Aggression is generally defined as action aimed at harming another person (Perry, Perry, and Boldizar 1994), and has been identified as a correlate of antisocial and delinquent behaviors.  Aggressive, disruptive behavioral responses as early as the first grade have repeatedly been shown to be important antecedents of later aggressive behavior, juvenile delinquency, conduct disorders, and drug abuse (Block, Block, and Keys 1988; Ensminger, Kellam, and Rubin 1983; Farrington et al. 1991; Kellam et al. 1983; McCord, 1988; Robins 1978; Schwartzman, Ledingham, and Serbin 1985; Shedler and Block1991;Tomas, Vlahov, and Anthony 1990; Tremblay et al. 1992).  Follow-up studies of aggressive children have revealed that they are as adults more likely to exhibit alcoholism, accidents, unemployment, divorce, and physical and psychiatric illness (Caspi, Elder, and Bem 1987; Farrington, 1983). In particular, aggressive children have been found to experience major adjustment problems in the areas of academic achievement and peer social relations (Kazdin 1987; Walker et al. 1987).

Many empirical studies have identified family variables as consistent covariates for early forms of aggression and later delinquency.  Families of antisocial children have been found to employ harsh and inconsistent discipline, have little positive parental involvement with the child, and exhibit poor monitoring and supervision of the child’s activities (Loeber and Dishion 1983; McCord, McCord, and Howard 1963).  Patterson et al. (1992) maintain that the relationship between a child’s coercive behavior and parental coercive behavior is reciprocal, with each magnifying the other.  They suggest that aggressive children encounter difficulty in school with teachers and peers, which exacerbates their own maladaptive tendencies.

There are gender differences in the display of aggressive behavior, although the research findings on this point are complex.   Maccoby and Jacklin (1974) concluded from an extensive review of the research on gender differences in aggression, that boys exhibited more aggressive behavior than girls from the age of two to 3 and continuing through life.  This gender difference is furthermore found across a variety of cultures.  Hyde (1984) conducted a meta-analytic survey a decade later, and found that gender differences in aggression are not large, but are well established.  Gender accounted for an average of five percent of the variance in aggression found in the combined-sex populations of the 75 studies in her sample.  The mean for males was on the average approximately a half standard deviation higher than the mean for females. The gender difference tended to be larger for children than for college students (accounting for 7 percent of the variance in studies of children and only 1 percent of the variance in studies of college students).    Hyde also noted that gender differences were larger when aggression was measured by direct observation, projective methods, or peer reports than when measured by self-reports, parent, or teacher reports.   The consensus among reviewers was that male-female differences in aggression are largest and most consistent for physical aggression (Hyde 1984; Maccoby and Jacklin 1974). However, Archer and Westeman (1981) found that much of the apparent difference in physical aggression may stem from extremely aggressive behavior by a few boys.   And the research by Murray et al. (1998) reports that much of the gender difference in aggressive behavior among younger students is due to the greater rates of boys’ physical destructive behavior. Since boys gradually decrease their use of physical aggression over time, by the 6^th grade the rates of aggression for boys and girls are similar. 

We also found little convergence in the aggressive behavior of boys and girls during the middle-school years.  See the figure just below which shows the lower overall aggression rates of girls, an increase over time for both boys and girls, but little convergence in the aggression rates in the latter years.

The next figure just below shows aggression means for different kinds of aggressive behavior.  It indicates that the boys rate higher for all types of aggression including relational aggression (teasing and yelling), and that the gap between boys and girls is similar across the types of aggression including property destruction.  It would suggest that boys and girls do not differ in the type of aggression they display, only in the frequency with which they display it.