3.3.5. Differences between logit regression methods

The logit regression is an iterative calculation method. The calculated results may differ slightly per program used (Graphpad, SPSS etc.).

This depends on the algorithms used and the calculation accuracy of the program used.

This program searches for the Bmax that gives the highest correlation of the calculated calibration curve.

The decision when the correlations are equal and the iteration stops is determined by the number of decimal places with which the correlations are compared.

This logit regression calculates with the natural logarithm eLog (ln). If a 10 Log is used, the calculated slope a and intercept b parameters will differ.