The idea that science should be free from bias is pretty straightforward. After all, the goal is to find reliable answers. However, there is a big problem called publication bias. This is when only certain results get published. It is a major issue and has been linked to the replication crisis in science. But there is more to this story than meets the eye.

    Some scientists might be able to work around this bias. For example, a scientist who knows about publication bias can use a Bayesian model to interpret published results without getting a skewed view. In some very specific cases, a scientist might even choose not to bother with policies that aim to reduce publication bias. Why? Because these policies can be time-consuming and might not seem worth the effort for the individual scientist.

    However, when looking at science as a whole, publication bias is a big problem. The benefits of science often reach the public through intermediaries. These are people who are not experts but still need to use or share scientific findings. The issue is that these intermediaries might not know how to account for publication bias. This means they could end up spreading biased information.

    So, even if scientists do not see the value in policies like mandatory pre-registration, it is still important to implement them. These policies help ensure that the information reaching the public is reliable. The benefits of these policies are not just for scientists but for everyone who uses or relies on scientific findings. Therefore, when deciding how to run scientific institutions, it is crucial to consider the perspective of these intermediaries. Their needs should be prioritized to make sure science benefits society as a whole.

How Bias in Science Publishing Affects Us All

Actions