It is well known that a gender difference in physical aggression emerges by the preschool years. We tested the hypothesis that the gender difference is partly due to changing tactics in peer interaction. Observations of girls' and boys' social initiatives and reactions to opportunities for conflict were made, using the Peer Interaction Coding System (PICS) in four independent samples of children between 9 and 36 months of age, which were aggregated to form a summary data set (N= 323), divided into two age bands (below or above 24 months of age). Linear mixed-model analyses revealed significant age by gender interactions in the use of bodily force in response to peers' initiatives and in the tendency to use bodily force at later stages of conflicts with peers. The gender difference in use of force was not explained by differences in the use of verbal tactics. These cross-sectional findings suggest that girls are initially more likely than boys to use reactive aggression, but then desist, whereas boys increase their use of force to defend their territory and possessions. The difference between older and younger girls likely reflects girls' abilities to regulate their behaviour in response to social challenges and the fact that girls are explicitly socialized to yield to peers' demands.

Observations of 12 pairs of 6-month-old infants interacting in a laboratory playroom in the presence of both mothers indicated that the estimated frequency, duration, and temporal distribution of one infant's distressed vocalizations were statistically independent of those of the peer. The absence of toys in the environment was a reliable predictor of the extent of an infant's distress, whereas the peer's estimated frequency and duration of distress and the infant's own gender and locomotor status were not. Infants watched their distressed peers during a majority of distress episodes and occasionally directed other behaviors to them but only rarely became distressed themselves. However, there were some indications in the pattern of distress in the trial when toys were absent that suggest one infant's distress had cumulative effects on the peer. Such cumulative effects would seem to be tempered by environmental conditions and the reactions of caregivers.

12 pairs of previously unacquainted 6-month-old infants were observed in both the presence and absence of toys. Interactive episodes that began when an infant touched the peer or a toy held by the peer were examined. The infants' immediate reactions to these events, which provoke conflict between toddlers, were neutral or positive. Furthermore, the patterns of occurrence of these behaviors indicated social influence over time: the outcome of 1 episode influenced the initiation of the next, and, when toys were present, 1 infant's tendencies to touch the peer and the peer's toys could be predicted from his or her partner's corresponding tendencies. The conclusion from previous research that in the absence of toys infants explore other infants needs further consideration: the frequency of peer contact when toys were absent was not reliably predicted by an infant's tendency to play with toys when they were present. It appears that the increase in peer contact that occurs in the absence of toys is mediated by corresponding increases in individual infants' gestural activity, which increases the likelihood of contact; however, gestural activity does not predict the initiation of contact when toys are present and does not predict an infant's tendency to reciprocate the peer's overture in either trial.

Observations of 12 pairs of 6-month-old infants interacting in a laboratory playroom in the presence of both mothers indicated that the estimated frequency, duration, and temporal distribution of one infant's distressed vocalizations were statistically independent of those of the peer. The absence of toys in the environment was a reliable predictor of the extent of an infant's distress, whereas the peer's estimated frequency and duration of distress and the infant's own gender and locomotor status were not. Infants watched their distressed peers during a majority of distress episodes and occasionally directed other behaviors to them but only rarely became distressed themselves. However, there were some indications in the pattern of distress in the trial when toys were absent that suggest one infant's distress had cumulative effects on the peer. Such cumulative effects would seem to be tempered by environmental conditions and the reactions of caregivers.

12 pairs of previously unacquainted 6-month-old infants were observed in both the presence and absence of toys. Interactive episodes that began when an infant touched the peer or a toy held by the peer were examined. The infants' immediate reactions to these events, which provoke conflict between toddlers, were neutral or positive. Furthermore, the patterns of occurrence of these behaviors indicated social influence over time: the outcome of 1 episode influenced the initiation of the next, and, when toys were present, 1 infant's tendencies to touch the peer and the peer's toys could be predicted from his or her partner's corresponding tendencies. The conclusion from previous research that in the absence of toys infants explore other infants needs further consideration: the frequency of peer contact when toys were absent was not reliably predicted by an infant's tendency to play with toys when they were present. It appears that the increase in peer contact that occurs in the absence of toys is mediated by corresponding increases in individual infants' gestural activity, which increases the likelihood of contact; however, gestural activity does not predict the initiation of contact when toys are present and does not predict an infant's tendency to reciprocate the peer's overture in either trial.