Identification of Gender in Domestic-Cat Faces with and without Training: Perceptual Learning of a Natural Categorization Task

Autor: Paul C. Quinn, Vanessa Palmer, Alan Slater
Rok vydání: 1999
Předmět:
Zdroj: Perception. 28:749-763
ISSN: 1468-4233
0301-0066
DOI: 10.1068/p2884
Popis: Three experiments were conducted to determine whether human observers could identify the gender of 40 domestic cats (20 female, 20 male) depicted in individual color photographs. In experiment la, observers performed at chance for photographs depicting whole cats, cat heads (bodies occluded), and cat bodies (heads occluded). Experiment lb showed that chance performance was also obtained when the photographs were full-face close-ups of the cats. Experiment 2a revealed that even with gender-identification training on 30 (15 female, 15 male) of the 40 face close-ups, observers were unable to generalize their training to reliably identify the gender of the 10 remaining test faces (5 female, 5 male). However, experiment 2b showed that gender-identification training with the 14 most accurately identified faces from experiment 1b (7 female, 7 male) was successful in raising gender identification of the 10 test faces above chance. Experiments 3a and 3b extended this facilitative effect of gender-identification training to a population of animal-care workers. The findings indicate that, with appropriate training, human observers can identify the gender of cat faces at an above-chance level. A perceptual category learning account emphasizing the on-line formation of differentiated male versus female prototypes during training is offered as an explanation of the findings.
Databáze: OpenAIRE