Sexual abuse, family environment, and psychological symptoms: on the validity of statistical control.