Log in
Register|
sasikumar ☆ Tamilnadu, India, 2008-04-02 08:08 (6249 d 18:39 ago) Posting: # 1742 Views: 9,251 |
|
|
Dear all, Please provide the solution for the problem? While calculating CV, I found that two formula 1. CV=100*((Sqrt(MSE)/µR) 2. CV=(Sqrt(exponent(MSE)-1)) I hope that first one will be used if data is without log transformation and latter one will be used if data is log transformed. I have doubt whether both formula giving the same result or different? Thanks in advance, S.Sasikumar |
|
Helmut ★★★   Vienna, Austria, 2008-04-02 15:10 (6249 d 11:37 ago) (edited on 2008-04-16 10:30) @ sasikumar Posting: # 1749 Views: 8,347 |
|
|
Dear Sasikumar! ❝ 1. CV=100*((Sqrt(MSE)/µR) ❝ 2. CV=(Sqrt(exponent(MSE)-1)) ❝ I hope that first one will be used if data is without log transformation and latter one will be used if data is log transformed. Both your assumptions are correct! ❝ I have doubt whether both formula giving the same result or different? Results most likely will be different. The former is based on the assumption of a normal distribution, the latter on the assumption of a lognormal distribution. In the context of BE testing, there's a general consensus that concentration-based data (AUC, Cmax) follow a lognormal distribution; therefore the Amount-based data (Ae: cumulative amount excreted) follow a normal distribution (additive model); therefore the first one should be used. Edit: Wrong statement ( — Dif-tor heh smusma 🖖🏼 Довге життя Україна! ![[image]](https://static.bebac.at/pics/Blue_and_yellow_ribbon_UA.png) Helmut Schütz ![[image]](https://static.bebac.at/img/CC by.png) The quality of responses received is directly proportional to the quality of the question asked. 🚮 Science Quotes |
|
yjlee168 ★★★  Kaohsiung, Taiwan, 2008-04-16 08:19 (6235 d 18:27 ago) @ Helmut Posting: # 1776 Views: 7,818 |
|
|
Hi, HS ❝ ❝ 1. CV=100*((Sqrt(MSE)/µR) ❝ ❝ 2. CV=(Sqrt(exponent(MSE)-1)) ❝ Results most likely will be different. ❝ The former is based on the assumption of a normal distribution, the latter on the assumption of a lognormal distribution. ❝ In the context of BE testing, there's a general consensus that concentration-based data (AUC, Cmax) follow a lognormal distribution; therefore the first formula (after ANOVA of log-transformed data - multiplicative model) should be used. Then should this be the "second" formula? ❝ Amount-based data (Ae: cumulative amount excreted) follow a normal distribution (additive model); therefore the first one should be used. This should be the right one. YJ |
|
Helmut ★★★ Vienna, Austria, 2008-04-16 12:25 (6235 d 14:22 ago) @ yjlee168 Posting: # 1777 Views: 7,870 |
|
|
Dear YJ, you are exactly right! I will correct my first post in order to avoid confusion… Thanks a lot. — Dif-tor heh smusma 🖖🏼 Довге життя Україна! Helmut Schütz The quality of responses received is directly proportional to the quality of the question asked. 🚮 Science Quotes |
Ing. Helmut Schütz