3.2.4. Review results

After pressing the “Logit” button, the slope, intercept, blank and Bmax (Rmax) regression parameters are calculated for an optimal fit for the calibration line.

Two graphs are shown.

The upper graph shows the dose-response points and curve after a logit-log transformation of dose.
The transformation formulae is printed along the axis.
The logit-log plot is a straight line.
This graph shows the data used for the calculation and shows the deleted points.
Use this chart to determine and remove outliers.

The lower graph is a representation of the values entered without transforming the dose and response.

The correlation is a quantity to judge how far the dose response points are from the calculated line.
With a correlation of 1.000 all points lie exact on the regression line and lower than a correlation of 0.000 in this program is not possible

Correlation is not a good parameter to use to assess a regression. This is because this parameter depends on the number of measurement points.

The correlation can be used in a validated test with a fixed number of measuring points for which it has been determined what the correlation should be for this test.

For an ELISA with five serially diluted measuring points in duplicate, an R> 0.9950 can be used. With six serially diluted measuring points in duplicate, an R> 0.9900 can be used.

If the correlation is lower, a outlier can be invalidated by making the measured value in the calibration line data italic or bold and pressing the “Logit” button.