Research on affective disorders in childhood and adolescence has appeared with increasing frequency in the psychiatric and pediatric literature. The active pace and ferment of research on this topic highlights the importance of accurate and reliable assessment for proper psychiatric diagnosis, for the identification of diagnostic subtypes in epidemic, genetic, and biologic studies, and for the measurement of severity of illness and treatment response.
Current objective techniques for assessing depression in child and adolescent psychiatry include a broad variety of measures. However, most of these fall into two main categories: self-report measures and structure interviews. Unfortunately, in many if not most research publications, the rationale for the choice of the particular instrument used to assess affective disorder is not explained. This may be due to the wish to focus on the primary research objective, combined with a readiness to rely on instruments that already appear established by wide use in the field.
Serious problems remain, however, in generating reliable and valid research assessments for affective disorders in children and adolescents. These include: (1) Selection of the content and format, including choice of symptoms or behaviors assessed, calibration of the range of response, and the time frame being reported. Also included here is the setting in which the information is being obtained. (2) Selection of the informant(s) best suited to provide this information (i.e., child, parent, teacher, nurse, peer). (3) Determination of the technique of administration so as to obtain the highest possible quality of information (i.e., when, where, by whom, and both the length of time and degree of effort required of the subject). Also included here, for both self-report and structured interview, is the choice and complexity of language used to probe for specific information.
The advantages of both self-report and interview techniques have been extensively reviewed elsewhere (Kazdin, 1981; Kazdin & Petti, 1982; Edelbrock & Costello, 1984; Costello, 1986). However, all these reviews conclude that further research on validation and an expansion of the range of assessment techniques is warranted This brings us to the objective of this annotation, which is to advocate direct, longitudinal observation as an addition to the available set of assessment measures for affective disorders in children and adolescents. To this end, we will briefly review the disadvantage of self-report and interview as assessment techniques, present the advantages of longitudinally based observations of affective symptoms, and review the handful of research reports in which this approach has been utilized successfully.
Limitations of Self-Report Measures
Despite their convenience and wide applicability, self-report measures have certain disadvantages. No self-report measure encompasses the entire range of affective symptoms described in DSM III-R (APA, 1987). Issues of compliance, the avoidance or denial of depressed affect (Cytryn & McKnew, 1974), and anxiety about revealing secrets or making mistakes (Birleson, Hudson, Buchanan, & Wolff, 1987) compromise these assessment modalities to varying degrees for different patients. Limitations in comprehension of items and reading ability related to chronological and developmental age are difficult to make allowance for or to interpret precisely. As a consequence, the assessment of affective disorder in special populations, such as noncompliant adolescents, preschoolers, and developmentally disabled children and adolescents continues to be problematic (Weismann, Oravaschel, & Padian, 1980; Kazdin & Petti, 1982). It is not clear to what extent the increase in affective symptoms that occurs with age is a measurement artifact and to what extent it reflects developmental/endogenous processes (Orvaschel, Weissman, Padian, & Lowe, 1981; Kazdin & Petti, 1982). …