Valentin Amrhein, Sander Greenland, Blake McShane, and over 800 signatories advocate for the abandonment of statistical significance in scientific research. They argue that the dichotomous approach to statistical significance, where results are labeled as either "statistically significant" or "not statistically significant," leads to misinterpretations and overconfident claims. The authors emphasize that a statistically non-significant result does not prove the null hypothesis and that statistically significant results do not prove any alternative hypothesis. They propose renaming confidence intervals as "compatibility intervals" to avoid overconfidence and encourage a more nuanced interpretation of results. The article highlights the widespread misuse of statistical significance, including the dismissal of potentially crucial effects and the overemphasis on significance thresholds. The authors call for a shift towards more detailed and scientific inferences, incorporating background evidence, study design, data quality, and underlying mechanisms. They suggest that decisions should be based on a comprehensive assessment of all potential consequences rather than solely on statistical significance. The article concludes by emphasizing the need for ongoing monitoring of statistical abuses in the scientific literature to ensure the continued improvement of research practices.Valentin Amrhein, Sander Greenland, Blake McShane, and over 800 signatories advocate for the abandonment of statistical significance in scientific research. They argue that the dichotomous approach to statistical significance, where results are labeled as either "statistically significant" or "not statistically significant," leads to misinterpretations and overconfident claims. The authors emphasize that a statistically non-significant result does not prove the null hypothesis and that statistically significant results do not prove any alternative hypothesis. They propose renaming confidence intervals as "compatibility intervals" to avoid overconfidence and encourage a more nuanced interpretation of results. The article highlights the widespread misuse of statistical significance, including the dismissal of potentially crucial effects and the overemphasis on significance thresholds. The authors call for a shift towards more detailed and scientific inferences, incorporating background evidence, study design, data quality, and underlying mechanisms. They suggest that decisions should be based on a comprehensive assessment of all potential consequences rather than solely on statistical significance. The article concludes by emphasizing the need for ongoing monitoring of statistical abuses in the scientific literature to ensure the continued improvement of research practices.