More support for Patterson, DeBaryshe, and Ramsey's (1989) theory of deviant group membership comes from a study conducted by Fuligni, Eccles, Barber, and Clements (2001). A group of 1,253 adolescents completed surveys in the 7th, 10th, and 12th grades in order to determine the effect that peers have on adolescent adjustment during high school. The sample, taken from several schools in Michigan, was fairly balanced by gender (54% females and 46% males), but it was predominantly Caucasian American. Although 7th grade adjustment was a significant predictor of later behavior problems, being involved with deviant peers and low family cohesion each added significant variance to the model. The authors theorized that having low family cohesion results in adolescents becoming overly involved with their peers. If those peers are deviant, then negative behaviors will emerge. A study by Capaldi, Dishion, Stoolmiller, and Yoerger (2001) further supported the theory that deviant peer group membership is related to later conduct problems. A sample of 206 predominantly Caucasian American (90%) and low to working class (75%) fourth grade boys were followed prospectively over a 10 year period. Results showed that engaging in antisocial behaviors between 9 and 12 years of age was related to continued antisocial behavior, as were deviant peer associations in early to mid-adolescence. The deviant peer associations were subsequently related to antisocial behaviors in late adolescence (as were previous antisocial behaviors) and were indirectly related, through later delinquent behaviors, to aggression toward an intimate partner in early adulthood. Overall, this study shows that engaging in antisocial behaviors may result in future associations with deviant peer groups, which could exacerbate the adolescent's delinquent behaviors.