A cluster-analytically derived typology: Feasible alternative to clinical diagnostic classification of children?

Abstract

Despite numerous studies of the relationships among symptoms manifested by children and adolescents, there have been few systematic attempts to group individuals on the basis of the syndromes identified. In the course of a treatment evaluation project requiring classification of both child patients and untreated controls, the present study was conducted to determine whether replicable types of children and adolescents could be identified by the cluster analysis of their IJR Behavior Checklist profiles. One clinical sample (N = 185) and two mixed clinical and normal subsamples (Ns of 358 and 373) were cluster-analyzed separately. Seven types that replicated across at least two subsamples were identified: (1) High Assets, Flat Symptom Profile; (2) Sociopathic, Academic Problems; (3) Moderate Assets, Egocentric, Incontinent; (4) Insecure, Somaticizing, Underachieving; (5) Aggressive, Overreactive; (6) Schizoid, Withdrawn, Anxious, Bizarre; and (7) Diffuse, Mixed Pathology. The issue posed by the title of the article was explored in the light of the findings.

Overt and covert antisocial behavior: Child and family characteristics among psychiatric inpatient children

“…Inpatient populations are known to show somewhat diffuse psychopathology, as reflected in multiple diagnoses or mixed syndromes of individual patients (e.g., Lessing, Williams, & Gil, 1982). This diffuse dysfunction could interfere with generating reliable differences for specific diagnoses.…”

A PIC profile typology of children and adolescents: I. Empirically derived alternative to traditional diagnosis.

“…Edelbrock and Achenbach (1980) used a hierarchical centroid clustering algorithm with the intraclass correlation as the measure of profile similarity to identify six subsequently replicated profile types for boys aged 6-11 and 12-16, and seven profile types for girls aged 6-l I and 12-16. Lessing, Williams, and Gill (1982) presented another attempt at developing a clusteranalytically based typology. The data base for this investigation was the I JR Behavior Checklist Parent Form (Lessing, Beiser, Krause, Dolinko, & Zagorin, 1973).…”

A Two-Year Follow-Up of Boys with Aggressive Conduct Disorder

“…Factor analyses [1,2] of the symptoms of children attending psychiatric clinics show that the dimensions of aggressiveness and antisocial behavior are relatively indepen dent of each other. Similarly, cluster analy ses which separated children according to their behavioral profiles distinguish the ag gressive from the antisocial or delinquent [4,13]. Summarizing these studies and reana lyzing them, Loeber and Schmaling [ 15] used multidimensional scaling to show that overt behaviors such as fighting fall at one pole of a dimension of antisocial behavior, while covert behaviors such as stealing fall at the opposite pole.…”