During my graduate studies in public health, I've been analyzing survey data, and I keep running into the problem of p-hacking—where researchers tweak analyses until they get a significant p-value. For instance, a colleague once tested twenty hypotheses and only reported the two that came out significant, which feels misleading. I argue that journals should ban the use of p-values entirely and require confidence intervals and effect sizes instead. What do you think—are p-values more harmful than helpful in statistical reporting?