Estimated Means Independent Samples t-test

Discuss statistics related things
Post Reply
DrVangsness
Posts: 1
Joined: Thu Sep 22, 2022 11:42 pm

Estimated Means Independent Samples t-test

Post by DrVangsness »

My students and I were running an independent samples t-test on some data that contains an equivalent number of observations per group. We noticed that the group means reported in the t-test table did not match the group means reported in the descriptive statistics table when the data is split by group. This would seem to suggest that a different calculation is being performed to obtain group means across the two platforms. Is this the case? And if so, what is the formula being used by the independent samples t-test?
User avatar
reason180
Posts: 274
Joined: Mon Jul 24, 2017 4:56 pm

Re: Estimated Means Independent Samples t-test

Post by reason180 »

The "t test table" can be made to report the difference between the two means; it does not report each mean.
Within the t test analysis, you can opt to output a separate table of group statistics which include the each of the two group means. I have never seen them differ from what is output in the Descriptives analysis.
User avatar
MAgojam
Posts: 421
Joined: Thu Jun 08, 2017 2:33 pm
Location: Parma (Italy)

Re: Estimated Means Independent Samples t-test

Post by MAgojam »

Hey @DrVangsness,
t-test.png
t-test.png (140.89 KiB) Viewed 3340 times
in jamovi Cloud, Desktop (current and solid) I can't see what you have kindly reported.
If there are no restrictions, could you share the .omv file you used with your students?
Maybe also indicating PC OS and jamovi release.

TA, Maurizio.
Post Reply