Computerized content analysis of natural language

摘要

The Gottschalk-Gleser Content Analysis Scales have been found to be a transculturally valid and reliable method of measuring the magnitude of a group of psychological states and traits from the content analysis of natural language, and as such these Scales have been found quite useful in medical and psychiatric research. Their use, however, requires time-consuming training of content analysis technicians to code scorable clauses at a correlation level of at least 0.80 with experts. A review of research aimed at the development of computer software, based on LISP, to score typescripts of speech samples on the Gottschalk-Gleser Scales has shown that such computerization is feasible and efficient for the Anxiety and Hostility Scales. The Cognitive Impairment and Social Alienation - Personal Disorganization Scales, however, require more study and research.