Perceptual and Motor Skills, 1991, 73, 893-894. O Perceptual and Motor Skills 1991

DISCRIMINATING ARTISTS FROM NONARTISTS BY THEIR EYE-FIXATION PATTERNS ' JAMES R. ANTES AND ARLINDA F. KRISTJANSON

University of North Dakoto Summary.-Discriminant analysis was used to differentiate 15 artists from 15 nonartists on the basis of their eye-fixation patterns. contributing significantly to the discriminant function were fixation densities on the less important aspects of familiar and unfamiliar paintings.

Individual differences in patterns of attention have been investigated by examining the eye-fixation patterns of groups of individuals who differ from each other on some dimension of background and experience. What has not been reported is the ability of standard eye-fixation measures to discriminate between the different groups. This report describes use of discriminant analysis to distinguish artists from nonartists on the basis of their eye-fixation patterns, using the data from a study by Kristjanson and Antes (1989). In that study 15 experienced artists and 15 nonartists viewed a series of paintings for 20 sec. each. Three of the paintings were familiar to both groups and three were familiar to neither group. Eye movements were recorded as the subjects viewed the paintings and the measures were broken into the following variables: fixation density (proportion of fixations per unit area) on centers of interest (defined a priori by a professional artist) and noncenters, mean duration of fixation on centers and noncenters, mean pupil diameter when fixating centers and noncenters, and mean interfixation distance. A stepwise discriminant analysis was performed on these data, using Wilks' criterion for inclusion, to discover which, if any, of these variables would differentiate artists from nonartists. A significant discriminant function resulted = 8.80, p < .05), with a canonical correlation of 3 2 7 . Significant contributors to the function were density of fixation on noncenters of interest of unfamiliar paintings (standardized discriminant function coefficient = 1.27) and fixation density on noncenters of interest of familiar paintings (-0.82). Related to correct categorization of the artists were higher density of fixation on noncenters of unfamiliar paintings and lower density of fixation on noncenters of familiar paintings, and vice versa for nonartists. This discriminant function correctly categorized 76.7% of the subjects (66.7% of the nonartists and 86.7% of the artists).

(xl

'Correspondence should be addressed to James R. Antes, Psychology Department, Box 7187 University Station, Grand Forks, ND 58202.

894

J. R. ANTES & A. F. KRISTJANSON

The pattern for unf arniliar paintings may reflect the artists' training to examine more subtle aspects of an art work. Why this pattern was not demonstrated with familiar paintings is not clear. Perhaps the paintings were so familiar to the artists (American Gothic, Mona Lisa, The Last Supper) that, for the artists, there were no unexplored subtleties. Regardless of the specific interpretation of these results, this research suggests there is value is using discriminant analysis to elucidate individual differences in patterns of attention. Additional exploration might expand the definition of the centers of interest by asking more artists and those of varied backgrounds to identify centers of interest in other paintings depicting a wide range of content. REFERENCE ANTES,J. R. (1989) Eye movement analysis of artists and nonartists viewing paintings. Visual Arts Research, 15(2), 2 1-30.

KRISTJANSON, A. F.,

&

Discriminating artists from nonartists by their eye-fixation patterns.

Discriminant analysis was used to differentiate 15 artists from 15 non-artists on the basis of their eye-fixation patterns. Contributing significantly...
59KB Sizes 0 Downloads 0 Views