S IRAJ AM A CAD EMY

Let (X , X , … , X ) be a random sample from a normal population

1
with mean 𝜇 and variance 𝜎 . Let 𝑋‾ = ∑ 𝑋 and
𝑛
1
𝑆 = ∑ (𝑋 − 𝑋‾) be the random variables corresponding to the
𝑛
sample mean and sample variance. Then, prove that 𝑋‾ and S are
independently distributed and has a 𝜒 distribution (chi-square
distribution) with ( 𝑛 − 1) degrees of freedom. 7
Define unbiased and consistent estimate. 2
Find the maximum likelihood estimate for the variance 𝜎 of a
normal (𝜇, 𝜎) population, when the parameter 𝜇 is known. Show that
the estimator is unbiased. 4+1
A bivariate sample of size 11 gave the results
𝑥‾ = 7, 𝑆 = 2, 𝑦‾ = 9, 𝑆 = 4 and 𝑟 = 0.5 where
𝑥‾ = Sample mean of random variable 𝑋,
𝑦 = Sample mean of random variable 𝑌.
𝑆 = Sample standard deviation of the random variable 𝑋.
𝑆 = Sample standard deviation of the random variable Y.
𝑟 = Correlation coefficient between 𝑋 and 𝑌.
It was later found that one pair of the sample values (𝑥 = 7, 𝑦 = 9)
was inaccurate and was rejected. How would the original value of 𝑟 be
affected by the rejection? 6

The marks obtained by 17 candidates in an examination have a

mean 57 and variance 64 . Find 95% confidence limits for the mean of
the population of marks, assuming it to be normal.
 S IRAJ AM A CAD EMY
. /
Given that ∫ 𝑒 𝑑𝑥 = 0.4750 . 6
√

: Simple and Composite hypothesis ,

Type I and Type II Error , Level of significance , Test
statistic and Power of a Test. 2×4
For a normal (𝜇, 𝜎) population, construct a test for the null hypothesis
𝐻 : 𝜎 = 𝜎 against 𝐻 : 𝜎 = 𝜎 on the basis of a sample from the pop-
ulation, where 𝜇 is known and 𝜎 , 𝜎 are two given unequal numbers.
6

3×2

Sample statistic, population parameter.

Sampling distribution
Method of maximum likelihood
Law of statistical regularity.
Prove that sample mean (𝑥‾) is an unbiased estimator for the
population mean whereas sample variance is not. 2+3
A population is defined by the density function
𝑒 /
𝑥
𝑓(𝑥, 𝛼) = 0<𝑥<∞
𝛼 Γ(𝑙)
𝑙 being a known constant. Estimate the parameter 𝛼 by the method of
maximum likelihood and show that the estimate is consistent and
unbiased. 5
 S IRAJ AM A CAD EMY
The variable 𝑋 is normally distributed with mean 68 cm and s.d.
2.5 cm. What should be the size of the sample whose mean shall not
differ from the population mean by more than 1 cm with probability
0.95? 4

State Neyman Pearson lemma to obtain the best powerful

critical region for testing hypothesis. Construct a test for the null
hypothesis 𝐻 : 𝑚 = 𝑚 against alternative 𝐻 : 𝑚 ≠ 𝑚 for a normal
(𝑚, 𝜎) population by the method of likelihood ratio testing. 2+6
Given the probability density function 𝑓(𝑥, 𝜃) = 𝜃𝑒 , 0 ≤ 𝑥 < ∞,
𝜃 > 0 of a population. The null hypothesis 𝐻 : 𝜃 = 2 against the one side
alternative hypothesis 𝐻 : 𝜃 > 2 will be tested on the following
procedure :
𝐻 would be rejected if a sample drawn from the population (𝑋) is
greater than or equal to 6 . Find the type 1 error and power of test.
5+2
What do you mean by Interval Estimation? Find an approximate
confidence interval with a given confidence coefficient 1 − 𝜖(0 < 𝜖 < 1)
for ' 𝑝 ' of a binomial (𝑛, 𝑝) population assuming 𝑛 is known and large.
5

