Publish without bias or perish without replications

Area of Science:

• Scientific methodology
• Research integrity
• Publication bias

Background:

• A significant portion of experimental results are unreproducible, fueling concerns about a replicability crisis in science.
• Publication bias, particularly against negative results, is a key factor influencing the scientific record.
• The impact of publication bias on scientific progress is debated, with some arguing it's detrimental and others suggesting it's not necessarily harmful.

Purpose of the Study:

• To propose a novel mechanism by which reducing publication bias can benefit scientific progress.
• To explore the relationship between publication bias and the quality of scientific research through mathematical modeling.

Main Methods:

• Development of a series of mathematical models.
• Simulation of a scientific community comprising discovery researchers (testing novel hypotheses) and confirmation researchers (testing known hypotheses).

Main Results:

• Reducing publication bias can unexpectedly increase the proportion of confirmation researchers conducting replications.
• An increased focus on confirmation research incentivizes scientists to produce high-quality work due to increased scrutiny.

Conclusions:

• Reducing publication bias offers an underappreciated pathway to enhancing scientific quality and progress.
• The study suggests that fostering an environment with less publication bias can lead to more rigorous scientific inquiry.