Gender differences in physical aggression are thought to appear early, around 2 years old, and further differentiate with socialization.
Baillargeon et al (2007) interviewed mothers about their toddlers' behaviour, and concluded that gender differences in aggression exist as early as 17 months, and do not become more differentiated in the following year.
Crick et al (2006) suggests that in preschool, boys are more physically aggressive than girls while girls are more relationally aggressive* than boys.
New results are to be presented later this year at the Annual Meeting of the
American Public Health Association.
Anson et al (2007) found that increased language ability significantly reduced physical aggression in 4-year-old boys, but not girls, .
Anson et al (2007) also plan to report that gender differences in physical aggression in their african american sample did not appear until age 6.
*
Mounts (1997) gives Crick's definition of relational aggression as "behavior specifically intended to hurt another child's friendships or feelings of inclusion in a peer group."
I am very interested to see more about Anson's study, because it actually appears to rely more on observations of the child than on parent interviews. While I am sympathetic that it is easier to ask the mother (in 99% of cases, the person who spends the most time with the child, according to Baillarger's paper) whether their child manifests certain behaviors, reporting is so clouded with societal expectations of male and female behaviour that I'd be just as pleased with interviews with non-parents. If I hear one more parent say "anyone who doesn't believe in gender differences should meet my children," I may have to punch them. How's that for aggression?