Infants were seated on a parent’s lap throughout the procedure, and parents listened to music over headphones so they were unaware of the auditory stimulus. Stimuli were presented on a 50 in. plasma monitor and stereo speakers using HABIT software (Cohen, Atkinson, & Chaput, 2004). Looking BMN 673 in vitro time was coded online by an experimenter blinded to both visual and audio presentation, and

inter-experimenter reliability for looking time was over 90%. The switch task was used (for a complete description of the task, see Werker et al., 1998). Infants were habituated to two objects paired with /buk/ and /puk/ in trials of a fixed length of 14 sec. When looking time reached 50% of the initial value over a four-trial moving window, the procedure automatically transitioned from the habituation phase to the test. Infants were then tested MG-132 datasheet on one of the objects in a same trial (the word–object pairing was the same as in habituation) and a switch trial (the pairing was switched). As is typical practice, both trials used the same visual stimulus, but the auditory stimulus

varied to either match or mismatch the object. After both experimental trials, infants were tested on a control trial, where a word from habituation was paired with a novel object to insure that the procedure was successful. Habituation trials were presented in pseudorandom order, with word–object pairing and test words counterbalanced across subjects. The same and switch trials were counterbalanced in the first two test positions, and the control trial was always presented third. Data were analyzed using a mixed design analysis of variance (ANOVA), with test condition (same, switch, and control) as the primary within-subject variable. We also included test order (same-first or switch-first) and science the word used for test (whether the same trial featured /buk/ or /puk/) as between-subjects factors. While these two factors were counterbalanced between subjects, it was important to demonstrate that they did not interact with our primary effect. We were particularly interested in the word used at test, as it was

possible that infants’ responses could have been affected by a preference for one of the words. This was important as one of our stimulus items, /buk/, is phonologically similar to “book,” a word known to 90% of children this age (Dale & Fenson, 1996). Lexical familiarity could have created difficulty mapping /buk/ because of lexical competition (Swingley & Aslin, 2007) or conversely could allow children to map the word more easily due to lexical support (Theissen, 2007). The analysis found a main effect of condition (same, switch, or control, F[2, 24] = 30.4, p < .001). Planned comparisons as shown in Figure 2 showed that the condition effect was driven entirely by looks to the control trial. The control trial was significantly different from same and switch trials, F(1, 12) = 57.1, p < .001, but there was no difference in looking time between same (M = 5.