question about the Scherrer method

[Angel L. Ortiz]

Dear users of the Rietveld method:   I would like to launch a question regarding the use of the Scherrer method of line-broadening analysis. When estimating the crystallite size in a undistorted material (free of lattice microstrains) using Scherrer method, should I use the integral breadth of the pure profile or its full width at haft maximum (of course, both corrected by instrumental broadening)?. Why?.   Thanks in advance.   Angel Luis.