Current Incentives for Scientists Lead to Underpowered Studies with Erroneous Conclusions.

Area of Science:

• Scientific research strategy
• Research ecosystem dynamics
• Publication and career success

Background:

• Scientific incentive structures often prioritize novel findings.
• Scientists aim to maximize their fitness, measured by publication records and career success.
• Existing systems may inadvertently promote suboptimal research strategies.

Purpose of the Study:

• To develop an optimality model for rational research strategies.
• To predict the optimal balance between seeking novel versus confirmatory research.
• To determine the ideal research effort allocation per study.

Main Methods:

• Developed an optimality model based on scientific incentive structures.
• Modeled research strategy in terms of effort allocation to novel vs. confirmatory studies.
• Used parameter values derived from the scientific literature.

Main Results:

• Researchers maximizing fitness should prioritize novel results.
• Optimal strategies involve conducting small studies with low statistical power (10%-40%).
• Approximately 50% of published studies report erroneous conclusions under these conditions.

Conclusions:

• Current scientific incentive structures conflict with maximizing research's scientific value.
• The scientific ecosystem may need adjustments to improve overall research quality.
• Reforming incentive structures could lead to more reliable scientific findings.