أزمة المصداقية في العلم: "المعدِّلون"، المحتالون، والتلاعب بالنتائج التجريبية

A new perspective on scientific fraud - How unannounced “edits” to research designs and model specifications are fueling a crisis of credibility in science. In “The Credibility Crisis in Science,” prominent sociologists Thomas Plumber and Erik Neumayer argue that the most influential types of fraud are underappreciated. While falsifying and manipulating data is widely recognized as fraud, “editing”—the deliberate selection of research designs and model specifications based on the results they provide—is not. As a result, the credibility crisis in science is more severe than scientists and the public alike believe. The book focuses on the idea that the real danger to science comes not just from "outright fraud" (falsifying data outright), but from gray practices that the authors call "tweaking" 1. The difference between fraud and tweaking: While data falsification is easy to condemn, "tweaking" is the deliberate selection of particular research designs or statistical models just because they give attractive results or support the researcher's hypothesis, which is very difficult to detect. 2. Hidden Crisis: The authors argue that the credibility crisis is deeper than the public believes, because these statistical manipulations have become a common practice to increase the chances of publication in prestigious journals. 3. Failure of traditional strategies: The book argues that current methods of detecting fraud will not work with “modulators,” because they use real data but target it to serve specific results. 4. Suggested solutions: The authors propose a classification system for data based on their susceptibility to manipulation, and suggest that journal editors and reviewers (not authors) select “robustness tests” to ensure the validity of the results.