3 different treatment groups (n=12 per group). Several behavioral parameters. There are different timepoints at which the behaviours have been recorded.
Based on the information you provided, you have two factors, time and treatment. With two way repeated ANOVA, you can analyze the effect of time (alone), treatment (alone) and the interaction between time and treatment. This statistical probe can be done with a bunch of softwares, Sigma Plot, SPSS, probably MS Excel, but I'm not sure that Prism has this feature.
You can check our previous behavioral study for more details:
https://www.researchgate.net/publication/253334212_Obestatin_prevents_analgesic_tolerance_to_morphine_and_reverses_the_effects_of_mild_morphine_withdrawal_in_mice
Nándor Lipták Thanks for your response. I'm confused because I have three groups.
The two-way ANOVA that Nándor Lipták mentioned is needed because you have two factors (time and treatment). It is independent from the number of experimental groups. You would need to run the ANOVA for each behavior parameter. Prism can perform a two-way ANOVA. Open Prism and navigate to the Help menu. You should find a "PRISM Statistics Guide" entry which gives a good introduction into the test and how to run it in Prism.
Roland Bock Thank you for your response. Please see the image I've uploaded. I have two more groups to the right with similar values. I thought I needed to do a one-way ANOVA?
This is difficult to answer without knowledge of the experimental design and scientific question. The scientific question you try to answer with this experiment should tell you what type of statistic you should use.
Your table setup suggests that you have multiple treatment groups (A ...) with 12-13 subjects in columns and a varying experimental factor in rows (dpi). Since the first row is labeled "baseline" I would assume that the rows represent responses/measurements to a changing variable. In this case the 2-way ANOVA is needed to answer if the treatment effect differs with the changing row variable.
A one-way ANOVA is appropriate when you would compare only the treatment (at baseline, for example) for all 3 groups, since the t-test is limited to the comparison of 2 experimental groups.
I would highly recommend to read through the online Prism statistics guide to give you more detailed information.
Roland Bock Yes, there are three groups (A, B, C) and each group has 24 subjects (animals). Rows represent different timepoints at which behaviors were recorded (baseline, 30 days, 60 days...). I want to know if the three groups have differences in behavior at these different timepoints. Do I still need a 2-way ANOVA? From my readings of different papers with similar data, I see 1-way ANOVA with Tukey/Bonferroni corrections.
From your description the same animal was tested with the same behavior on different time points and you are comparing multiple treatment groups: hence the proper statistical comparison is using the 2-way repeated measures ANOVA (2-way for your 2 factors (treatment and time), repeated measures: same subject was tested multiple times (time factor)).
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Biology
- Biomolecules
- Nucleic Acids
- RNA