Kinematic Models of Subjective Complexity in Handwritten Signatures

This research seeks to develop a model using kinematic features that best explains variability in Forensic Document Examiner’s (FDE’s) perception of complexity. One hundred and twenty-three individuals provided five genuine signatures each. All signatures were collected on a tablet running MovAlyzeR® software, which recorded the kinematic data while the subject was signing. Scanned copies of the signatures were sent to five FDEs who were asked to rate the complexity on a five-point scale. For the purpose of this study, complexity was defined as how difficult it would be to simulate the signature without detection by an FDE. The effects of signature style on stroke kinematics were statistically significant for several variables, so separate models were developed for the three signature styles. Kinematic models ranging from three to eight parameters explained 71%, 76%, and 79% of the variation in complexity judgment for text-based, stylized, and mixed-style signatures, respectively. Our results confirm those from previous work that the number of strokes (or the number of turning points) in a signature is a strong predictor of FDE complexity assessment, adding that other kinematic variables such as stroke velocity, slant, horizontal stroke length, and total length of all segments of the signature, together with number of strokes, account for a greater portion of variability in complexity perception. These results underscore the importance of sensitive kinematic analyses of temporal and spatial features in understanding signature complexity.