Skip to content
Statculator
P-Values & Critical Values

P-Value from t Score Calculator

Free p-value from t score calculator. Enter t and df to get the two-tailed, left- or right-tailed p-value, the decision at your α, a shaded t curve and steps.

Free, no sign-up Runs in your browserBy Hira Noor, BS MathematicsUpdated

Try

Negative values are fine.

Decimals are allowed, for example a Welch df of 17.63.

Tail

Left for H₁: “less than”, right for H₁: “greater than”, two-tailed for H₁: “not equal to”.

Leave blank to get the p-value without a decision.

Results

p-value (two-tailed)
0.0366
p = 2 × P(T ≥ |2.31|), df = 14
Critical t at α = 0.05
±2.1448

Significant at α = 0.05: reject H₀

p = 0.0366 ≤ 0.05. Equivalently, |t| = 2.31 is at or beyond the critical value 2.1448.

Left tail, P(T ≤ t)
0.9817
Right tail, P(T ≥ t)
0.0183
Two-tailed, 2 · P(T ≥ |t|)
0.0366
Degrees of freedom
14
Same value read as a z-score (two-tailed)
0.0209
Incomplete-beta argument, x = df/(df + t²)
0.724

Interpretation: If H₀ is true, a t statistic at least 2.31 away from 0 in either direction would turn up about once in every 27 repetitions with 14 degrees of freedom. Treating t as a z-score would give p = 0.0209 instead; the gap comes from the heavier tails of the t distribution and shrinks as df grows.

t distribution curve with the shaded probability region, markers at t = 2.31, -2.31-2.31t = 2.31-4-3-2-101234t distribution, df = 14
Shaded area = p-value = 0.0366 (both tails together). Dashed line: your t statistic.

Decision at common significance levels (two-tailed, df = 14)

αCritical t*Reject H₀?
0.1±1.7613Yes
0.05±2.1448Yes
0.025±2.5096No
0.01±2.9768No
0.005±3.3257No
0.001±4.1405No

p = 0.0366. “Yes” means p ≤ α, the same as the observed t falling beyond the critical value.

Step-by-step solution

  1. 1

    Write the p-value as a probability

    T follows a t distribution with 14 degrees of freedom when H₀ is true. A two-tailed p-value adds the areas beyond −|t| and +|t|, which are equal because the curve is symmetric.

    p = 2 × P(T ≥ |2.31|), T ~ t(14)
  2. 2

    Convert t to the incomplete-beta argument

    The t distribution’s tail area is a regularized incomplete beta function evaluated at x = df / (df + t²).

    x = df / (df + t²) = 14 / (14 + 2.31²) = 0.724034
  3. 3

    Find the area beyond |t|

    P(T ≥ 2.31) = ½ × Iₓ(df/2, ½) = ½ × I(0.724034; 7, 0.5) = 0.018324
  4. 4

    Map the tail area to the p-value

    Double the one-sided area to count both tails.

    p = 2 × 0.018324 = 0.0366
  5. 5

    Decision: reject H₀

    p = 0.0366 ≤ α = 0.05
    Critical value t* = ±2.1448 (df = 14)

What is a P-Value from t Calculator?

A p-value from t calculator turns a t statistic and its degrees of freedom into a p-value: the probability of a t at least as extreme as yours if the null hypothesis is true. For a two-tailed test it is the area beyond ±|t| under the Student t curve, so t = 2.31 with df = 14 gives p = 0.0366, significant at α = 0.05. Fewer degrees of freedom mean heavier tails, so the same t gives a larger p-value.

What the P-Value from t Calculator does

Enter a t statistic (also called a t score or t value) and its degrees of freedom to get the exact left-tail, right-tail and two-tailed p-values, the critical t at your α, the reject or fail-to-reject decision, and a t curve with the p-value area shaded. A table shows the decision at six common α levels, and the steps show the incomplete-beta calculation with your numbers. It works as a p-value from t score calculator, a t stat to p-value converter or a t score p-value calculator: the input is always t and df. Degrees of freedom can be any positive number, including the fractional values that Welch’s t-test produces.