Let 𝑥 , 𝑥 , … , 𝑥 be a random sample of size 𝑛 taken from a

normal population with mean zero and standard deviation 𝜎. Show that
1
∑ 𝑥 is an unbiased estimate of 𝜎 . What do you mean by a
𝑛
consistent estimator? Examine whether the above estimate is consistent
or not. 4+1+2
 S IRAJ AM A CAD EMY
the method of maximum likehood. Estimate by the method
of maximum likehood, the parameter 𝛼 of a continuous population
having density function (1 − 𝛼)𝑥 (0 < 𝑥 < 1). 2+5
The population of scores of 10-years children in a test is known to
have a standard deviation 5.2 . If a random sample of size 20 , shows a
mean of 16.9, find 95% confidence limits for the mean score of the
population assuming that the population is normal.
⁄
Given that ∫∙ 𝑒 𝑑𝑥 = 0.025 . 6
√

What is meant by statistical hypothesis? What are the types of

error of decision committed in testing of statistical hypothesis? Explain
the concept of a critical region and power of test. Explain how the best
critical region is determined. 2+2+2+2
Let 𝑝 denote the probability of getting a head when a given coin is
1
tossed once. Suppose that the hypothesis 𝐻 : 𝑝 = is rejected in favour
2
3
of 𝐻 : 𝑝 = if more than 3 heads are obtained out of 5 throws of a coin.
4
Find the probabilities of Type I and type II errors. 6
A random sample (𝑥 , 𝑥 , … . 𝑥 ) of size 𝑛 is taken from a population
whose variance is 𝜎 . If 𝑠 = ∑ (𝑥 − 𝑥‾) , 𝑥‾ = ∑ 𝑥 , then
( )
show that the variance of the sampling distribution of 𝑠 is given by
𝜇 − ⋅ 𝜎 , where 𝜇 is the fourth central moment of population.
6

Derive an expression for an acute angle between two regression

lines of two random variates X and Y. Discuss the cases 𝜌 = 0 and
 S IRAJ AM A CAD EMY
𝜌 = ±1, where 𝜌 is the correlation coefficient between X and Y.
4+1+1
A random variable 𝑋 can take all non-negative integral values and
𝑝(𝑋 = 𝑖) = 𝑝(1 − 𝑝) , 𝑖 = 0,1,2, … where, 𝑝 is a parameter in (0,1). Find
the maximum likelihood estimate of 𝑝 on the basis of sample size n from
the population of X. How can you decide whether this estimate is
unbiased, as well as consistent? 5+2+2
Find the confidence interval for the parameter 𝑚 of a normal (m, 𝜎)
population on the basis of a sample of size n having confidence
coefficient (1 − 𝜀), assume 𝜎 is known. 5

What is null hypothesis? When do we discard these? What can

we infer about the sample characteristics? 4
What is a scatter diagram? How can we fit a parabolic curve for a
given bivariate tabular data (𝑥 , 𝑦 ), 𝑖 = 1,2,3 … 𝑛 ? What is the goodness
of fit? 2+4+2
If 𝑥 ≥ 1 be the critical region of testing of hypothesis H : 𝜃 = 2
against the alternative. H : 𝜃 = 1, on the basis of the single observation
from the population 𝑓(𝑥, 𝜃) = 𝜃𝑟 , 𝑥 ≥ 0, find the probabilities of
Type I and Type II errors. 8

What is meant by Statistical Hypothesis? Explain the term

Empirical distribution , Sample characteristics. 1+2+2
Let 𝑋 , 𝑋 , … . . , 𝑋 be a random sample of size 𝑛 taken from a
normal population with mean zero and standard deviation 𝜎.
𝑥
Show that ∑ is an unbiased estimate of 𝜎 . Examine whether this
𝑛
 S IRAJ AM A CAD EMY
estimate is consistent . 6
Find the sampling distribution of the sample mean for the Poission
distribution. 6
Show that for a population of random variable X with probability
1
density function 𝑓(𝑥) = , 0 ≤ 𝑥 ≤ 𝜃, the maximum likelihood
𝜃
estimator is the largest member of the sample values. 5

