Quality weight
Select the Quality weight you want to use for configuration optimization.
The following options are available from the drop-down list:
- Coefficient of determination
- F-ratio regression
- Slope
Coefficient of determination
The configuration optimizer selects ranges with higher R² value.
Important: The Coefficient of determination is the default Quality weight option of PLA. If you consider using one of the other options, you must confirm suitability of your method for you assay.
F-ratio regression
The configuration optimizer selects ranges with higher F-ratio regression.
Note: With the F-ratio regression option, you must select the Residual-error ANOVA option for ANOVA model. If you use the Pure Error ANOVA option, the calculation is aborted with an error message. If F-ratio regression cannot be calculated, the value is set to '0'.
Slope
The configuration optimizer selects ranges with greater absolute value of the Slope of the restricted regression model.
Note: The Quality weight is used during range selection. It only affects the calculation if the 'Optimization on' option is set in the current Data selection scheme.
The behavior of Quality weight also depends on which option you select for the Configuration optimizer allocation range.
- If you select Best range: The configuration optimizer first finds all configurations with minimal severity index. Of these, the configuration with optimal Quality weight is determined.
- If you select Maximum range: The configuration optimizer first finds all configurations with minimal severity index. It then selects the configurations with maximal number of dose steps. Of these, the configuration with optimal Quality weight is determined.
This element has the following attributes:
| Data type | Multiplicity | Usage | Default value |
| xs:token | 0...1 | optional | none |
Context:
/Quantitative response assay/Analysis/Data selection schemes/