This is the last step of every t procedure: one-sample, two-sample, Welch and paired t-tests, tests of regression slopes and tests of a correlation coefficient. Use it when software or a textbook gives you t but not p, or when a printed t-table only brackets the answer between two columns.

How to use the P-Value from t Calculator

  1. Enter the t statistic exactly as computed, including its sign.
  2. Enter the Degrees of freedom (df) for your test, such as n − 1 for a one-sample or paired test or n₁ + n₂ − 2 for a pooled two-sample test. The Degrees of Freedom Calculator finds df for other tests.
  3. Pick the Tail that matches the alternative hypothesis you stated before collecting data: two-tailed for “not equal to”, right for “greater than”, left for “less than”.
  4. Optionally set the Significance level (α), then read the p-value, the decision and the shaded t curve. The step-by-step solution shows the incomplete-beta calculation.

How to determine the p-value from a t-test

Finding the p-value from a t score takes three things: the t statistic, its degrees of freedom and the direction of the alternative hypothesis. The p-value is the area under the t(df) curve beyond your t, in the direction H₁ points.

  1. Compute t. For a one-sample test, t = (x̄ − μ₀) / (s / √n). For other tests, t = estimate / standard error.
  2. Work out df. One-sample or paired: n − 1. Pooled two-sample: n₁ + n₂ − 2. Welch: the Welch–Satterthwaite value, usually a decimal. Regression slope: n − k − 1.
  3. Pick the tail. Right-tailed: p = P(T ≥ t). Left-tailed: p = P(T ≤ t). Two-tailed: p = 2 × P(T ≥ |t|).
  4. Find the area. This calculator, Excel or R gives it exactly. With a printed t-table, find the row for your df and the two critical values your |t| falls between; the p-value lies between their α levels.
  5. Compare with α. If p ≤ α, reject H₀. If p > α, fail to reject H₀.

How to find the p-value for a two-tailed test

A two-tailed p-value counts results as extreme as yours in either direction. Because the t curve is symmetric about 0, the area beyond −|t| equals the area beyond +|t|, so you find one tail and double it: p = 2 × P(T ≥ |t|). The sign of t does not matter; t = −2.31 and t = 2.31 with df = 14 both give p = 0.0366.

A one-tailed p-value is half the two-tailed value when t points in the direction H₁ predicts (0.0183 here). When t points the other way, the one-tailed p-value is 1 minus that half (0.9817), and the result can never be significant. Choose the tail before looking at the data; switching to one tail after seeing t inflates the false-positive rate.

P-value from t formula

For a t statistic with ν degrees of freedom, the area beyond |t| on one side has a closed form in terms of the regularized incomplete beta function Iₓ(a, b).

One-sided area beyond |t|
P(T ≥ |t|) = ½ × Iₓ(ν/2, ½), x = ν / (ν + t²)
Two-tailed p-value
p = 2 × P(T ≥ |t|) = Iₓ(ν/2, ½)
Right-tailed and left-tailed p-values
p_right = P(T ≥ t), p_left = P(T ≤ t) = 1 − p_right
t density
f(t) = Γ((ν + 1)/2) / [√(νπ) Γ(ν/2)] × (1 + t²/ν)^(−(ν + 1)/2)
Formula symbols
SymbolMeaning
tobserved test statistic
ν (df)degrees of freedom, any positive real number
Ta random variable with a Student t distribution on ν df
Iₓ(a, b)regularized incomplete beta function

As ν grows, the t distribution approaches the standard normal, so for large df the p-value is nearly the same as the p-value from z.

The distribution follows the NIST/SEMATECH e-Handbook definition of the t distribution. Critical values here agree with its table of t critical values, and the p-values match R’s pt and SciPy’s t.sf.

