Binomial Attribute Sampling as an Aid in Defining a Representative Field of View in Microscopy
Statistical analyses are routinely applied to microscope data from micrographs (measurements, shape factors) and, using electron microscopes, from x-ray microanalytical systems (net counts, peak ratios). However selection of a field of view which is recorded on a micrograph and implied to be a representative image of the bulk of the sample (at that magnification) is done intuitively. Experienced microscopists do consider a number of factors in selecting and preparing a sample and then choose from many instrument operating parameters before they view the sample. At this point assuming that the best microscopy practice has been exercised there is still the problem of a great number of fields of view to select from. High magnifications exacerbate this selection problem: in an Ultra High Resolution Field Emission Scanning Electron Microscope sharp images of routine samples are easily obtained at magnifications of X100,000 - at this magnification, even with a small sample size (4mm × 7mm), there are 28 million mutually exclusive fields of view.