Addressing selective reporting of experiments – the case for predefined exclusion criteria

Abstract

Articles describing experimental data in the life sciences are meant to tell a clear story to the reader. This means that not every experimental attempt ends up published, as failed experiments and uninformative data are typically filtered out by researchers. Freedom to exclude data from an article, however, can lead to reporting bias when exclusion decisions are made after results are in. We discuss how to balance clarity and thoroughness in biomedical research reporting, and suggest that predefined criteria for experimental validity might help in solving this conflict.