Abstract
HYPOTHESIS: Basic visual functions develop rapidly during the first year of life. Since infants' endogenous attention system is not yet quite matured, visual salience has an almost exclusive role in controlling their visual attention. However, no previous research has attempted to systematically study the relationship between detectability and salience in infants and to compare the salience of different visual features.
METHODS: We measured detection thresholds and salience relations between iso-detectable stimuli using the forced-choice preferential looking technique in 5- and 6-month-old infants. Detection: Stimuli consisted of a dense 20x20 array of randomly oriented Gabor patches, where a 4x4 region that differed from the background elements in either color (red saturation: 6–31% , background: 0%) or spatial frequency (1.5–4.5 cpd, background: 1 cpd) appeared either on the left or the right side of the field. Salience: Two equally detectable (65% preference for color vs. SF) stimuli were pitted against each other on the same background as in the detection task. Gaze directions were coded in both experiments. A total of thirteen 5-month-old and twelve 6-month-old infants participated.
RESULTS: The detection thresholds measured were very similar in the two age groups for color and somewhat lower for the older infants for SF. In our preliminary results detectability nearly always predicted salience.
CONCLUSION: Salience preference seems to parallel the underlying maturation of the developing visual system. Our conclusions confirm and extend the predictions of an early model of infants' visual preference by Banks and Salapatek (1981). In our future studies we plan to explore other regions of the feature space.