Questionnaire/ scale or scoring development

In various fields of science, it is a common practice to use questionnaire based tool or approach for data collection. However in many studies, well validated questionnaires or scales tailored to local research settings are scarcely available and such researchers invariably develop their own customized scales or questionnaires. Importantly, use of previously established scale in newer setting or context and application of newly developed scale impose a rigorous development and validation testing so that such scale is accurate (valid) and precise (reliable).

Reliability testing of any questionnaire or scale include internal consistency reliability (e.g. by Cronbach alpha), test-retest and inter rater reliability (using correlation coefficient or Kappa agreement). Validity testing includes face and content validity, construct validity (using factor analysis), concurrent validity and predictive validity (by regression modeling).

In addition, use of a published English language questionnaire in native language (say Hindi or Urdu or Punjabi) would necessitate translation and back translation methodology. Finally, researchers are at times keen to convert their prediction model or risk factors into a scoring algorithm or composite score for future use.