comment this author: Even before reading the article, I knew that 0.05 is an arbitrary p-value chosen at some point for convenience. I've seen studies where lower p-values are used to reject the null hypothesis, as well as higher values being used. That said, I think there is a good point for evaluating the necessary p-value for each study. Intuitively, the available sample size plays a role in such determination. On statistical significance and practical importance, I thought of two examples. One is about the drug trial. Let's say the new drug significantly reduces the duration of one side effect by 50%, yet it does nothing to a range of other side effects. Let's say patients experience headaches, nausea, and fatigue. While the duration of nausea is decreased, other side effects remain. The result is statistically important, but maybe not materially significant. Another example is that a new arthritis drug reduces joint pain but causes dizziness, impacting mobility. Again - it is statistically significant but practically unimportant. Those might not be clean-cut examples as they involve other factors. Let's say a new genetic modification increases the weight of carrots by 0.001%. It may not be statistically significant, but it is