Rebecca W. answered • 02/21/21
Harvard Phd explains statistics clearly from basic to multivariate
A little more information about the study is needed.
Did each case receive just one of the treatments (e.g., pressure 1, pressure 2, control)? If so, you can do an independent samples one way ANOVA (analysis of variance to compare means across groups). If you use SPSS, the easiest analysis is the ONEWAY procedure.
Did each case receive all three treatments? If so, you have a one way repeated measures ANOVA. If you are using SPSS, you would use the GLM (General Linear Model) procedure. (Note, this is different from the generalized linear model procedure). Within GLM you can select repeated measures as your type of analysis. The menu selections are somewhat complex and you might want to look at a Youtube video to see how these work.
With two outcome variables, a journal editor or teacher might conceivably ask for an analysis that considers these variables at the same time (this would be a MANOVA, multivariate analysis of variance). I guess from your question that you are working at a more basic level. You can just do one ANOVA for each of the two outcome variables.
In either analysis, you might want to do follow up (post hoc) tests to examine whether pairs of group means differ:
pressure 1 vs pressure
pressure 1 vs control
pressure 2 vs control
In addition, for either of these analyses, you can report effect size information. The effect size h2 (pronounced eta squared) can be obtained by computing SSbetween groups/ SStotal. This is interpreted like a squared correlation. That is, if eta squared = .30, then 30% of the variance in outcome variable scores is associated with treatment group membership.
If you take the square root of eta squared, you obtain eta. Eta is like r, exept that r assumes a linear association between variables (type of treatment, outcome variable) and eta does not.
