Premium
Testing accuracy of analytical methods by regression
Author(s) -
Kemény S.,
Deák A.,
Bánfai B.
Publication year - 2009
Publication title -
journal of chemometrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.47
H-Index - 92
eISSN - 1099-128X
pISSN - 0886-9383
DOI - 10.1002/cem.1219
Subject(s) - statistics , regression , range (aeronautics) , confidence interval , mathematics , interval (graph theory) , regression analysis , linear regression , statistical hypothesis testing , computer science , econometrics , combinatorics , engineering , aerospace engineering
The usual procedures for testing the accuracy of analytical methods are criticized. Testing the average bias, as well as zero intercept and unit slope are all questionable procedures. Application of the general regression hypothesis test is applicable, but its pair of hypotheses is not relevant. The relevant question is: if the bias exceeds an acceptable level or not? This is an interval hypothesis, that may be properly tested using the two one‐sided t ‐tests (TOST) approach. It is practically executed by calculating the 1–2 β confidence band for the bias, and checking if it is within the allowed interval throughout the concentration range. If the bias does not depend on the concentration, this method is identical to the TOST procedure for the average bias. Copyright © 2009 John Wiley & Sons, Ltd.