NCJ Number
190133
Date Published
2001
Length
43 pages
Annotation
Motivated by several rulings in U.S. courts concerning expert testimony in general and handwriting testimony in particular, this research sought to validate objectively the hypothesis that handwriting is individualistic.
Abstract
Handwriting samples were obtained from 1,500 individuals representative of the U.S. population with regard to gender, age, ethnic groups, etc. The analysis of differences in handwriting was done by using computer algorithms for extracting features from scanned images of handwriting. Attributes characteristic of the handwriting were obtained, including line separation, slant, and character shapes. These attributes, which are a subset of attributes used by expert document examiners, were used to establish individuality quantitatively by using machine learning approaches. Using global attributes of handwriting and very few characters in the writing, the ability to determine the writer with a high degree (98 percent) of confidence was established. Concluding that these results are statistically inferable over the entire U.S. population, the study validated the individuality hypothesis with 95-percent confidence. By considering finer features of the handwriting, the researchers believe individuality in handwriting can reach a confidence level near 100 percent. 3 tables and 21 figures
Date Published: January 1, 2001