
Self reference and familiarity in handwriting recognition
Author(s) -
Walter Gerbino,
Elisa Mattaloni
Publication year - 2011
Publication title -
journal of vision
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.126
H-Index - 113
ISSN - 1534-7362
DOI - 10.1167/11.11.827
Subject(s) - handwriting , psychology , computer science , speech recognition , cognitive psychology , artificial intelligence
We studied handwriting recognition in an X-AB task, by asking participants to discriminate if a handwritten target word was produced by the same writer as one of two comparison words, shown in left/right positions. Three types of writers have been considered: self, best friend, stranger. We collected a large sample of handwritten words by 10 pairs of best-friend children, attending 3rd and 5th primary school classes, and tested every child individually after 1 and 2 months, using the following AB pairs: self vs. best friend; self vs. stranger; best friend vs. stranger; stranger vs. another stranger. Therefore, discrimination could benefit from three factors: agency (present only in self handwritings), familiarity (present in self and best-friend, but not stranger, handwritings), and form similarity (differentiating various handwritings). Agency and familiarity effects were revealed by the superiority of performance in trials in which self and best-friend handwritings were compared to stranger handwritings, over those involving the handwritings of two different strangers. Taking discrimination accuracy by stranger participants as a dissimilarity value, we used a multidimensional scaling approach to find the relative distances of the 20 handwritings. Such distances were significantly distorted by agency and familiarity (i.e.; distances were modified when AB pairs involving self and best-friend handwritings were considered). Children data confirmed previous results obtained in our lab with university students, using a similar paradigm, as well as conclusions by Chen et al. (2008) in their study on the identification of Chinese handwritings