Chronicity and the General Health Questionnaire

We propose a new scoring for Goldberg's (1972) General Health Questionnaire. We argue that the response 'no more than usual', to an item describing pathology, should be treated as an indicator of chronic illness rather than of good health, and we score these responses accordingly. We give evidence that this set of responses is associated with other measures of neurotic illness, and that the revised scoring provides a better prediction of caseness than the conventional scoring. The revised scoring is more strongly associated with trait neuroticism, and is more stable in repeated measurement. It is recommended in preference to the conventional scoring for most research and epidemiological purposes.