A comment on Hirst's 'Evaluating the Malaise Inventory'

An item analysis of the Malaise Inventory by Hirst (1983), reported in this journal, suggested that there no was no evidence of a single dimension of emotional disturbance underlying the set of items. Hirst concluded that Malaise scores were untrustworthy for testing empirical hypotheses about degrees of stress. Consideration of Hirst's paper leads to the conclusion that an inappropriate methodology was used for the factor analysis, since the correlation coefficients were based on dichotomous items that were simplifications of continuous variables. A confirmatory factor analysis of the Malaise scores of 200 mothers of severely mentally handicapped children finds that the Malaise Inventory yields a single moderate factor of stress.