topic Re: Comparing mean values across quintiles statistically in SAS Programming

Comparing mean values across quintiles statistically

Corinthian94 — Mon, 21 Mar 2022 15:07:22 GMT

Hi all,

I have a dataset I'm performing an analysis on with a continuous outcome and categorical exposure that is parsed into 5 quintiles. I'm looking to compare the mean value of the outcome in each quintile to each other statistically but am unsure on the best way to do this in SAS. Would anyone have some advice on this? I used the following code to attain the means across quintiles for the outcome variable but have not been able to figure out how to compare them.

proc glm data=thomas.ecl_analytic;

class ecl_nosmoke_cat;

model weight_change_kg = ecl_nosmoke_cat;

means ecl_nosmoke_cat;

run;

Thank you in advance for the help!

Re: Comparing mean values across quintiles statistically

PaigeMiller — Mon, 21 Mar 2022 15:10:16 GMT

means ecl_nosmoke_cat/lines t;

allows the pairwise comparison of the mean via a t-test.

Re: Comparing mean values across quintiles statistically

sbxkoenk — Mon, 21 Mar 2022 15:19:22 GMT

On top of the answer by @PaigeMiller ,

you might want to :

check the Statistical Assumptions for Using PROC GLM
... like requesting a homogeneity of variance test
use LSmeans instead of arithmetic means
correct for the inflation of the type I ERROR probability when doing multiple comparisons (I do not think the t option for performing pairwise t tests is doing that automatically).

Thanks,
Koen

Re: Comparing mean values across quintiles statistically

PaigeMiller — Mon, 21 Mar 2022 16:38:26 GMT

@sbxkoenk wrote:

On top of the answer by @PaigeMiller ,

you might want to :

check the Statistical Assumptions for Using PROC GLM

... like requesting a homogeneity of variance test

use LSmeans instead of arithmetic means

correct for the inflation of the type I ERROR probability when doing multiple comparisons (I do not think the t option for performing pairwise t tests is doing that automatically).

Thanks,
Koen

Three good points, and one I am skeptical of. For a one-way design, are not the means and LSMeans equal?

Re: Comparing mean values across quintiles statistically

sbxkoenk — Mon, 21 Mar 2022 17:14:06 GMT

Three good points, and one I am skeptical of. For a one-way design, are not the means and LSMeans equal?

I may be completely mistaken, but in case there are missing data (missing values), I think Means and LSMeans differ (even for a one-way design).
No time to test it thoroughly now, so definitely not for blindly taking it for granted !!

But it is a fact, indeed, that increasingly complex treatment arrangements and experimental designs and also covariates being present (ANCOVA) gives more reason to use LS-means, rather than ordinary means.
And a one-way design is not exactly 'complex'.

Thanks,

Koen

Re: Comparing mean values across quintiles statistically

Corinthian94 — Tue, 22 Mar 2022 13:22:17 GMT

Thank you both! I was able to figure it out with your advice and some help from my advisor.