Estimate of the random match frequency of acquired characteristics in footwear: Part II - Impressions in dust

Abstract

This study serves as Part II of an investigation into the random match frequency of randomly acquired characteristics (RAC-RMF) in footwear evidence. In Part I, RAC-RMF was estimated in a dataset of laboratory-simulated crime scene impressions deposited in blood. For Part II, a second dataset was created composed of impressions deposited in dust on paper or tile, with the latter lifted using gelatin or Mylar film. A total of 1,513 RACs were identified from more than 160 dust impressions and compared to RACs with positional similarity in test impressions from 1,299 non-mated outsoles. RACs of any size deposited in dust exhibited a 31% decrease in shoes with non-zero RAC-RMFs as compared to their mated test impressions, while those deposited in blood exhibited a 45% increase. When only considering shoes with at least one RAC deemed forensically-reliable (length ⩾ 2.8 mm), 3.1% of shoes contributing dust impressions and 3.4% of shoes contributing blood impressions exhibited relative RAC-RMFs at a value ⩾ 0.0008. Although each dataset resulted in a comparable rate for encountering non-zero RAC-RMFs, the estimate for dust was based on twice the number of RAC comparisons (154,477) than those performed when assessing blood (77,566). While these results are considered specific to the non-mated impressions and methods of analysis described herein, and continued work is required before rates can be fully understood and reported in forensic casework, this study encountered non-zero RAC-RMFs for shoes exhibiting at least one forensically-reliable RAC at a more frequent rate than any estimates previously reported.

Keywords: Crime scene impressions; Dust impressions; Footwear; Percent area overlap; Random match frequency; Randomly acquired characteristics.