PLA 3.0 is able to detect statistical outliers in your bioassays. Of course, outlier detection is optional, but if activated, the program supports four different tests for finding statistical outliers. All four tests can be executed iteratively during your biostatistical analysis. Statistical outliers will be flagged and the program excludes them from further analysis. There are also suitability tests available to ensure the validity of the results.
Treatment-based Outlier Detection
- Dixon Outlier Detection
- Grubb's Test for Outliers
- Standard Deviation Test
Model-based Outlier Detection
- Externally Studentized Residual
This test checks for the significance of an outlier on the regression model.
In addition to outlier detection any data point can be excluded from analysis by marking it as a technical outlier.