P-value Calculator — from z, t, or Chi-Square
StatisticsFreeNo sign-up
A p-value calculator turns a test statistic into the probability of a result at least that extreme if the null hypothesis is true. For z = 2.25 in a two-tailed test, the area beyond 2.25 in each tail of the normal curve is 0.0122, so p = 2 × 0.0122 = 0.0244. That is below α = 0.05, so the result is statistically significant.
If H₀ were true and the test’s assumptions hold, a z statistic at least as far from 0 as 2.25 in either direction would turn up in about 2.4% of samples. That is rare enough, at your α, to count as evidence against H₀.
- It is not the probability that H₀ is true, or that the result happened “by chance”.
- It does not tell you how large or important the effect is. Report an effect size and a confidence interval too.
- p ≤ 0.05 versus p > 0.05 is a convention, not a law; p = 0.049 and p = 0.051 are almost the same evidence.
Based on the American Statistical Association’s 2016 statement on p-values (principles 2, 3 and 5).
Show steps
- Hypotheses. H₀: μ = μ₀ (the population value in the null hypothesis). H₁: μ ≠ μ₀ (two-tailed).
- Under H₀ the statistic follows the standard normal distribution. Its left-tail area is Φ(z), the standard normal CDF, computed here from the complementary error function: P(Z ≥ z) = ½ erfc(z/√2).
- Two-tailed: double the tail beyond |z| = 2.25. p = 2 × P(X ≥ 2.25) = 2 × 0.0122 = 0.0244.
- Compare with α = 0.05: p = 0.0244 ≤ 0.05, so reject H₀.
- Same decision with critical values: reject H₀ when z is at most −1.96 or at least 1.96. Here z = 2.25, so it is in the rejection region.
How to use the p-value calculator
- Pick the statistic your test produced: z-score, t-score, chi-square, F-ratio or a correlation r.
- Choose the tail from your alternative hypothesis: left-tailed for H₁: <, right-tailed for H₁: >, two-tailed for H₁: ≠.
- Enter the statistic and its degrees of freedom (none for z; df for t and χ²; df₁ and df₂ for F; the sample size n for r).
- Set the significance level α (0.05 unless your course or study says otherwise) and press Find p-value.
You get the p-value, both tail areas, the critical value(s) for your α, the decision, an APA-style line such as p = .024, a plain-language interpretation, the distribution curve with the p-value area shaded, and every step of the working.
What the p-value means
A p-value answers one narrow question: if the null hypothesis H₀ were true, and the test’s assumptions held, how often would random sampling give a statistic at least as extreme as yours? A small p-value means your data would be unusual under H₀, which counts as evidence against it.
The American Statistical Association’s 2016 statement on p-values lists six principles. Three are worth keeping in mind every time you read a result:
- A p-value is not the probability that H₀ is true, and not the probability that the data were produced by chance alone.
- A p-value does not measure the size of an effect or how important a result is.
- Conclusions should not rest only on whether p crosses 0.05. Report the effect size and a confidence interval as well.
A large p-value does not prove H₀. It only says the data are not strong enough evidence against it, often because the sample is small.
P-value formulas for z, t, chi-square and F
Every p-value is an area under the curve of the statistic’s distribution when H₀ is true. Write the statistic as x and the cumulative distribution function as F(x):
| Tail | p-value | Used for |
|---|---|---|
| Left-tailed | p = F(x) | H₁: parameter < claimed value |
| Right-tailed | p = 1 − F(x) | H₁: parameter > claimed value; all goodness-of-fit, independence, ANOVA and regression F-tests |
| Two-tailed, z or t | p = 2 × (1 − F(|x|)) | H₁: parameter ≠ claimed value |
| Two-tailed, χ² or F | p = 2 × min(F(x), 1 − F(x)) | tests about one variance or a ratio of two variances |
The calculator evaluates these areas with standard special functions rather than a printed table: the normal tail as ½ erfc(z/√2), the t and F tails with the regularized incomplete beta function, and the chi-square tail with the regularized incomplete gamma function. Tiny p-values are shown in scientific notation instead of being rounded to 0.
