Validation of Toolmark Comparisons Made At Different Vertical and Horizontal Angles.

Numerous studies have focused on determining whether objective statistical methods can be used to discriminate between known matches and nonmatches when comparing laboratory prepared toolmarks. This study involved an analysis of striated toolmarks made as a function of varying vertical and horizontal angles of attack. Comparisons based on experimental data show that replicate toolmarks from the same tool show high correlation values at identical vertical and horizontal angles, with the correlation decreasing as the angular difference increases, especially for horizontal angular changes. Comparisons between nonmatching samples produce low correlation values that remain unchanged as horizontal angular differences increase. While complete statistical separation was not achieved between matching and nonmatching samples, there is evidence demonstrating that toolmarks can be identified if the variation in horizontal angle is within 10°. The experiment shows that computer-aided comparison techniques could be viable for identification with the proper statistical algorithm.