Investigating error in saw mark minimum kerf width measurements

J Forensic Sci. 2023 Jan;68(1):257-266. doi: 10.1111/1556-4029.15178. Epub 2022 Nov 21.

Abstract

Minimum kerf width (MKW) is often assessed in medicolegal cases of skeletal sharp force trauma and can provide information about the tool used. Previously published saw MKW research focuses on the relationship between MKW and saw blade set width (width of the saw blade including tooth set). Despite these publications using various measurement tools and methods to collect MKW, error in MKW measurements is not well-documented. The goal of this research was to investigate MKW measurement precision in terms of intraobserver error, interobserver error, and the effects of measurement modality (e.g., calipers versus stereo microscope). The study sample consisted of 351 incomplete cuts made with 27 diverse saws. MKW was collected using three methods: (1) inserting the internal caliper jaws into the incomplete kerf, (2) placing the external caliper jaws on the cortical surface, and (3) using a digital stereo microscope and associated measurement software. For each method, intraobserver error and interobserver error were assessed. Differences in measurements taken with the three measurement modalities were assessed for each observer as well. Relative technical error of measurement (rTEM) and coefficients of reliability indicate that internal caliper and stereo microscope measurements had the lowest intra- and interobserver error (rTEM = 3.72% to 6.15%; r = 0.98-0.996). External caliper measurements performed the worst (rTEM: intra = 8.53% and inter = 21.32%). There was higher precision between internal caliper and stereo microscope measurements than with the external caliper measurements. This research highlights the need for measurement standardization.

Keywords: dismemberment; forensic anthropology; measurement definition; measurement error; saw mark analysis; sharp force trauma.

MeSH terms

  • Corpse Dismemberment*
  • Humans
  • Microscopy*
  • Observer Variation
  • Reproducibility of Results