Every Book Back multiple-choice question from Tests of Significance - Basic Concepts and Large Sample Tests (12th Standard Statistics, Samacheer Kalvi) — each with the correct option highlighted and a clear, worked explanation. Free to read in English and Tamil.
Q1
Asked in board exam:
March 2023
Standard error of the sample mean is
- A. \( \sigma^2 \)
- B. \( \dfrac{\sigma}{n} \)
- C. \( \dfrac{\sigma}{\sqrt{n}} \)Correct
- D. \( \dfrac{\sqrt{n}}{\sigma} \)
Explanation. For iid observations with variance \( \sigma^2 \), \( V(\bar{X}) = \dfrac{n\sigma^2}{n^2} = \dfrac{\sigma^2}{n} \). The standard error is the standard deviation of this sampling distribution, \( SE(\bar{X}) = \dfrac{\sigma}{\sqrt{n}} \).
Q2
When \( n \) is large and \( \sigma^2 \) is unknown, \( \sigma^2 \) is replaced in the test statistic by
- A. Sample mean
- B. Sample varianceCorrect
- C. Sample standard deviation
- D. Sample proportion
Explanation. In the large sample test for the mean with unknown variance (Section 1.10), the population variance \( \sigma^2 \) is replaced by the sample variance \( S^2 \), giving \( Z = \dfrac{\bar{X} - \mu_0}{S/\sqrt{n}} \).
Q3
The critical value (table value) of the test statistic at the level of significance \( \alpha \) for a two-tailed large sample test is
- A. \( z_{\alpha/2} \)Correct
- B. \( z_{\alpha} \)
- C. \( -z_{\alpha} \)
- D. \( -z_{\alpha/2} \)
Explanation. For a two-sided alternative the area \( \alpha \) is split equally between the two tails, so the book's table gives the critical value \( z_e = z_{\alpha/2} \) and rejects \( H_0 \) when \( |z_0| \geq z_{\alpha/2} \).
Q4
In general, large sample theory is applicable when
- A. \( n \geq 100 \)
- B. \( n \geq 50 \)
- C. \( n \geq 40 \)
- D. \( n \geq 30 \)Correct
Explanation. Note 6 of the chapter: when the sample size is greater than or equal to 30 the sample is termed a large sample, and a sample of size less than 30 is a small sample.
Q5
Asked in board exam:
March 2026
When \( H_1 \) is a one-sided (right) alternative hypothesis, the critical region is determined by
- A. both right and left tails
- B. neither right nor left tail
- C. right tailCorrect
- D. left tail
Explanation. For \( H_1: \mu > \mu_0 \) the rejection rule is \( z_0 > z_{\alpha} \), so the area \( \alpha \) lies at the right end of the sampling distribution. Such a test is a right-tailed test.
Q6
Critical value at 5% level of significance for two-tailed large sample test is
- A. 1.645
- B. 2.33
- C. 2.58
- D. 1.96Correct
Explanation. Table 1.6: for a two-sided alternative at \( \alpha = 0.05 \), \( z_{\alpha/2} = z_{0.025} = 1.96 \). The value 1.645 is the one-tailed 5% value and 2.58 is the two-tailed 1% value.
Q7
For testing \( H_0: \mu = \mu_0 \) against \( H_1: \mu < \mu_0 \), what is the critical value at \( \alpha = 0.01 \)?
- A. 1.645
- B. \( -1.645 \)
- C. \( -2.33 \)Correct
- D. 2.33
Explanation. \( H_1: \mu < \mu_0 \) is a one-sided (left) alternative, so the critical value is \( -z_{\alpha} \). At 1%, Table 1.6 gives \( -z_{0.01} = -2.33 \).
Q8
Asked in board exam:
March 2026
When the alternative hypothesis is \( H_1: \mu \neq \mu_0 \), the critical region will be determined by
- A. both right and left tailsCorrect
- B. neither right nor left tail
- C. right tail
- D. left tail
Explanation. For a two-sided alternative \( H_0 \) is rejected when \( |z_0| \geq z_{\alpha/2} \), that is, for values in either tail, each tail holding area \( \alpha/2 \). It is a two-tailed test.
Q9
Asked in board exam:
March 2023
Rejecting \( H_0 \), when it is true is called
- A. type I errorCorrect
- B. type II error
- C. sampling error
- D. standard error
Explanation. From the decision table (Table 1.5), the error committed by rejecting \( H_0 \) when \( H_0 \) is really true is the type I error. Not rejecting a false \( H_0 \) is the type II error.
Q10
Asked in board exam:
March 2025
What is the standard error of the sample proportion under \( H_0 \)?
- A. \( \sqrt{\dfrac{PQ}{n}} \)Correct
- B. \( \sqrt{\dfrac{pq}{n}} \)
- C. \( \dfrac{PQ}{n} \)
- D. \( \dfrac{pq}{n} \)
Explanation. Table 1.4 gives \( SE(p) = \sqrt{\dfrac{PQ}{n}} \), where \( P \) is the population proportion and \( Q = 1 - P \). Under \( H_0: P = p_0 \) it becomes \( \sqrt{p_0q_0/n} \). Options (c) and (d) are variances, not standard errors.
