T Distribution Study Pack

Kibin's free study pack on T Distribution includes a 6-section study guide, 25 quiz questions, 30 flashcards, and 5 open-ended Explain review questions. Sign up free to track your progress toward mastery, plus upload your own notes and recordings to create personalized study packs organized by course.

Last updated May 28, 2026

Topic mastery0%

T Distribution Study Guide

Master the t distribution — from degrees of freedom and heavier tails to t-tests and confidence intervals for unknown population standard deviations. See how the t and z distributions converge as sample size grows.

Key Takeaways

  • The t distribution is a symmetric, bell-shaped probability distribution used when estimating population parameters from small samples or when the population standard deviation is unknown.
  • Unlike the standard normal (z) distribution, the t distribution has heavier tails, which accounts for the additional uncertainty introduced by estimating the population standard deviation with the sample standard deviation.
  • Each t distribution is defined by a single parameter called degrees of freedom (df), calculated as n − 1 for a one-sample problem, where n is the sample size.
  • As degrees of freedom increase, the t distribution converges toward the standard normal distribution; at roughly df ≥ 30, the two are practically indistinguishable.
  • The t distribution is the basis for constructing confidence intervals and conducting hypothesis tests (t-tests) for population means when σ is unknown.
  • Critical values from the t distribution are always larger in magnitude than their z counterparts at the same significance level, producing wider confidence intervals that reflect greater uncertainty in small samples.

Why the t Distribution Exists: The Problem with Unknown Population Variance

When a researcher wants to make inferences about a population mean, the ideal tool is the standard normal distribution — but only if the population standard deviation (σ) is known. In practice, σ is almost never known, and using the sample standard deviation (s) as a substitute introduces extra variability that the normal distribution does not account for.

Limitation of the z-Statistic

  • The z-statistic (x̄ − μ) / (σ / √n) assumes σ is a known constant, so every source of randomness comes only from the sample mean x̄.
  • When s replaces σ, the ratio becomes random in both the numerator and denominator, making the resulting statistic follow a different, heavier-tailed distribution.
  • Using z critical values in this situation underestimates the true variability and produces confidence intervals that are too narrow, increasing the risk of incorrect conclusions.

Origin of the t Distribution

  • William Sealy Gosset derived the t distribution in 1908 while working at the Guinness Brewery, publishing under the pseudonym 'Student' — which is why the distribution is often called Student's t distribution.
  • Gosset needed a mathematically rigorous way to draw conclusions from very small batch samples, where assuming known σ was unrealistic.
  • The t distribution formally describes the sampling distribution of the statistic (x̄ − μ) / (s / √n) when the underlying population is approximately normal.

Shape and Properties of the t Distribution

The t distribution shares several features with the standard normal curve but differs in ways that matter significantly for statistical inference, especially at small sample sizes.

Shared Features with the Standard Normal Distribution

  • Both distributions are symmetric around a mean of zero.
  • Both are unimodal and bell-shaped, with total area under the curve equal to 1.
  • Both extend infinitely in both directions, though probabilities in the extreme tails become vanishingly small.

Heavier Tails and What They Represent

  • The t distribution has more area in its tails than the standard normal distribution — a property described as having heavier or fatter tails.
  • This extra tail area reflects the real possibility of obtaining an unusually large or small sample standard deviation by chance, which inflates the t-statistic beyond what a z-score would reach.
  • As a direct consequence, the critical values (t*) needed to capture a given percentage of the distribution are larger in absolute value than the corresponding z critical values.

Degrees of Freedom and Shape Change

  • The t distribution is not a single curve but a family of curves, each uniquely defined by its degrees of freedom (df).
  • At df = 1 (extremely small sample), the tails are very heavy and the peak is low and flat.
  • As df increases, the tails shrink and the peak rises, so the curve grows progressively closer to the standard normal shape.
  • At approximately df = 30 or more, the practical difference between t and z critical values is negligible for most applications.

Degrees of Freedom: The Governing Parameter

Degrees of freedom determine which specific t distribution to use, and understanding what they represent conceptually clarifies why they equal n − 1 rather than n.

Conceptual Meaning of Degrees of Freedom

  • Degrees of freedom measure the number of independent pieces of information available to estimate a parameter.
  • When computing the sample standard deviation s from a sample of size n, the sample mean x̄ must be calculated first and is then treated as fixed.
  • Once x̄ is fixed, only n − 1 of the individual deviations (xi − x̄) can vary freely; the last deviation is determined by the constraint that all deviations must sum to zero.

Calculating Degrees of Freedom in Common Scenarios

  • One-sample t-test or single confidence interval: df = n − 1.
  • Two-sample t-test for independent groups: df depends on whether equal variances are assumed; the conservative formula uses df = min(n₁ − 1, n₂ − 1), while the Welch approximation produces a non-integer df calculated from both sample sizes and variances.
  • Paired t-test: treat the differences as a single sample, so df = n_pairs − 1.

Effect of Degrees of Freedom on Critical Values

  • A t distribution with df = 5 has a two-tailed critical value of approximately 2.571 at the 95% confidence level, compared to z* = 1.960 for the standard normal.
  • At df = 30, the t critical value drops to approximately 2.042, much closer to 1.960.
  • Consulting a t-table or software output always requires specifying df to identify the correct critical value.

Unlock the rest of this study guide

  • Access the full study pack
  • Track your mastery and be test-day ready
  • Upload your own notes to build personalized study guides, quizzes, flashcards, and more
Sign up free →

About this Study Pack

Created by Kibin to help students review key concepts, prepare for exams, and study more effectively. This Study Pack was checked for accuracy and curriculum alignment using authoritative educational sources. See sources below.

Sources

More in Statistics

See all topics →

Browse other courses

See all courses →