P-value
Statistical hypothesis testing, Null hypothesis, Statistical power
The focus on p-value produced a “significance filter” where the studies that happened to observe a significant p-value (or those that actively seek one) got published. This can be nicely exposed by examining all published p-values or z-scores (see Zwet2020significance).
A simple alternative to focusing on p-value is paying attention to the effect size and confidence interval. Sullivan2012using
- https://en.wikipedia.org/wiki/P-value
- P-values: misunderstood and misused
- The fickle P value generates irreproducible results
A New Look at P Values for Randomized Clinical Trials