SEARCH

SEARCH BY CITATION

SYNOPSIS

The present study compared reported frequencies of headache symptoms assessed by dichotomous and interval rating scales. Reliability estimates, measured by per cent agreement, were obtained between the two scales for 26 common symptoms. Ninety-eight undergraduate students and 25 chronic headache patients completed the symptom checklist using both rating procedures. Results suggested very low agreement between scales. In addition, a curvilinear relationship was discovered between per cent agreement and subjects' ratings of degree of headache problem. Severe and non-problem headache groups were more reliable across scales than slight or moderate headache sufferers. Data also suggested wide variability in reliabilities for specific symptoms although symptom classes (vascular, autonomic, musculoskeletal) were similar in per cent agreement with a slight tendency for autonomic symptoms to be more reliable. Migraine patients were more reliable than chronic muscle contraction (CMC) individuals and these differences were consistent across symptom classes. Overall, the data support the hypothesis that an interval scale is a more sensitive index of headache symptomatology characteristics than dichotomous ratings. Results are discussed with regard to the role of reliability in questionnaire research and its facilitative effect in reducing error variance across headache investigations.