Profile classification strategies for the Personality Inventory for Children.

Abstract

Two profile classification strategies were examined for use with the 12 clinical scales of the Personality Inventory for Children (PIC) : reciprocal twopoint code types and code type factor clusters. The PIC profiles of 556 children, who comprised six of the criterion groups used in PIC scale development, and the 6-16-year-old normative sample (« = 2,390) were classified by each system. Differences in frequencies across criterion groups of all the examined classification categories were significant and conceptually meaningful. Only 12% of the normative profiles had two or more Personality Inventory for Children clinical scales with T scores greater than 69. Additional study of a heterogeneous sample of 691 child guidance protocols compared code type classification rates using traditional rules with a revised classification system based upon previous actuarial study of individual profile scales. The possible interpretive significance of obtained sex and age differences in code type frequencies is discussed, as are proposed efforts to develop optimum code type interpretations.

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

“…The criterion group composition for this cluster was relatively heterogeneous. Because the proportion of children who actually obtain within-normal-limits PIC profiles within a child guidance population is about 4% (DeHorn, Lachar, & Gdowski, 1979), it seemed likely that Cluster 1 contained a considerable number of at least minimally elevated profiles. In fact, 34% of Split 1 and 40% of Split 2, Cluster 1 protocols obtained spike profiles, in which only 1 of the 12 substantive scales was elevated beyond the normal range.…”

Factor analysis of the PIC profiles of clinic-referred children

“…It is reasonable to predict that profile analysis of the PIC will receive increasing attention in research and clinical practice in the future. In fact, efforts (Dehorn, Lachar, & Gdowski, 1979;Gdowski, Lachar, & Butkus, 1980) already have been reported in grouping PIC scales into clusters and using these clusters as the basis for developing actuarial classification and interpretive systems. However, the identified scale clusters so far have been formed mainly on a logical rather than an empirical basis.…”

Broad-band dimensions of psychopathology: Factor scales for the Personality Inventory for Children.

“…Initial reliability and validity data are presented in the PIC manual (Wirt et al,, 1977). Gdowski (1979a, 1979b) have specifically documented the predictive accuracy of the profile scales, and several ad-ditional studies have provided other evidence of test validity and generalizability (DeHorn, Lachar, & Gdowski, 1979;Kurdek, Blisk, & Siesky, 1981;Lachar, Butkus, & Hryhorczuk, 1978;Leon, Kendall, & Garber, 1980;Porter, 1980;Voelker, Lachar, & Gdowski, in press;Wirt & Lachar, 1981).…”