Worked example

Worked example: paired test scores

Problem. Fifteen students take a test before and after a revision workshop. A paired t-test on the improvements gives t = 2.31 with df = 15 − 1 = 14. Find the two-tailed p-value and decide at α = 0.05.

  1. 01Two-tailed probability statement: p = 2 × P(T ≥ 2.31) with T ~ t(14).
  2. 02Incomplete-beta argument: x = 14 / (14 + 2.31²) = 14 / 19.3361 = 0.724034.
  3. 03One-sided area: P(T ≥ 2.31) = ½ × I(0.724034; 7, 0.5) = 0.018324.
  4. 04Two-tailed p-value: p = 2 × 0.018324 = 0.0366.
  5. 05The critical value for α = 0.05 and df = 14 is ±2.1448, and 2.31 > 2.1448.

Answer. p ≈ 0.0366 < 0.05, so reject H₀: the mean improvement is statistically significant. Had the same t been read from the normal curve, p would be 0.0209, which shows how much the small df matters.

How to calculate the p-value from t by hand

With only a printed t-table you can bracket the p-value. Take the worked example above, t = 2.31 with df = 14. The df = 14 row of a two-tailed table lists 2.145 under α = 0.05 and 2.624 under α = 0.02. Since 2.145 < 2.31 < 2.624, the two-tailed p-value lies between 0.02 and 0.05. The exact answer, 0.0366, is inside that range.

For df = 1 and df = 2 the t distribution has a simple closed form, so you can get the exact p-value on any scientific calculator:

  • df = 1 (the Cauchy distribution): two-tailed p = 1 − (2/π) × arctan(|t|). For t = 1, p = 1 − (2/π)(π/4) = 0.5.
  • df = 2: two-tailed p = 1 − |t| / √(t² + 2). For t = 2, p = 1 − 2 / √6 = 1 − 0.8165 = 0.1835.
  • Larger df: use the incomplete-beta formula above, a calculator or software. For df above about 100, the normal curve is a close approximation (see the p-value from z).

t statistic to p-value table

Two-tailed p-values for common t statistics and degrees of freedom. Read across from your df to the t nearest yours; for an exact value, enter it in the calculator above. Down each column the p-value falls toward the normal-curve value in the last row, which is why t = 2 is “significant at 5%” for large samples but not for small ones. Halve any entry for a one-tailed p-value in the predicted direction.

Two-tailed p-value, 2 × P(T ≥ |t|), by df
dft = 1.0t = 1.5t = 2.0t = 2.5t = 3.0t = 4.0
10.50000.37430.29520.24220.20480.1560
20.42260.27240.18350.12960.09550.0572
30.39100.23060.13930.08770.05770.0280
50.36320.19390.10190.05450.03010.0103
100.34090.16450.07340.03140.01330.0025
150.33320.15440.06390.02450.00900.0012
200.32930.14920.05930.02120.00710.0007
300.32530.14410.05460.01810.00540.0004
600.32130.13890.05000.01520.00390.0002
1200.31930.13620.04780.01380.00330.0001
∞ (normal)0.31730.13360.04550.01240.00270.0001

P-value from t in Excel, R, Python and on a TI-84

Every package computes the same area. The examples use t = 2.31 and df = 14, which give a two-tailed p-value of 0.0366 and a right-tailed p-value of 0.0183. Excel’s T.DIST.2T and T.DIST.RT need a non-negative t, so wrap a negative statistic in ABS for the two-tailed value.

ToolTwo-tailed p-valueRight-tailed p-value
Excel / Google Sheets=T.DIST.2T(ABS(2.31), 14)=T.DIST.RT(2.31, 14)
R2 * pt(-abs(2.31), 14)pt(2.31, 14, lower.tail = FALSE)
Python (SciPy)2 * t.sf(abs(2.31), 14)t.sf(2.31, 14)
TI-83 / TI-842 * tcdf(2.31, 1E99, 14)tcdf(2.31, 1E99, 14)

Reading and reporting the p-value

The p-value is the probability of a t statistic at least as extreme as the one observed, assuming H₀ is true. It is not the probability that H₀ is true, and a p-value above α does not prove H₀; it means the data are compatible with it.

