Hi all

I have conducted a robust mixed design ANOVA using WRS2 in R. My data are time taken in seconds, and not normally distributed. I cannot remove outliers as the experiment is based around provoking outliers in performance. Being new to both R and this type of statistical analyses, if anyone could advise on post-hoc tests I would be grateful.

The WRS2 package uses Huber's Psi. I can generate an output no problem. However, I have no idea on how to report the post-hoc comparisons other than stating whether results are above or below the p-value. Can anyone please advise on how to report the output? Below is the R output for the between-,within-, and interaction effects.

Thanks in advance

Between effect:

sppba(formula = Seconds ~ Group * Phase, id = Participant, data = master,

est = "mom", nboot = 500)

Test statistics:

Estimate

Control -21.342

High -18.985

Low 2.357

Test whether the corrresponding population parameters are the same:

p-value: 0

Within effect:

sppbb(formula = Seconds ~ Group * Phase, id = Participant, data = master,

est = "mom", nboot = 500)

Test statistics:

[1] -5.117

Test whether the corrresponding population parameters are the same:

p-value: 0.004

Interaction:

sppbi(formula = Seconds ~ Group * Phase, id = Participant, data = master,

est = "mom", nboot = 500)

Test statistics:

Estimate

Control Time_1-Time_2 -19.1595

High Time_1-Time_2 -0.3791

Low Time_1-Time_2 18.7804

Test whether the corrresponding population parameters are the same:

p-value: 0

More Shane Barry's questions See All
Similar questions and discussions