All particle measurement systems divide the size (or shape measure) axis into small classes, and counts in each class are accumulated. The result is histogram data, which gives a good indication of the actual size or shape measure distribution. The larger the number of divisions, the better the accuracy of results will be. From the histogram data, the standard measures of the distribution center and spread can be calculated. These include Mean, Standard deviation, and Mode. The Particle Insight also determines Harmonic Mean, Coefficent of variance, and Skewness.
Another standard way of quantizing a distribution is to report cumulative percentiles. For example the 50th percentile is the value such that half of the sample has size (or other measure) below that value. It is typical to report the 10%, 50% and 90% percentiles. The PI can report five percentiles that are user-settable.
Instruments that can only determine numerical counts in the size classes, on the basis of a single size measure, report “number” data. These systems must assume that particles are spheres in order to convert number data into surface area or volume data. Such results can be highly inaccurate if particles are non-spherical. An image analyzer that has the ability to fit particles to shape patterns can determine area and volume data much more accurately. An analyzer that accumulates data on every particle can compute mean, standard deviation etc. directly from that data without the need for histograms, resulting in the highest possible accuracy of all measurement methods.
Also refer to Statistical Reporting Basics.