Publication Type: Individual Poster
Abstract: Background and Aims: After Wynn (1992) reported that 5-month-olds can “compute the precise results of simple additions and subtractions,” her phenomenon was replicated in several laboratories (but not in some). Nonetheless, her interpretation of her data remains controversial. The current study evaluated the hypothesis that infants’ behaviors are best explained by their perceptual familiarity with the test displays, not computation.

Methods: This study was identical to Moore (1997)—which replicated Wynn using digital displays of checkerboards—with one exception that did not alter the situation’s computational characteristics: the 89 participants were familiarized with the “computationally unexpected” displays they would see in some test trials. In this situation, infants in a ‘1 + 1’ condition should prefer 1-object test displays if they are adding (because a 1-object result is unexpected) but they might prefer 2-object displays after being adequately familiarized with the 1-object displays. Likewise, infants in the ‘2 - 1’ condition should prefer 2-object displays if they are subtracting, but 1-object displays if familiarization with 2-object displays influences their behavior.

Key Results: Looking time analyses revealed a 4-way interaction that was interpretable: while the girls' means reflected the 3-way pattern predicted by Wynn (1992), the boys' means reflected the 3-way pattern predicted by the familiarity hypothesis, F(1, 85) = 4.80, p < .04. Although neither males nor females in the “addition” condition differed from those in the “subtraction” condition, differences emerged when the pattern of the boys’ and girls’ means were considered relative to each other. Girls in the “addition” condition increased looking (from baseline to test) at 1-object displays, whereas girls in the “subtraction” condition increased looking at 2-object displays; conversely, boys in the “addition” condition increased looking at 2-object displays, whereas boys in the “subtraction” condition increased looking at 1-object displays.

Conclusions: Girls’ behaviors in Wynn-style experiments apparently do not reflect their familiarity with the test displays; even after familiarization, girls’ behaviors after “addition” and “subtraction” events supported Wynn’s (1992) hypothesis, not the familiarity hypothesis. Boys, in contrast, were influenced by familiarity; their data do not indicate the formation of expectations based on computations. This sex difference probably reflects the faster rate of girls’ brain development; relatively immature boys might be more likely to respond to a display’s perceptual characteristics, whereas girls might be beginning to apprehend such displays with emerging cognitive skills. It remains premature to draw firm conclusions about the computational abilities of 5-month-olds in general.