U.S. flag

An official website of the United States government, Department of Justice.

Writer Identification Using Hand-Printed and Non-Hand-Printed Questioned Documents

NCJ Number
Journal of Forensic Sciences Volume: 48 Issue: 6 Dated: November 2003 Pages: 1391-1395
Moshe Kam Ph.D.; Erwei Lin M.S.
Date Published
November 2003
5 pages
This article discusses the proficiency of forensic document examiners (FDE's).
The proficiency of FDE's was the subject of several controlled studies conducted in the last decade. These tests examined the performance of FDE's and layperson in handwriting identification and in authentication of signatures. Recently, several Federal district court judges have required information on the abilities of FDE's in the specific area of hand-printed documents. A re-analysis of the data from the 1996 controlled proficiency test is presented here. This re-analysis treats hand-printed (HP) and non-hand-printed (NHP) documents separately. Information is provided on FDE and layperson writer identification performance using cursive (©) and non-cursive (NC) documents. Each test-taker was given an unknown package (six original handwritten documents, not necessarily by the same writer), and a database package (24 original handwritten documents, not necessarily by the same writer). The test-taker was asked to match each one of the documents in the unknown package to all the documents in the database package that were written by the same person. The re-analysis of this study showed that, for each group of test-takers separately, the Kolmogorov-Smirnov (KS) test did not detect statistically significant differences between data on HP documents and data on NHP documents. Within each document type the data provided by FDE's were found to be significantly different from the data provided by laypersons. The performance of FDE's was found to be much better than the performance of laypersons for both hand-printed and non-hand-printed documents. Similar results were obtained when cursive vs. non-cursive and hand-printed vs. cursive comparisons were made. Data supplied by each test-taker group for any pair of document types were not found by the KS test to be statistically different. When data from different test-taker groups were compared within the same document type, the data provided by FDE's and the data provided by laypersons were found to be significantly statistically different. FDE's had much better performance than the laypersons in each one of the handwriting types. 3 figures, 5 tables, 13 references