Everybody Ps

“Significant or not significant?” You’ve had classes on this. You have had practice on this. You are motivated to perform this statistical practice of examining p values. P values are everywhere in your training and in the published literature.

If you have not heard this in your training, begin to listen up for how limited this use of the beloved p is. We can do so much better.

So then why? Why have you already been drilled in p values? Why do journals sometimes still only publish results with findings of p < 0.05? Ultimately, getting a computer to generate a p value and then examining whether it is above or below a threshold is truly easy. It makes results interpretation a breeze. There is no room for nuance, for examining patterns, for comparing to the wider world of published literature or real-world ranges of the constructs being studied, or for other messy considerations. (In practice, much messiness may have actually gone into forcing that p value just a tad lower, or in interpreting a p value as “marginal”). Perhaps more importantly, it appears so very objective. It’s math! It’s a rule! Everybody does it! This ease and perceived objectivity of p value interpretation has led to its widespread overuse. If your goal is to make the world a better place, and to do this by discerning truth, learn to interpret results without p values. P values are no portal to truth.