In APA style, give t with its df in brackets, then the exact p-value: t(14) = 2.31, p = .037. Report p to two or three decimals and write p < .001 for anything smaller. To check a published result such as “t(28) = 2.45”, enter t = 2.45 and df = 28 here: the two-tailed p-value is 0.0208, so a printed “p = .021” is consistent and a printed “p = .002” is not.

Common use cases

  • Converting the t from a hand-worked one-sample, two-sample or paired t-test into a p-value.
  • Finishing a Welch’s t-test, whose fractional df a printed table cannot handle.
  • Testing a regression slope or coefficient, where t = estimate / standard error with n − k − 1 df.
  • Checking a t and df reported in a journal article, such as “t(28) = 2.45”, against the p-value printed next to it.

Assumptions and limitations

  • The p-value assumes the statistic follows a t distribution under H₀, which in turn requires roughly normal data (or differences) and independent observations. For small, skewed samples consider a nonparametric alternative such as the Wilcoxon signed-rank or Mann–Whitney test.
  • The df must match the test. Using n instead of n − 1, or pooled df for a Welch statistic, gives the wrong p-value.
  • One-tailed p-values are valid only for a direction chosen in advance; after the fact, report the two-tailed value.
  • A small p-value shows a difference is unlikely to be zero, not that it is large. Report a confidence interval and an effect size such as Cohen’s d alongside it.

Frequently asked questions

How do you convert a t-score to a p-value?

You need the t value, its degrees of freedom and the tail of the test. The p-value is the area under the t distribution with those df beyond your t: to the right for a right-tailed test, to the left for a left-tailed test, or twice the area beyond |t| for a two-tailed test. For t = 2.31 and df = 14 the two-tailed p-value is 0.0366.

How do you calculate the p-value in a t-test?

Compute the t statistic from your data, work out its degrees of freedom (n − 1 for a one-sample or paired test, n₁ + n₂ − 2 for a pooled two-sample test), then find the tail area beyond t on the t(df) distribution. Excel’s =T.DIST.2T(ABS(t), df) and R’s 2 * pt(-abs(t), df) give the two-tailed value.

Can degrees of freedom be a decimal?

Yes. Welch’s t-test and some mixed-model methods produce fractional df such as 17.63. The t distribution is defined for any positive df, and this calculator uses the value exactly instead of rounding it down as a printed table would.

Why is my p-value from t larger than the p-value from z?

The t distribution has heavier tails than the normal distribution because it accounts for estimating the standard deviation from the sample. More area lies beyond any given value, so the same statistic gives a larger p-value under t, especially when df is small.

What t value is significant at 0.05?

It depends on df. For a two-tailed test at α = 0.05 the critical value is 12.706 for df = 1, 2.228 for df = 10, 2.042 for df = 30 and approaches 1.960 as df grows. The T Critical Value Calculator gives the exact value for any df.

Is a t-test p-value two-tailed by default?

Most software, including R’s t.test and Excel’s T.TEST with tails = 2, reports two-tailed p-values unless told otherwise. A one-tailed p-value is half the two-tailed value when t points in the hypothesized direction.

Can a p-value from t be negative or greater than 1?

No. A p-value is a probability, so it always lies between 0 and 1. A negative t simply means the sample estimate is below the null value; its two-tailed p-value is the same as for the positive t. A one-tailed p-value above 0.5 means t points away from the direction H₁ predicts.

Cite this page

Statculator. (2026, November 14). P-Value from t Calculator. Retrieved November 14, 2026, from https://statculator.com/p-values-critical-values/p-value-from-t-calculator/