Skip to content
UPCOMING EVENTS:UX, Product & Market Research Afterwork23. Apr.@Packhaus WienDetailsInsights & Research Breakfast16. Mai@Packhaus WienDetailsVibecoding & Agentic Coding for App Development22. Mai@Packhaus WienDetails
UPCOMING EVENTS:UX, Product & Market Research Afterwork23. Apr.@Packhaus WienDetailsInsights & Research Breakfast16. Mai@Packhaus WienDetailsVibecoding & Agentic Coding for App Development22. Mai@Packhaus WienDetails

P-Value

The probability of observing your data (or something more extreme) if there were truly no effect. Widely used, widely misunderstood, and never sufficient on its own to make a decision.

Definition: The probability of observing your data (or something more extreme) if there were truly no effect. Widely used, widely misunderstood, and never sufficient on its own to make a decision.

A p-value tells you how likely your observed result would be if the null hypothesis were true—that is, if there were no real difference or effect. A small p-value means the data would be surprising under the assumption of no effect.

What P < 0.05 Actually Means

If p = 0.03, there is a 3% chance of seeing a result this extreme assuming the null hypothesis is true. It does not mean:

  • There is a 97% chance the effect is real
  • There is a 3% chance you are wrong
  • The effect is large or practically meaningful

The 0.05 threshold is a convention, not a natural law. Fisher originally proposed it as a rough guide, not a binary decision rule.

Why P-Values Are Not Enough

  • Large samples detect trivial effects: With enough participants, you can get p < 0.001 for a difference so small it has no practical impact
  • Small samples miss real effects: A genuine improvement might produce p = 0.12 simply because you did not have enough data
  • P-values say nothing about magnitude: They tell you whether an effect exists, not whether it matters

What to Do Instead

Report effect size alongside your p-value. A statistically significant result with a tiny effect size is not worth acting on. A non-significant result with a large effect size may warrant a larger study. Use confidence intervals to communicate the range of plausible values—far more informative than a single yes-or-no threshold.

P-Value - Definition | UX Research Glossary | Busch Labs