Q11
The test statistic for testing the equality of two population means, when the population variances are known is
- A. \( \dfrac{\bar{X} - \bar{Y}}{\dfrac{\sigma_X^2}{m} + \dfrac{\sigma_Y^2}{n}} \)
- B. \( \dfrac{\bar{X} - \bar{Y}}{\dfrac{s_X^2}{m} + \dfrac{s_Y^2}{n}} \)
- C. \( \dfrac{\bar{X} - \bar{Y}}{\sqrt{\dfrac{\sigma_X^2}{m} + \dfrac{\sigma_Y^2}{n}}} \)Correct
- D. \( \dfrac{\bar{X} - \bar{Y}}{\sqrt{\dfrac{s_X^2}{m} + \dfrac{s_Y^2}{n}}} \)
Explanation. The difference \( \bar{X} - \bar{Y} \) is divided by its standard error \( \sqrt{\sigma_X^2/m + \sigma_Y^2/n} \), using the known variances. Option (d) is the statistic used when the variances are unknown.
Q12
When the population variances are known and equal, the statistic \( Z = \dfrac{\bar{X} - \bar{Y}}{\sigma\sqrt{\dfrac{1}{m} + \dfrac{1}{n}}} \) is used to test the null hypothesis
- A. \( H_0: \mu = \mu_0 \)
- B. \( H_0: \mu_X = \mu_Y \)Correct
- C. \( H_0: P_X = P_Y \)
- D. \( H_0: P = p_0 \)
Explanation. When \( \sigma_X^2 = \sigma_Y^2 = \sigma^2 \), the statistic for the equality of two population means reduces to \( (\bar{X} - \bar{Y})/\sigma\sqrt{1/m + 1/n} \). It tests \( H_0: \mu_X = \mu_Y \), as in Example 1.10.
Q13
What is the standard error of the difference between two sample proportions, \( (p_X - p_Y) \)?
- A. \( \sqrt{\hat{p}\hat{q}\left(\dfrac{1}{m} + \dfrac{1}{n}\right)} \)Correct
- B. \( \sqrt{\hat{p}\hat{q}}\left(\dfrac{1}{m} + \dfrac{1}{n}\right) \)
- C. \( \hat{p}\hat{q}\sqrt{\left(\dfrac{1}{m} + \dfrac{1}{n}\right)} \)
- D. \( \hat{p}\hat{q}\left(\dfrac{1}{m} + \dfrac{1}{n}\right) \)
Explanation. When \( P_X \) and \( P_Y \) are unknown, Table 1.4 gives \( SE(p_X - p_Y) = \sqrt{\hat{p}\hat{q}(1/m + 1/n)} \), with the pooled \( \hat{p} = \dfrac{mp_X + np_Y}{m+n} \) and \( \hat{q} = 1 - \hat{p} \). The square root covers the whole product.
Q14
Asked in board exam:
March 2026
What is level of significance?
- A. P(Type I Error)
- B. P(Type II Error)
- C. upper limit of P(Type I Error)Correct
- D. upper limit of P(Type II Error)
Explanation. The level of significance \( \alpha \) is the maximum probability with which we are willing to tolerate a type I error, that is, the upper limit of P(Type I error). It is fixed before the sample is drawn.
Q15
Large sample test for testing population proportion is valid, when
- A. \( n \) is large, \( n \geq 30 \), \( np > 5 \), \( n(1-p) > 5 \)Correct
- B. \( n \) is large, \( n \geq 30 \), \( np > 5 \), \( n(1-p) < 5 \)
- C. \( n \) is large, \( n \geq 30 \), \( np > 5 \), \( np(1-p) > 5 \)
- D. \( n \) is large, \( n \geq 30 \), \( np > 5 \), \( np(1-p) < 5 \)
Explanation. Section 1.13 and the Points to Remember: the statistic for a proportion follows N(0, 1) only when \( n \geq 30 \), \( np > 5 \) and \( n(1-p) > 5 \). In Example 1.12, \( np = 560 \) and \( n(1-p) = 440 \).
Q16
Large sample test can be applied for two-sample problems, for testing the equality of two population means, when
- A. \( m + n \geq 30 \)
- B. \( m \geq 30, n \geq 30 \)Correct
- C. \( m \geq 30, m + n \geq 30 \)
- D. \( mn \geq 30 \)
Explanation. For two samples the test is a large sample test only when the sizes of both samples are greater than or equal to 30, that is, \( m \geq 30 \) and \( n \geq 30 \). A large total \( m + n \) is not enough.
Q17
Large sample test is applicable, when the parent population is
- A. normal distribution only
- B. binomial distribution
- C. Poisson distribution
- D. any probability distributionCorrect
Explanation. For large \( n \) the test statistics have approximately the N(0, 1) sampling distribution whatever the parent population, so large sample tests can be used for any probability distribution, not only the normal.
Q18
Asked in board exam:
March 2025
What is the rejection rule, based on large sample, for testing \( H_0 \) against one-sided (left) alternative hypothesis?
- A. \( |z_0| \geq z_{\alpha/2} \)
- B. \( z_0 < -z_{\alpha} \)Correct
- C. \( z_0 \leq -z_{\alpha/2} \)
- D. \( z_0 > z_{\alpha} \)
Explanation. For a left alternative such as \( H_1: \mu < \mu_0 \), the critical value is \( -z_{\alpha} \) and \( H_0 \) is rejected when \( z_0 < -z_{\alpha} \). Option (a) is the two-tailed rule and (d) the right-tailed rule.