The fixation task used in this study measures the amount of time a child fixates or looks at a given picture as it is repeated over six trials and then is followed by a novel picture on the seventh. Two series of slides were used. The first was a redundant nonsocial visual stimulus: six trials of a slide showing 20 chromatic straight lines and a seventh of chromatic curved lines. The second series, a social array, was a chromatic schematic representation of a family shown for six trials, and a seventh presentation of the same schematic without color. Each slide was shown for 30 seconds and followed by a black slide for 30 seconds before the next presentation. Fixation time was recorded by a stopwatch as the amount of time, in seconds, the child looked at each picture before turning away the first time. Three basic measures of attention were obtained: response decrement, response recovery, and mean initial viewing time. Intercorrelations for the measures are tabulated. The highest correlations are those involving mean initial viewing time. In an age x sex s SES analysis of variance on the longitudinal sample, separately by year, age differences were not significant for the nonsocial series for any measure, and for the social series, age was marginally significant for recovery scores in Year 2, with older subjects looking less. Results of a repeated-measures ANOVA (age x sex x SES) performed on the sample using scores for mean initial viewing time, recovery, and decrement scores for each series separately showed that the only variable to reach statistical significance was SES measured by amount of mother's education. (For related documents, see TM 003 174-180, 181-198.) (DB)