Classification (General, Including Diagnosis)

Introduction

This entry treats formal classification procedures, not psychological models of classification behaviour. After specifying the terminology, fundamental concepts of classification are briefly introduced, followed by a short review of the empirical basis of classification. Finally, assignment procedures and the evaluation of a classification system are emphasized.

According to Gordon (1996: 65) classification ‘is concerned with the investigation of a set of objects in order to establish whether or not they fall naturally into groups (or classes, or clusters) of objects with the property that objects in the same group are similar to one another and different from objects in other groups’.

Differential and clinical psychology frequently solve problems of classification in several different areas: persons are characterized by typologies, one finds classifications of tasks and situations, intervention procedures are analysed and ordered in this way, and above all diagnosis as the assignment of persons to (nosological) categories is based on classification. Many examples from psychology and the social sciences are provided by and referred to in Reinecke and Tarnai (2000). Several classificatory systems for clinical assessment are found in Baumann and Perrez (1990). These systems are classified as distortion of psychological functions (like learning, memory, sensory-motor skills, sleep, emotion and motivation), distortion of patterns of functions (neuroses, depression, psychosomatic, schizophrenic, distortions specific to children, adolescents and old people), and distortions to interpersonal systems (in school, work organizations or community).

Terminology

The objects mentioned in Gordon's definition are also called cases, persons, patients, clients, elements, units, exemplars, specimens or items. This reflects the interdisciplinary research tradition of the field. Physics provides in the table of elements a well known example, and biology can be seen as an ongoing struggle to give a systematic, i.e. taxonomic, overview of all living beings. In clinical psychology and psychiatry, DSM (Diagnostic and Statistical Manual of Mental Disorders) as well as ICD (International Classification of Diseases) are used extensively.

What Gordon calls ‘groups’ is usually called classes or, more formally, sets and partitions, or a taxon in biology. In the behavioural and social sciences, quite often neither the kind nor the number of classes are known in advance and have to be determined in the process of establishing a classificatory system. They are the target of permanent revision to incorporate the increasing knowledge in a subject area.

Searching for classes means analysing the relationships between objects. Two fundamentally different though not mutually exclusive bases exist for this analysis: either judgements of similarity are studied, or patterns of features describing the objects are compared.

Similarity may be either based directly on judgements, expert opinions etc. on likeness or of belonging together, or on confusion frequency or other behavioural observations. Or the similarities may be derived from co-occurrence patterns or correlations between properties. A large variety of coefficients exists to derive similarity measures on which procedures like cluster analysis or multidimensional scaling might be based.

The objects may be characterized by properties, traits, characteristics, symptoms, features or variables. They describe the variability between the objects by categories. These categories exist in at least two values (states, labels). The variables may be qualitative or quantitative, discrete or continuous, and of any scale level. Within the context of classification, it is useful to consider sets of variables, also called profiles, vectors, syndromes, or feature patterns.

...