Graphical Evaluation of Certainty of Smokeless PowderPopulation Association Using Computer-Intensive Statistics

Published

October 16, 2008

Author(s)

William MacCrehan, J D. Secl, David L. Duewer

Abstract

Forensic evidential evaluations often require comparison of questioned and known measurements to determine if they have a common origin. Visual display of this comparison using both classical and computer-intensive statistics can be used to determine inclusion of the questioned measurement(s) into the known population. In this study, compositional measurements of commercial ammunition cartridge powder from two manufacturers were used to develop a graphical presentation of the population distributions. Ogive plots demonstrated normal distribution of all cartridge measurements in each of the ammunition boxes. The Monte Carlo Bootstrap (MCB) method of computer-intensive statistics was used to evaluate the two data sets. In MCB analysis, finite measurement sets are resampled to provide much larger data sets for evaluation of population distribution and confidence intervals. Using the MCB graphical approach, the concurrence of questioned measurement(s) with a defined fraction of the known population can be determined with a stated level of confidence. The two ammunition populations were evaluated by MCB using 1000 iterations of three, six, and nine measurements and confidence intervals were constructed. An artificial bimodal data set was also created from the two boxes to demonstrate the applicability of MCB analysis to non-normal population distributions.

Citation

Journal of Forensic Sciences

Pub Type

Journals

Keywords

classical statistics, computer-intensive statistics, forensic science, graphical data display, handgun ammunition, Monte Carlo bootstrap method, smokeless powder composition

Citation

MacCrehan, W. , Secl, J. and Duewer, D. (2008), Graphical Evaluation of Certainty of Smokeless PowderPopulation Association Using Computer-Intensive Statistics, Journal of Forensic Sciences (Accessed March 1, 2026)