Biases in scientific studies are thought to be among the main factors that affect the reproducibility of results. A new study has assessed some of these biases and how they affect the results of studies.
This study, which was published in the Proceedings of the National Academy of Sciences covered all disciplines, was conducted by Daniele Fanelli of the Meta-Research Innovation Center at Standford University.
The researchers found that biases varied in different fields. Among other biases, the researchers found that authors of “small, early, and highly cited” studies are at risk of overrating their findings. On the other hand, authors of papers in non-peer reviewed journals were found to have a likelihood of underestimating their findings.
The researchers also found that there is a tendency for early studies and studies conducted in the US to focus on “extreme effects.” Some of the other factors that were found to influence the likelihood of inflating the results include inexperience, long-distance collaborations, and lack of individual integrity.