Reproducibility and significance of measurements of sedimentary size distributions

Reproducibility of graphically determined statistical measures obtained from sedimentary particle-size distributions is affected by differences in techniques, laboratories, and operators. Major sources of error lie in the mechanics of preparing and using the graphs. Careful use of standard techniques, mutual calibration of results and the use of probability graph paper can reduce the differences in means and standard deviations obtained from different laboratories to where they do not exceed a relatively low tolerance. Greater caution is required to obtain comparable measures of skewness and kurtosis.