Logger Pro calculates the uncertainty of the parameters in fitted lines and curves using all of the data points by means of standard curve fitting algorithms. Conceptually, as a fit parameter is less constrained by the data points, typically due to scatter of the points about the fitted line, the lower constraint is reflected in a larger uncertainty.

Some specific curricula use an alternate technique for estimating the minimum and maximum slope of a fitted line using only two points, the left- and right-most points, along with their uncertainty ranges. Logger Pro does not directly support this technique, but it can easily be performed for linear fits by using the Moveable Fit feature.

To do this,

1. Construct a graph of your data with error bars plotted.
2. Perform a standard linear fit to get the slope of the fit representing all the points.
3. Choose Model from the Analyze menu, and select the linear equation; note the inserted helper object with slope and intercept displayed.
4. Double click on the helper object and select the option for Enable Line Drag. The movable linear fit can now be adjusted by dragging one of the three handles on the line. The center handle adjusts the Y-intercept, the outer two handles can be used to rotate the line about the far edge. If the X axis is scaled from zero, then the right handle will control the slope while leaving the Y-intercept alone.
5. Using the draggable line handles, move the line so that it runs from the top of the error bar on the left-most point to the bottom of the error bar on the right-most. The slope of this line is your minimum slope.
6. Add another Model and set it to use line drag.
7. Using the draggable line handles, move the line so that it runs from the bottom of the error bar on the left-most point to the top of the error bar on the right-most. The slope of this line is your maximum slope.

Note that if the whole set of points is used for determining the best slope, and the above method is used to determine the min and max slopes, it is possible to have the best slope not be between the min and max slopes. This can happen since the min and max are determined using only two points, and not all the data.