A high p-value means your data did not provide strong evidence against the default assumption (called the null hypothesis) that nothing interesting is going on. In most fields, “high” means any p-value above 0.05, though the exact cutoff depends on the study. A p-value of 0.72, for instance, suggests the results you observed would be perfectly common in a world where there’s no real effect, so you have little reason to claim one exists.
What a P-Value Actually Measures
A p-value answers a narrow question: if there were truly no effect, how likely would you be to see data at least as extreme as what you collected? It’s a measure of compatibility between your data and the assumption of “no difference” or “no relationship.” A small p-value (say, 0.003) means the data would be very unusual under that assumption, which gives you reason to doubt it. A high p-value means the data fit comfortably with the no-effect scenario.
Most research uses 0.05 as the dividing line. That threshold traces back to early 20th-century statistics and reflects a willingness to accept a 5% chance of being wrong when declaring an effect real. Some fields use stricter cutoffs like 0.01 (a 1% risk) for higher-stakes decisions. Anything above whichever threshold you’ve chosen counts as “high” in that context.
Why a High P-Value Is Not Proof of Nothing
This is the single most important thing to understand: a high p-value does not prove the null hypothesis is true. It does not mean there is no effect. It means you didn’t find convincing evidence of one. The distinction matters. “We found no evidence of fire” is not the same as “there is no fire,” especially if you only glanced out the window for two seconds.
The American Statistical Association released a formal statement emphasizing this point. One of its six core principles states that a large p-value is not evidence supporting your alternative hypothesis, because many different explanations could be consistent with the observed data. A high p-value could mean the treatment genuinely does nothing, or it could mean the study simply wasn’t powerful enough to detect a real effect.
Small Samples Can Hide Real Effects
One of the most common reasons for a high p-value is an undersized study. Statistical power is the probability that a study will detect a real effect if one exists. When a study has too few participants, power drops, and the p-value can come back high even when something meaningful is happening.
Consider a concrete example from a medical education study. Researchers compared two groups and found mean scores of 30.1 and 28.5, yielding a p-value of 0.06. On paper, that’s “not significant.” But the calculated effect size was 0.5, which is considered medium. The researchers determined that expanding from about 30 participants per group to 60 per group would raise statistical power to 80%, likely producing a significant result. The effect was probably real the whole time; the study was just too small to confirm it.
This is why interpreting a high p-value requires context. Before concluding that nothing is going on, you need to ask whether the study had enough participants to detect a plausible effect in the first place.
Type II Errors: The False Negative Problem
When a study produces a high p-value and you conclude there’s no effect, but one actually exists, that’s called a Type II error (or a false negative). Think of it like a court acquitting someone who actually committed the crime. The evidence wasn’t strong enough to convict, but that doesn’t mean the person was innocent.
The probability of making a Type II error is labeled beta. Statistical power is simply 1 minus beta. A study with 80% power still has a 20% chance of missing a real effect and returning a high p-value. Results above the significance threshold, as one research team put it, “do not imply that there is no association in the population; they only mean that the association observed in the sample is small compared with what could have occurred by chance alone.”
Confidence Intervals Tell You More
A high p-value has a direct relationship with confidence intervals. When the p-value is above 0.05, the corresponding 95% confidence interval for the effect will contain the null value (typically zero for a difference, or 1.0 for a ratio). That null value sitting inside the interval is another way of saying: zero effect is a plausible explanation for these data.
But confidence intervals give you something a p-value alone cannot: a range of plausible effect sizes. A 95% confidence interval of -2 to +15 contains zero, so the p-value will be above 0.05. Yet that same interval also contains some fairly large positive effects. Seeing the full range helps you judge whether the study was simply inconclusive versus truly showing no meaningful effect. If the interval is narrow and tightly clustered around zero, there’s genuinely not much happening. If it’s wide, you just don’t have enough information yet.
Common Misconceptions
A p-value of 0.20 does not mean there’s a 20% chance the null hypothesis is correct. The p-value assumes the null hypothesis is already true and then asks how surprising the data are. It cannot loop back and tell you the probability of the hypothesis itself being true. This is a subtle but critical distinction that even experienced researchers sometimes get wrong.
A high p-value also does not measure the importance of a result. Practical significance and statistical significance are separate concepts. A clinical trial with 10,000 participants might find that a drug causes 0.5 kg of weight loss with a tiny p-value. That result is statistically significant but probably irrelevant to patients. The reverse happens too: a small study might return a high p-value for a treatment that actually produces meaningful improvement, simply because the sample was too small. The p-value tells you about the statistical evidence in your dataset, not about whether the effect matters in real life.
How to Read a High P-Value in Practice
When you encounter a high p-value in a study, paper, or report, run through a short mental checklist. First, check the sample size. A study with 15 participants per group has very different implications from one with 500. Second, look at the effect size or the confidence interval if they’re reported. A high p-value paired with a large but imprecise effect estimate means the study was underpowered, not that the effect is absent. Third, consider the context. One study returning p = 0.12 doesn’t settle a question, especially if other studies on the same topic have found effects.
Scientific journals are increasingly moving away from treating the 0.05 line as a hard boundary. The International Journal of Exercise Science, for example, now requires authors to report exact p-values (like p = 0.072) rather than blanket statements of “not significant.” This reflects a broader shift: a p-value is a continuous measure of evidence, not a binary verdict. A result at p = 0.06 is not fundamentally different from one at p = 0.04, even though they fall on opposite sides of the traditional cutoff.
The bottom line is that a high p-value means the data are consistent with no effect, but that’s not the same as proving no effect exists. It’s a statement about the strength of your evidence, not about the truth of your hypothesis.