Which tail should you use?
Decide the tail from the research question before you see the data. Use two-tailed when any difference matters (“is the mean different from 50?”). Use one-tailed only when a difference in the other direction would mean the same as no difference (“does the new method reduce errors?”). Switching tails after seeing the data overstates the evidence.
Chi-square tests of goodness of fit and independence, and the F-tests from ANOVA and regression, are always right-tailed: only large values of the statistic count against H₀.
Worked examples
p-value from a z-score
A test gives z = 2.25 with H₁: μ ≠ μ₀. The normal CDF is Φ(2.25) = 0.9878, so the right tail is 0.0122 and the two-tailed p-value is 2 × 0.0122 = 0.0244. Since 0.0244 ≤ 0.05, reject H₀. Equivalently, 2.25 is beyond the critical values ±1.96.
p-value from a t-score
A one-sample t-test with n = 16 gives t = 2.4, so df = 15. The area above 2.4 is 0.0149, and the two-tailed p-value is 0.0298. The two-tailed critical value is 2.131, and 2.4 is beyond it, so reject H₀ at α = 0.05. If H₁ had been μ > μ₀ (decided in advance), p = 0.0149.
p-value from a chi-square statistic
A test of independence on a 3 × 3 table gives χ² = 11.2 with df = (3 − 1)(3 − 1) = 4. The right-tail area is p = 0.0244, and the critical value at 0.05 is 9.488. Reject H₀: the two variables appear to be associated.
p-value from an F-ratio
A one-way ANOVA compares 3 groups of 5 people (N = 15), so df₁ = 3 − 1 = 2 and df₂ = 15 − 3 = 12. With F = 4.5 the right-tail area is p = 0.0348, and the critical value is 3.885. Reject H₀: at least one group mean differs.
p-value for a correlation
With r = 0.45 from n = 30 pairs, t = 0.45 × √28 / √(1 − 0.2025) = 2.666 with df = 28, and the two-tailed p-value is 0.0126. The correlation is statistically significant; r² = 20.3% of the variation is shared.
Critical values and p-values: quick reference
A statistic beyond the critical value always has p < α. These standard normal (z) values are the ones most worth remembering:
| α | One-tailed z | Two-tailed z |
|---|---|---|
| 0.10 | 1.282 | ±1.645 |
| 0.05 | 1.645 | ±1.960 |
| 0.01 | 2.326 | ±2.576 |
| 0.001 | 3.090 | ±3.291 |
For t the critical values are larger when df is small and shrink toward the z values as df grows. The two-tailed 0.05 value is 12.706 for df = 1, 2.571 for df = 5, 2.228 for df = 10, 2.042 for df = 30 and 1.980 for df = 120.
How to read the result
- p ≤ α: reject H₀; the result is statistically significant at that level. Say “the data provide evidence that…”, not “the data prove…”.
- p > α: fail to reject H₀. Do not write “accept H₀” or “there is no effect”.
- Report it fully: give the statistic, df and exact p, for example t(15) = 2.40, p = .030, rather than only “p < 0.05”.
Common p-value mistakes
| Mistake | Fix |
|---|---|
| Using the one-tailed area for a two-tailed test | Double the tail area for z and t when H₁ says ≠. |
| Dropping the minus sign of z or t | The sign decides which tail is left or right in a one-tailed test. |
| Wrong degrees of freedom | t: n − 1 for one sample; χ²: k − 1 or (r − 1)(c − 1); F: k − 1 and N − k. |
| Rounding the statistic too early | Keep at least 3 decimals in z, t, χ² or F before finding p. |
| Reading p as the chance H₀ is true | p assumes H₀ is true; it cannot also measure how likely H₀ is. |
Checking homework with the calculator
Work the problem by hand first: write H₀ and H₁, compute the statistic, and estimate p from a table (for example, z = 2.25 lies between the 0.025 and 0.01 one-tailed columns). Then enter the statistic here; the exact p should fall inside your table’s range. A mismatch usually points to the wrong tail or the wrong df.
