Performance comparisons of Likert and binary version of SF-36 Version 1.6 across ECRHSII adults population.

SF-36 questionnaire is a multi-item multi-dimensional questionnaire using well-accepted Likert standards, but little attention is paid to the specific proprieties of the ordered categorical data recorded. An optimal scaling, that takes account of the non-metric properties of the data considering ordinal responses as nominal responses, was used to reduce SF-36 Likert item responses to give a smarter and faster binary form but with same excellent psychometric characteristics. The content and features of the binary form are compared with the original questionnaire regarding scaling assumptions, reliability, and physical/mental summary measures, and know-groups clinical validity. Data from ECRHS II survey of 12 countries (n=8854 form 25 European and extra-European centres) were analyzed to cross-validate the binary questionnaire and scoring algorithms. Results indicate that the changes to the binary response formats of the SF-36 scales meet at least the same standards without jeopardizing the underling structure of the original SF-36. This changes will likely further enhance the responsiveness of the SF-36. The respect of the most important SF-36 features, and the better quality offered information respect to the shorter versions, should make the binary version suitable in administering both in clinical field (clinical trials), heavy diseases, and by telephone way (reduction of times and costs).