Geometric means and adjusted means (LSMs) [General Sta­tis­tics]

posted by Helmut Homepage – Vienna, Austria, 2018-07-19 15:01 (2402 d 15:00 ago) – Posting: # 19083
Views: 6,295

Hi Elena,

extending what Ohlbe wrote…

❝ The ratio in your table is not a geometric mean ratio and is therefore rightly not labelled GMR. It is a ratio of LSMEAN, not of geometric means. It makes no difference if your study is balanced (i.e. same number of subjects with sequence TR and sequence RT, in which case the LSMEAN is equal to the geometric mean), but if you have a drop-out and your study gets unbalanced you will get some differences.


See this post for an example of unbalanced sequences and especially the PS.

Dif-tor heh smusma 🖖🏼 Довге життя Україна! [image]
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes

Complete thread:

UA Flag
Activity
 Admin contact
23,380 posts in 4,914 threads, 1,665 registered users;
97 visitors (0 registered, 97 guests [including 15 identified bots]).
Forum time: 05:01 CET (Europe/Vienna)

When people learn no tools of judgment
and merely follow their hopes,
the seeds of political manipulation are sown.    Stephen Jay Gould

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5