Abstract

Selection bias is well known to affect surveys and epidemiological studies. There have been numerous methods proposed to reduce its effects, so many that researchers may be unclear which method is most suitable for their study; the wide choice may even deter some researchers, for fear of choosing a sub-optimal approach. We propose a straightforward tool to inform researchers of the most promising methods available to reduce selection bias and to assist the search for an appropriate method given their study design and details. We demonstrate the tool using three exam- ples where selection bias may occur; the tool quickly eliminates inappropriate methods and guides the researcher towards those to consider implementing. If more studies con- sider selection bias and adopt methods to reduce it, valuable time and resources will be saved, and should lead to more focused research towards disease prevention or cure.