Tests Based on Sampling Distributions - I: 12th Statistics MCQ
Chapter 2 of 12th Statistics covers small sample tests: Student's t-test for a single mean, for two independent means with pooled standard deviation and the paired t-test, and chi-square tests for a population variance, independence of attributes in contingency tables and goodness of fit of binomial and Poisson distributions.
Study this chapter
Asked in previous board exams
4 questions from this chapter came in 4 past public exams. Open any of them to see the answer and explanation.
-
March 2026
If n is the degree of freedom of chi-square distribution then its variance is :See answer in the solved paper →
-
March 2025
For testing the hypothesis concerning variance of a normal population __________ is used.See answer in the solved paper →
-
March 2023
The degrees of freedom of the chi-square test statistic for the contingency table of order 5 × 4 :See answer in the solved paper →
-
March 2020
If n is the degree of freedom of Chi-square distribution, then its variance is :See answer in the solved paper →
About Tests Based on Sampling Distributions - I
Medium ~70 min study
Tests Based on Sampling Distributions - I is the second chapter of the Tamil Nadu State Board Class 12 Statistics book. It opens with W.S. Gosset, the brewery chemist who published as "Student", and explains why the large sample procedures of Chapter 1 fail when the sample size is less than 30.
The first half develops Student's t-distribution, its properties and three uses: the test for a single normal mean when the variance is unknown, the test for equality of two means using the pooled standard deviation, and the paired t-test for correlated samples. The second half, introduced through Karl Pearson's 1900 paper, covers the chi-square distribution and its tests for a population variance, independence of attributes and goodness of fit.
This page brings together the book-back one mark questions with answers worked out step by step, and additional questions on test statistics, degrees of freedom, expected frequencies and rejection rules.
What you'll learn
- State the definition and properties of Student's t-distribution.
- Test a normal population mean from a small sample when the variance is unknown.
- Test the equality of two means with independent samples and with paired samples.
- Test whether a normal population has a specified variance using the chi-square test.
- Test the independence of attributes from an m x n contingency table.
- Test the goodness of fit of binomial, Poisson and ratio-based distributions.
Before you start
- Tests of significance, hypotheses, level of significance and critical region from Chapter 1.
- Mean, standard deviation and variance of a sample.
- Binomial and Poisson distributions from Class 11.
Topics covered in this chapter
Tests Based on Sampling Distributions - I explained
What this chapter covers
Student's t-distribution
If X is standard normal and Y is an independent chi-square variable with n degrees of freedom, \( X/\sqrt{Y/n} \) has the t-distribution with n degrees of freedom. It is symmetrical with mean zero, flatter in the middle than the normal curve, has more area in the tails and approaches N(0,1) as the degrees of freedom grow.
t-tests for small samples
- Single mean: \( t = \dfrac{\bar{x}-\mu_0}{s/\sqrt{n}} \) with n-1 degrees of freedom.
- Two independent means: \( t = \dfrac{\bar{x}-\bar{y}}{s_p\sqrt{1/m+1/n}} \) with m+n-2 degrees of freedom.
- Paired t-test: \( t = \dfrac{\bar{d}}{s/\sqrt{n}} \) with n-1 degrees of freedom.
Chi-square distribution
The square of a standard normal variable is chi-square with 1 degree of freedom. The distribution is continuous, has a single parameter n, mean n and variance 2n, and is additive.
Chi-square tests
- Variance: \( \chi^2 = (n-1)s^2/\sigma_0^2 \) with n-1 degrees of freedom.
- Independence of attributes: \( E_{ij} = R_i C_j / N \), degrees of freedom (m-1)(n-1).
- Goodness of fit: \( \sum (O_i-E_i)^2/E_i \) with k-1-s degrees of freedom, pooling classes whose expected frequency is below 5.
Common mistakes to avoid
- Taking n instead of n-1 as the degrees of freedom of the single mean t-test.
- Dividing by m+n instead of m+n-2 when computing the pooled variance.
- Using the two-sample t-test for before-and-after data, which needs the paired t-test.
- Counting cells instead of using (m-1)(n-1) for the degrees of freedom of a contingency table.
- Forgetting to pool classes with expected frequency below 5 and to subtract estimated parameters in k-1-s.
Test yourself on these with the practice test, then check the worked reasoning in the solved MCQs.
Frequently asked questions
Why is the t-distribution used for small samples?
When the sample size is less than 30, the sampling distributions of test statistics are far from normal, so the standard normal procedures of Chapter 1 cannot be applied. The t-distribution, found by W.S. Gosset, is used instead when the population is normal and its variance is unknown.
What are the properties of the t-distribution?
It is symmetrical with mean zero; the normal curve is higher in the middle, while the t curve has a greater spread and more area in its tails; it is asymptotic to the X-axis; its shape depends on the degrees of freedom and it approaches the standard normal distribution as they increase.
When is the paired t-test applied?
It is applied when the two samples are correlated in pairs, such as scores or sales of the same persons before and after training. The differences \( d_i \) are used and \( t = \bar{d}/(s/\sqrt{n}) \) follows the t-distribution with n-1 degrees of freedom.
What is the pooled standard deviation?
In the test for equality of two means with independent samples, the common unknown variance is estimated by \( s_p^2 = \dfrac{(m-1)s_x^2+(n-1)s_y^2}{m+n-2} \). Its square root, the pooled or combined standard deviation, appears in the denominator of the t statistic.
What are the conditions for the chi-square test of independence?
The total frequency N should be reasonably large, say greater than 50, and no expected cell frequency should be less than 5. Small cells are pooled with neighbouring cells. Expected frequencies are row total times column total divided by the sample size.
How are degrees of freedom found in a goodness of fit test?
They are k-1-s, where k is the number of classes after pooling those with expected frequency below 5 and s is the number of parameters estimated from the data. Fitting a binomial with p estimated or a Poisson with the mean estimated uses s = 1.
Last updated 28 October 2026