Residual variance and fit probability, Logistic weighting – Bio-Rad Bio-Plex Software® Upgrades and Conversions User Manual
Page 216
Bio-Plex Manager Software 6.1 User Guide | Determining Outliers
204
Residual Variance and Fit Probability
Other statistical parameters that can be used to evaluate and adjust goodness
of fit include spiked recovery, residual variance, fit probability and weighting.
Residual variance measures how well a curve fits the individual data points. Fit
probability provides statistical evaluation of residual variance to determine
acceptability. The values of fit probability range from 1.0 (perfect fit) to
0 (no fit).
These fit probability and residual variance values are reported by Bio-Plex
Manager, as shown in this figure.
Figure 188. Fit probability and residual variance
Logistic Weighting
If you are using Logistic 5PL or Logistic 4PL, Bio-Plex Manager will
automatically weight the points in the curve. There are eight models for
calculating the variance in the logistic weighting algorithm.
The default setting for Variance Model is Power Law Variance. Weighting
coefficient a will be automatically calculated.
BioPlex_6.book Page 204 Friday, September 23, 2011 11:55 AM