randombadger ☆ UK, 2011-06-29 18:20 (5108 d 01:30 ago) Posting: # 7195 Views: 4,222 |
|
Hi all, I'd like to calculate the SE of the Geometric LSM however my colleague and I disagree on which formula is correct! Based on the logged data, we have the LSMean & SE from the Mixed Model. In the past I've calculated SE Geom LSM = EXP(LSMean) * SE (based on former colleague's recommendations) whereas my colleague suggests simply exponentiating SE. Any thoughts? Does anyone have a reference? Thanks, RB |