~geometric mean of range [Bioanalytics]

posted by Helmut Homepage – Vienna, Austria, 2014-06-18 18:07 (3997 d 06:20 ago) – Posting: # 13084
Views: 14,488

Hi Nobody,

❝ And the weighing is applied for heteroscedasticy, […] weighing increases accuracy of samples calculated from the calibration function, especially in the lower range.


Correct. Now think about the confidence bands of the regression: They are two hyperbolas. The confidence band is narrowest at x|y. Both with geometric spacing of calibrators and weighting we shift the means towards the lower end. Therefore, the accuracy close to the LLOQ increases (narrower CI). But there is no free lunch. Accuracy at the upper end decreases.

❝ So, where am I wrong? :-)


In my example it might matter whether we look at x=16 or x=65. ;-)

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
23,424 posts in 4,927 threads, 1,675 registered users;
68 visitors (0 registered, 68 guests [including 14 identified bots]).
Forum time: 00:27 CEST (Europe/Vienna)

Every man gets a narrower and narrower field of knowledge
in which he must be an expert in order to compete with other people.
The specialist knows more and more about less and less
and finally knows everything about nothing.    Konrad Lorenz

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