An appraisal of the computation of statistical parameters in grain size analysis

A computer program has been written to overcome the limitations of two published programs that compute statistical parameters in grain size analysis. This program uses all available data, obtained from standard sedimentological analyses, regardless of phi interval, and computes statistical parameters by direct integration of a linear approximation to the cumulative curve. This has the advantage that the distribution curve is known and hence the limitations of the parameters obtained can be assessed. Furthermore, it is argued, with examples, that this approximation introduces no greater degree of inaccuracy than is already present in standard sampling and analytical procedures.