Original title: Teaching Thoughtful Statistical Inference: Stories for the World Beyond p < 0.05
Peter Martin · September 2026
Many statistics students do not understand hypothesis tests and p values. A p value is a number used to judge if data supports a claim. Across published science, researchers often apply these tools poorly and interpret them incorrectly. They also lean too heavily on a rule called "p < 0.05" to claim statistical significance, which hides real flaws in their evidence. To fix this, the author argues that teachers should embed inference within broader statistical thinking. Instructors should show how real research actually uses these methods, demonstrate good habits, and reveal what fails when people misinterpret, misuse, or cheat p values. Sharing stories from the history of statistics and science provides rich examples that show students the real uses and traps of hypothesis tests.
Imagine you are flipping a coin to see if it is fair. You flip ten times, get nine heads, and declare the coin unfair. But what if you only stopped after seeing nine heads? Your stopping rule makes your conclusion unreliable. In this paper, researchers sometimes choose when to stop or what to report based on how results look, which biases the evidence they present. Historical stories show how these hidden choices distort statistical conclusions.
Every year, countless research papers contain flawed conclusions because researchers misuse p-values, the most common tool for deciding if a result is real. This problem extends to medicine, psychology, economics, and other fields where decisions rely on statistical evidence. By using historical stories to teach how inference really works, educators can help students recognize good from bad practice, reducing the flood of unreliable findings.
The paper proposes a teaching approach but does not test it experimentally. It does not provide evidence that using historical narratives improves student understanding or reduces misuse. The abstract outlines a conceptual argument, not a curriculum or empirical study. Its scope is limited to identifying a problem in statistics education and suggesting a pedagogical direction without concrete implementation details.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Teaching Thoughtful Statistical Inference: Stories for the World Beyond p < 0.05 — Peter Martin — Teaching Statistics — doi:10.1002/test.70049
Abstract licence: CC BY
Explanation generated by Synthesis from the abstract — it is not the original paper.