A sample of 2,453 grade school children were followed into young adulthood through record sources. Teacher interviews provided information about low-peer-status children that was assessed in relation to subsequent delinquency for both sexes and young adult criminality for males. A multivariate design evaluated the joint effects of social class, a measure of family disturbance, and childhood problem behavior factors as antecedents of delinquency. Childhood aggression emerged as the most prominent antecedent factor for males but not for females. Social class and family disturbance were associated with aggression but did not have significant direct effects on delinquency. Aggression was related to severity of delinquency. Dispositional status, reflecting severity, was the best indicator of which delinquent males would have adult criminal records. A causal model is presented.

Fifty-one male patients with hospital diagnoses of schizophrenia in the military service were followed up through Veterans Administration hospital records 22 years after their initial service hospitalization. Results revealed that severity of thought disorder did not predict outcome, but measures of affectivity and interpersonal competence did so with high significance. The authors attempt to explain why affectivity has received so little attention in the research literature. They also suggest that recent research diagnostic criteria for schizophrenia are inadequate, because the criteria focus almost exclusively on the less predictive symptoms of disordered thought and exclude assessment of affectivity and interpersonal competence. Kraepelin (1919) and Bleuler (19SO) considered disorders of thought and of affect to be primary or fundamental symptoms of schizophrenia, and many researchers and clinicians since 'then have affirmed the importance of both deficits as central features of the syndrome. It is striking, therefore, that recent efforts to specify criteria for differentiating schizophrenia from other psychopathologies have focused almost exclusively on manifestations of thought disorder. Schneider's (1959) First Rank Symptoms do not include an assessment of affective deficits. Both Spitzer, Endicott, and Robins (Note 1) and Feighner, Robins, Guze, Woodruff, Winokur, and Munoz (1972) clearly enumerate the thought disorder symptoms necessary forThe major results of this study were presented at the meeting of the Eastern Psychological Association in Boston, April 1977. The authors wish to acknowledge the cooperation provided by the Veterans Administration Hospital, Bedford, Massachusetts. Special thanks are due Denny Marvinney, who was one of the research assistants for this project.Requests for reprints should be sent to Raymond

Reliability as measured by the extent of agreement is often a problem for complex global judgments. Empirically, the use of multiple raters improved reliability consistent with predictions from the Spearman-Brown formula. Implications for the reliability of clinical diagnosis are suggested.

A sample of 2453 grade school children were followed through record sources into young adulthood. Data on peer status and social class were collected, and a subset was subsequently located in high school yearbooks. Those with mental health treatment during young adulthood were determined. Significant relationships were found between peer status and high school activity, and young adult adjustment, with low peer status children having two to three times greater risk for mental health treatment contact in young adulthood.

A sample of child guidance clinic patients who received an adult diagnosis of schizophrenia were followed through record sources into middle adulthood. The determination of within-sample differences in long-term adult outcome provided predictive criteria. Childhood symptoms were combined into rationally derived symptom scales that were included in a IS-variable matrix. Factor analysis yielded four orthogonal factors: Unsocialized Aggressiveness, Low IQ -Poor School Achievement, Neurotic, and Schizoid. Clusters of types were derived from factor scores and related to adult outcome ratings. The schizoid type was related to poor outcome, as was the neurotic -low-IQ type. Unsocialized aggressive and neurotic types were associated with more favorable outcomes.