Identify groups that differ statistically in gt_summary

Question

I just recently discovered gtsummary and I'm impressed by the ease of use and the amount of work that our team won't have to do creating summary tables of our results. Thanks!

My question: With more than two groups, the p-value included with add_p() refers to the global test. How can I obtain the information about the post hoc comparaisons between groups? In some journals, we see the use of superscripted letters.

I looked for the add_difference() option but it prints the difference between two groups. I also thought that add_q() would help me.

Thanks in advance!

score 0 · Answer 1 · answered Nov 09 '22 at 19:47

0

There is no built-in function for pairwise differences, but this post illustrates how to incorporate them How to calculate SMD between 3 groups or more?

answered Nov 09 '22 at 19:47

Daniel D. Sjoberg

8,820
2
12
28

Is there a way to identify only the groups that are statistically different? – Tchuvak Nov 10 '22 at 20:48
Yeah, just apply the function to the variables you want instead of all – Daniel D. Sjoberg Nov 10 '22 at 22:26

Paul Schmidt · Answer 2 · 2022-11-13T07:48:59.970

0

I have tried to summarize the what, how and why (not) about the "Compact Letter Display" you are talking about in this chapter: https://schmidtpaul.github.io/dsfair_quarto/summaryarticles/compactletterdisplay.html

Does that help?

edited Nov 13 '22 at 07:48

answered Nov 11 '22 at 14:41

Paul Schmidt

1,072
10
23

This is interesting, but I'm looking for a way to easily integrate the compact letters in a summary table. – Tchuvak Nov 12 '22 at 02:39

Identify groups that differ statistically in gt_summary

2 Answers2