Confidence Interval For Tmax [Nonparametrics]

posted by Helmut Homepage – Vienna, Austria, 2017-07-30 16:41 (2433 d 20:30 ago) – Posting: # 17630
Views: 4,799

Hi Mohamed,

first of all, please read this post and follow the forum’s policy (you confirmed with the registration that you have read and accepted it)! I had to edit all of your posts… :-(

❝ […] However, I found a SAS report that generate a Non parametric analysis test on Ln transformed Tmax;


tmax comes from an ordinal scale and therefore, follows a discrete distribution (dependent on the sampling schedule). Hence, a nonparametric analysis is correct in principle. However, transformations are only allowed for data coming from a ratio scale (continuous with a true zero point).
Therefore, the report you found is crap. SAS is a fine piece of software but only as good as the user. This one probably was not qualified for his/her job.

Common language will give you a hint whether an additive model (raw data) or a multiplicative model (log-transformed data) should be used. Every time you say “X is Y% of Z” or “X is Y% lower than Z” you should apply a log-transformation (e.g., Cmax). If you say “the difference of X to Z is Y” you should use the raw data (e.g., tmax).

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
22,957 posts in 4,819 threads, 1,636 registered users;
106 visitors (0 registered, 106 guests [including 3 identified bots]).
Forum time: 12:12 CET (Europe/Vienna)

With four parameters I can fit an elephant,
and with five I can make him wiggle his trunk.    John von Neumann

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