Bias in Studies on Diagnostic Performance
Introducing Chapter 8, we emphasized that statistical testing of the null hypothesis H0 and potential acceptance of the experimental hypothesis H1 can be reliably done only if a study is not flawed by substantial biases, namely systematic distortions which could explain the observed difference as an alternative to a real difference between the compared samples.
KeywordsDiagnostic Performance Carpal Tunnel Syndrome Conventional Coronary Angiography Verification Bias Lead Time Bias
Unable to display preview. Download preview PDF.