The reduced sample consisted of 300 infants: 17 were excluded due to having low birth weight or premature birth status; 5 were excluded due to atypical developmental status (parents provided documentation of these issues at the 4-year lab visit); 1 was excluded due to not having usable EEG data at either 5 or 10 months; 87 were excluded due to not having one or more variables among the age 2 mediator and age 3 cognitive outcomes. The reduced sample had a similar demographic profile as the full sample: 51 percent were male; 75 percent were of White/non-Hispanic background; 73 percent had at least one parent with a 4-year college degree. Participants excluded from the sample tended to have parents who were less educated than participants included in the sample, t(408) = −2.29, p < .05. However, the two groups did not differ significantly in terms of sex, t(408) = 1.55, p = .12, minority status, t(408) = 1.56, p = .12, or temperament, t(394) = 1.35, p = .18. Of the 300, 295 participated at the 5-month lab visit (5
