borrowed notes is not significant (p > 0.05). notes alone, is significant (p = 0.004), while book alone vs.
For example, the Tukey HSD test, book alone vs. For each, you will see the results of each pairwise comparision. The output for these three tests is presented below. Notice that to do these tests you need to specify what level of a you want to use. Some of the most common are Tukey's HSD, Fisher's LSD, and Scheffe (a very conservative post hoc test). All you need to do is to click on the box of the kind of post hoc test that you want to do.
Two-way ANOVA allows us to add a second factor (independent variable) One-way ANOVA examines whether the means of two or more groups differ For instance, when I carry out a paired t-test in specific group of 9 subjects, in SPSS the degrees of freedom is 9 whereas it is between 89-102 in R (see attached images).SPSS for Windows: 2-way (independent groups) ANOVA Procedures I guess the crucial point might be the different degrees of freedom in both softwares. This confuses me and I believe I should get similar results as in SPSS, since only paired-sample t-tests are carried out.
#Threeway anova spss code code#
The code runs, but I get different results compared to the SPSS output (in terms of p-values and t-values). Ref <- lsmeans(anova_all,~recency*positivity|group)Ĭomps <- contrast(ref,method="pairwise", adjust="none")
Within=.("recency", "positivity"), between=.("group")) This is feasible with the following code and using the lsmeans and the afex packages: anova_all <- aov_ez("code", "err", stim.lng, +
Now, I want to do all possible post-hoc t-tests, separately for the three groups. The design is the following: Positivity (2 Levels) x Recency (2 Levels) x Group (3 Levels)Īs a result I have a three-way interaction. I did a between subjects repeated measurement ANOVA in R, giving me the same results as SPSS.