Bias in results of randomised trials: Direction, threshold, and context

The project consists of four studies: 1) a scoping review to describe the principles and terminology for assessing the direction of bias, minimally relevant degree of bias, and interaction between sources of bias in results of randomised clinical trials. 2) a cross-sectional study and expert sruvey to analyse the mechanisms of direction of bias. 3) a vignette study to assess the threshold for minimal relevant bias as perceived by clinical research methodologists and systematic reviews authors. 4) a comparative analysis to identify likely and possible sources of contextual variation in estimates of observer bias.