Volume 18, Issue 2 (April 1973)
Numerical Correlation and Evaluation in the Comparison of Evidentiary Materials
One of the neglected areas in criminalistic laboratory techniques is the ability of the criminalist to compare and evaluate, quantitatively, sets of numerical data. At the present time, there does not exist in practice any uniformly accepted approach to determining a “figure-of-merit” (or rather a quantitative, reliable expression of the degree of match) for sets of data. The need for such a technique arose in a task designed to apply the phenomena of thermoluminescence to criminalistics . This task, funded and supported by NASA, was conducted through the Civil Systems Program Office at the Jet Propulsion Laboratory. In this specific application, each piece of physical evidence provided, after processing, a continuous curve. In comparing the curve of an unknown with that of an exemplar, a need was established for a quantitative expression of the extent to which the two curves matched. A well established statistical procedure was applied and was found to be fully satisfactory in resolving the problem. The same technique, without modifications was also found to be applicable to the analysis of emission spectrography data, neutron activation analysis (NAA), gradient density measurements, and any other criminalistic technique where sets of numerical data are determined. The main attribute of this developed technique is that it allows the criminalist to make judgements on the quality of the evidentiary determinations.