Displaying and Interpreting Curve Fit Results

Click the Results button to display the Curve Fit Results dialog for details on the results of the latest curve fit. This will display the Curve Fit Results dialog. Use What's This help for information on the individual controls for this dialog.

The results of the latest curve fit are displayed in a table in the Curve Fit Results dialog window. Click the Copy button to copy the results to the clipboard for use in another application. Click the Annotate button to copy the results as an annotation on your graph.

Note that the curve fit annotation on your graph is automatically updated if you perform another curve fit on the same data. However, it is not updated if you simply change a parameter.

Since the curve fit results table is tabbed, you can adjust the tabulator of the table by changing the default tab size for your annotations. Click here for more information on changing the tab size for annotations. In general, a tab size of about 10-12 characters works best.

To change which parameters are dispalyed, click the Options Command in the View menu, and click the Curve Fit tab. You can turn each parameter (including the index and comment) on and off.

Interpreting Results

The results table displays the latest estimates for all the bands, as well as the actual number of iterations taken to fit.

In addition, a goodness-of-fit value is displayed. This number is an error-mean-squared (EMS) value calculated at the conclusion of the fit. It is calculated by summing the squares of the differences between the experimental and curve fit data for each data point (each multiplied by a weighting factor defined as the reciprocal of the experimental data for that point), divided by the number of data points minus the number of parameters. The smaller the value, the less error in the fit.

For most XPS data, values of less than 10 are often desired, and values less than 2 or 3 are often attainable. However, note that the average intensity of the data affects the goodness-of-fit value. Higher intensity data may result in a higher goodness-of-fit value than the same exact fit for lower intensity data. Transforming your data from counts/seconds to counts will therefore affect the goodness-of-fit value if the average intensity of the data changes greatly. It is therefore important to use this value as a relative, as opposed to an absolute measure of the curve fit.

Also keep in mind that a lower goodness-of-fit value simply means that the artificial curve more closely matches the experimental data. It does not necessarily imply that the peaks associated with the bands accurately reflect the true chemical nature of the peaks present in the data.

Tips for Getting Better Curve Fits

Additional Curve Fitting Topics