When it "interpolates" an X value from a standard curve, Prism finds the X value corresponding to the entered Y value on the best-fit curve. The line or curve doesn't go exactly through every experimentally measured point, so the interpolated X value for a Y value won't exactly match the observed X value for that Y. Some find it helpful to look at the patterns of how the actual X values differ from the back calculated X values. Prism does not do this automatically, but use these steps to get the job done.
In other words, the goal of back calculating is to interpolate each of the Y values which the model was fit to to find out what X value they correspond to. Then the goal is to compare those interpolated (back calculated) X values with the actual X values.
1. Enter the X and Y values for the standard curve.
2. If you want to back calculate Y values from entered X values: Select and copy the X values, and paste them lower on the table. If using Prism 6 or later, use Paste link so these pasted values will update if you edit or replace the original X values. When you paste link, the pasted values are surrounded by a blue box. Note that these rows will have only X values and no Y values.
If you want to back calculate X values from entered Y values. Select and copy the Y values, and paste them lower on the table. If using Prism 6 or later, use Paste link so these pasted values will update if you edit or replace the original Y values. When you paste link, the pasted values are surrounded by a blue box. Note that these rows will have only Y values and no X values.
3. Choose nonlinear (or linear) regression and choose the option to interpolate from a standard cuve.
4. Find the results tab with the interpolated values.
5. Make whatever graph or analysis you want to compare the back calculated values with the entered values. There don't appear to be any standard ways to do this.