The Measurement of Performance in Probabilistic Diagnosis
- 1 October 1978
- journal article
- research article
- Published by Georg Thieme Verlag KG in Methods of Information in Medicine
- Vol. 17 (04), 227-237
- https://doi.org/10.1055/s-0038-1636442
Abstract
Attention is focused on one important aspect of good performance in probabilistic diagnosis, the »reliability« (external validity) of probabilistic assertions: a diagnostic alternative claimed to be 90% certain, say, must occur neither more nor less than nine times out of ten on the average. Statistical measures are offered by which departures from such perfect reliability can be estimated, and statistical tests are developed in order to test the hypothesis of perfect reliability. The specific reliability defects looked for include overconfident diagnoses and so-called size bias (common diseases being over diagnosed). Reliability is contrasted with discriminatory power and other performance aspects. The illustrative data derive from a study of computer-aided diagnosis of the acute abdomen.Keywords
This publication has 6 references indexed in Scilit:
- Reliability of Subjective Probability Forecasts of Precipitation and TemperatureJournal of the Royal Statistical Society Series C: Applied Statistics, 1977
- Credible Interval Temperature Forecasting: Some Experimental ResultsMonthly Weather Review, 1974
- Multigroup diagnosis of electrocardiogramsComputers and Biomedical Research, 1973
- A computer diagnostic system for the diagnosis of prolonged undifferentiating liver diseaseThe American Journal of Medicine, 1969
- The Computer Evaluation and Diagnosis of Congenital Heart Disease, Using Roentgenographic FindingsRadiology, 1966
- Discrimination Analysis in MedicineJournal of the Royal Statistical Society: Series D (The Statistician), 1964