Sources
- Wasserstein, R. L., and Lazar, N. A. (2016). The ASA Statement on p-Values: Context, Process, and Purpose. The American Statistician, 70(2), 129–133 (checked November 2026).
P-value practice quiz: 10 questions
Test yourself with 10 multiple-choice questions. Choose an answer and press Submit to see if you are right and why. Your score appears at the end.
The area above 1.5 is 0.0668. Two-tailed doubles it: 2 × 0.0668 = 0.1336.
Left-tailed p = Φ(−2.0) = 0.0228. 0.0455 would be the two-tailed value.
0.03 ≤ 0.05, so the result is significant at the 0.05 level and H₀ is rejected.
A large p-value only means the evidence against H₀ is weak. It does not prove H₀ or give the chance that H₀ is true.
Only large χ² values count against independence, so the p-value is the right-tail area.
Each tail holds 0.025, and the z with 0.025 above it is 1.96. 1.645 is the one-tailed 0.05 value.
For a symmetric statistic, the one-tailed p-value is half the two-tailed value: 0.08 / 2 = 0.04. The tail must be chosen before seeing the data.
Principle 5: a p-value, or statistical significance, does not measure the size of an effect or the importance of a result.
Quiz complete
Frequently asked questions
What is the p-value in statistics?
The p-value is the probability, assuming the null hypothesis is true, of getting a test statistic at least as extreme as the one observed. Small values (usually 0.05 or less) are taken as evidence against the null hypothesis.
How do you find the p-value for a t-test?
Compute t, find df (n − 1 for one sample), then find the area beyond t under the t curve with that df. Double it for a two-tailed test. For t = 2.4 with df = 15 the two-tailed p-value is 0.0298.
How do you find the p-value on a TI-84 calculator?
Use the DISTR menu. For a two-tailed z, enter 2 × normalcdf(2.25, 1E99, 0, 1). For t use tcdf(lower, upper, df), for chi-square χ²cdf(lower, upper, df), and for F Fcdf(lower, upper, df1, df2). The STAT > TESTS menu also prints p directly.
How do you find the p-value without a calculator?
Use a printed z, t, χ² or F table. Find your df row, see which two columns your statistic lies between, and read the range of tail areas from the column headings. Tables only give a range, such as 0.01 < p < 0.025.
Is a p-value a percentage?
A p-value is a probability between 0 and 1, but you can say it as a percentage: p = 0.0244 means about 2.4% of samples would be at least this extreme if H₀ were true. It is not the percentage chance that H₀ is true.
Can a p-value be 0 or negative?
No. A p-value is an area, so it is always above 0 and at most 1. Software may print 0.000 when p is very small; report it as p < .001 instead.
Related calculators
- Z-score calculator: standardize a value before finding p
- Chi-square calculator: goodness of fit and independence tests
- Confidence interval calculator: the interval to report next to p
- Standard deviation calculator: the spread used in z and t
- Sample size calculator: plan a study with enough power
More statistics calculators
Related to This Article
More math articles
- 3rd Grade STAAR Math Worksheets: FREE & Printable
- PayPal Fee Calculator
- ATI TEAS 7 Math FREE Sample Practice Questions
- Double Consciousness and Intersectionality
- Free Smarter Balanced Grade 5 Math Worksheets — 49 Printable Single-Skill PDFs with Solutions
- 5 Tips on How to Score Higher on SAT Math
- Appointment and removal powers
- Free Grade 7 English Worksheets for Louisiana Students
- Top 10 Tips to Create a PSAT/NMSQT Math Study Plan
- Campaign work




What people say about "P-Value Calculator From z, t, Chi-Square or F (With Steps)"?
No one replied yet.