If you have performed a refinement, but find that there are additional peaks in the pattern that are not being indexed, you will likely want to identify the source for these peaks.
One possibility is that the symmetry of the material has been lowered. GSAS-II provides a fairly simple way to look for that using the visualization in the Unit Cells List data tree entry. This is discussed in §28.3.1.0.
More commonly, there is an unexpected phase present. In the case of neutron scattering, if it is also possible that the peaks are from magnetic scattering. To help identify an unexpected phase or to find the k-vector for the expanded magnetic cell, you will want to have accurate positions for the peaks that are not being indexed in the pattern. For this one can fit peaks to the difference plot, where the these extra peaks should be amongst the most promising features. This is done using “Extra Peak mode” which is enabled, either by clicking on the button labeled “Switch to Extra Peak mode” at the top of the data window, or set turn on the checkmark setting for the “Add impurity/subgrp/magnetic peaks.” The two options do exactly the same thing.
Once “Extra Peak” mode has been set, the intensity from the peaks in the peak list will be added to the computed pattern from the Rietveld (or Le Bail/Pawley) fit. These intensities will be plotted in green. The cyan difference curve will show obs-(calc+peaks)