| Popis: |
Randomized-Controlled Trials are the gold standard of experimental sciences. However, randomized trials are not innately perfect—they require careful stewardship to ensure that they produce valid and meaningful conclusions. Therefore, it is essential for scientists and scientific editors alike to be well versed in threats to experimental hygiene , which can cause randomized experiment to nonetheless produce biased or incorrect conclusions. These threats can arise at a variety of stages, from experiment conception and design to data analysis. While this chapter is not exhaustive, it highlights six major and common threats that journal editors may face: unbalanced groups, duration bias, imperfect compliance, attrition, spillover effects, and behavioral responses. This information can help editors protect their journals and readers from errors, scientific misconduct, and fraud. |
