Premium
Clinical trials based on patients' preferences A new test of statistical significance
Author(s) -
Acland J. D.
Publication year - 1961
Publication title -
clinical pharmacology and therapeutics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.941
H-Index - 188
eISSN - 1532-6535
pISSN - 0009-9236
DOI - 10.1002/cpt196125581
Subject(s) - statistical significance , concordance , test (biology) , clinical significance , statistical hypothesis testing , statistics , clinical trial , preference , psychology , statistical analysis , value (mathematics) , significance testing , medicine , econometrics , mathematics , biology , paleontology
When patients' preferences are used. to assess the value of alternative treatments in a comparative clinical trial, tied preferences are often obtained as a result of the poor discriminatory ability of the experimental subiects. It is maintained that the weight attached to a given preponderance of positive choices for one treatment should decrease as the number of tied preferences increases. However, tied preferences are neglected altogether in two recommended sequential statistical tests, while the correction for ties used in another statistical test, based on the coefficient of concordance, actually causes the significance leoels to increase with the number of ties. A new test of statistical significance is described which can be used in clinical trials based on paiients' preierences, if large numbers of ties are recorded. The presence of tied prejerences louiers markedly the significance leoels obtained in this test, The errors which may be caused by neglecting tied preferences are shoum to be quantitatively important.