measured the inhibition percentages for the 6 bacterial treatments and I want to know if there is a significant difference between the inhibition percentages registered with single and combined bacteria (is the inhibition percentage registered with B1xR significantly different from that registered with B1 alone and so on). What is the most suitable statistical test in this case?
Best regards.
I think an interaction term in a 3x2 ANOVA would tell you whether the effect of B (1, 2 or 3) depends on the level of R (R or no R).
If you want to further test combinations at different levels then I would look into pairwise (orthogonal) contrasts, can get some really neat planned analyses going with these.
George Evangelou, Thank you very much for your answer, a colleague suggested doing an independent samples t-test for each pair (between B1 and B1xR and so on) what do you think about that?
Ps: the inhibition percentages are means (i used three replicates for each bacterial treatment).
Emm W Kay Indeed t-test's will be what follows from the ANOVA so you get that plus more information. You will get to see whether B and/or R have a independent effects of their own. Further, should there be a significant interaction, you will be able to follow up with the t-tests you need (apriori if planned beforehand) or all t-tests (post-hocs which will need to have corrected alpha levels).
I suppose if you really are only interested in B1 vs B1*R, B2 vs B2*R and B3 vs B3*R then you could just do the t-tests. Depends only on your research question I guess!
- Badges
- Science topic
- Similar topics
- Philosophy
- Philosophy Of Science
- Reasoning