Dixon test
Provides a statistical test for the detection of data outliers based on: Dixon, Wilfrid J. 1953. 'Processing data for outliers.' Biometrics 9 (1): 74-89. The Dixon test is also described in the United States Pharmacopeia: USP 28 NF 23 (2005), p. 2521.
The test is based on replicates. Therefore, no outliers are detected for dose steps with less than three observations. Sensitivity is defined by the respective alpha value.
Note: Our Dixon test implements corrections concerning the calculation of critical values based on Rorabacher.* Yet, for reasons of compatibility, we use the alpha level provided in the original Dixon paper where the test is defined as one-sided test. The significance level of the corresponding two-sided test is alpha/2.
* Rorabacher, David B. 1991. 'Statistical treatment for rejection of deviant values: Critical values of Dixon's "Q" parameter and related subrange ratios at the 95% confidence level.' Analytical Chemistry 63 (2): 139–146.
This element has the following attributes:
| Data type | Multiplicity | Usage | Default value |
| Node | 1...1 | required | none |
Context:
/Dose-response analysis/Analysis/Data selection schemes/Data selection scheme/alternatives/