The variance of scores on a standard mathematics test for all

high school students of Calcutta was 50 in 2000. A random sample of
scores of 20 high school students who took the test this year gave a
variance of 70 . Test at 5% level of significance if the variance of the
current scores on this test differ from 50 .
[Given: 𝜒 . = 8.97 and 𝜒 . = 32.8 ] 6
Define the sample correlation coefficient 𝛾 for the sample
(𝑥 , 𝑦 ), (𝑥 , 𝑦 ), … … … , (𝑥 , 𝑦 ) from a bivariate population. The least
square regression line of 𝑦 on 𝑥 and of 𝑥 on 𝑦 are respectively
𝑥 + 3𝑦 = 0 and 3𝑥 + 2𝑦 = 0. If standard deviation 𝜎 = 1, find the least
square line of 𝑣 on 𝑢 where 𝑢 = 𝑥 + 𝑦 and 𝑣 = 𝑥 − 𝑦. 2+3+3
Given the density function
1
𝑓(𝑥, 𝜃) = , if 0 ≤ 𝑥 ≤ 𝜃,
𝜃
= 0, elsewhere,

and that you are testing the null hypothesis H : 𝜃 = 1 against 𝐻 : 𝜃 = 2

by means of a single observed value 𝑋 . Determine the size of Type I and
Type Il error if you choose the interval 0.5 ≤ 𝑥 as the critical region.
6
 S IRAJ AM A CAD EMY

Population , Sample.
Prove that the sample variance is not an unbiased estimator of
population variance. 2+4
Define Standard error of a sample statistic. Derive the standard
error of sample proportion in case of simple random sampling without
replacement (SRSWOR). 1+4
A poll is taken in which 592 out of 985 randomly selected voters
indicate their preference for a certain candidate. Construct a 95%
confidence interval to estimate the true proportion of voters who prefer
the candidate. [ Given that ∫ .
𝑒 𝑑𝑥 = 0.025 ] 5
√

Find an 𝑚. 𝑙. 𝑒 for 𝜇 in Poisson Distribution. 4

Describe the method of curve fitting to a bivariate data by least

square method. Establish the normal equations for a parabolic curve to
a given set of data (𝑥 , 𝑦 )(𝑖 = 1,2, … . . 𝑛). How can one measure the
goodness of the fit? 6+4
What is hypothesis testing? What is the power of a test? Explain
Type I and Type II error with suitable example, how it is used to accept
or reject a notional hypothesis. 2+2+3+3

Define an unbiased and consistent estimate. Prove that sample

mean is always unbiased and consistent estimate of the population
mean. 2+5
 S IRAJ AM A CAD EMY
Find sampling distribution of the mean of a binomial (𝑛, 𝑝) variate.
6
Find the maximum likelihood estimate of 𝜎 for a normal (𝑚, 𝜎)
population if 𝑚 is known and show that the estimate is unbiased and
consistent . 4+1+2

Define Confidence Interval. Find the confidence interval for the

parameter ' 𝑚 ' of a normal (𝑚, 𝜎) population on the basis of a sample of
size n having confidence coefficient (𝑖 − 𝛼). Assume that ' 𝜎 ' is known
(0 < 𝛼 < 1). 1+5
State Neyman Pearson's theorem. Using Neyman Pearson's
theorem, construct a test of null hypothesis H : 𝜎 = 𝜎 , against the
hypothesis 𝐻: 𝜎 = 𝜎 for the normal (𝑚, 𝜎) population. 7
Given the population density function 𝑓(𝑥, 𝜃) = 𝜃𝑒 ; 0 ≤ 𝑥 < ∞,
𝜃 > 0.The null hypothesis 𝐻 : 𝜃 = 2 against the onesided alternative
𝐻 : 𝜃 > 2 will be tested on the following procedure: 𝐻 , should be
rejected if a sample drawn from the population (𝑥) is greater than or
equal to 6. Find the type I error and power of the test. 5+1+1