Accuracy and Precision in Medical Researches;Common Mistakes and Misinterpretations

Siamak Sabour, Ommolbanin Abbasnezhad, Samaneh Mozaffarian, Hajar Nazari Kangavari

Background: When we use a single test in clinical care, for appropriate management of patients and correct diagnosis in clinical care the validity and reliability of that single test is important. The purpose of this study was to evaluate the statistical issues about validity and reliability used in medical papers focusing on common mistakes and misinterpretations. METHODS: The articles about validity and reliability published in PubMed in 2012- 2015, were searched using MESH term. 200 most relevant papers with our topic were reviewed for assessing the correctness of methodology and statistical tests used to assess validity and reliability. Results: our study showed that the clinical researchers make many mistakes in assessing of validity and reliability of a single test. In more than half of the papers the methodology and statistical tests used for evaluating of validity and reliability of a single test were incorrect or incomplete. Conclusion: In analysis of validity and reliability of a test in published papers, there are many mistakes and clinical researchers need to gain more knowledge about that.