GATE 2022 General Aptitude (GA)

Q.1 – Q.5 Carry ONE mark each.

Q.1 Mr. X speaks _________ Japanese _________ Chinese.

(A) neither / or

(B) either / nor

(C) neither / nor

(D) also / but

Q.2 A sum of money is to be distributed among P, Q, R, and S in the

proportion 5 : 2 : 4 : 3, respectively.

If R gets ₹ 1000 more than S, what is the share of Q (in ₹)?

(A) 500

(B) 1000

(C) 1500

(D) 2000

Page 1
Q.3 A trapezium has vertices marked as P, Q, R and S (in that order anticlockwise).
The side PQ is parallel to side SR.

Further, it is given that, PQ = 11 cm, QR = 4 cm, RS = 6 cm and SP = 3 cm.

What is the shortest distance between PQ and SR (in cm)?

(A) 1.80

(B) 2.40

(C) 4.20

(D) 5.76

Page 2
Q.4 The figure shows a grid formed by a collection of unit squares. The unshaded
unit square in the grid represents a hole.

What is the maximum number of squares without a “hole in the interior” that
can be formed within the 4 × 4 grid using the unit squares as building blocks?

(A) 15

(B) 20

(C) 21

(D) 26

Page 3
Q.5 An art gallery engages a security guard to ensure that the items displayed are
protected. The diagram below represents the plan of the gallery where the
boundary walls are opaque. The location the security guard posted is identified
such that all the inner space (shaded region in the plan) of the gallery is within
the line of sight of the security guard.

If the security guard does not move around the posted location and has a 360o
view, which one of the following correctly represents the set of ALL possible
locations among the locations P, Q, R and S, where the security guard can be
posted to watch over the entire inner space of the gallery.

(A) P and Q

(B) Q

(C) Q and S

(D) R and S

Page 4
Q. 6 – Q. 10 Carry TWO marks each.

Q.6 Mosquitoes pose a threat to human health. Controlling mosquitoes using

chemicals may have undesired consequences. In Florida, authorities have used
genetically modified mosquitoes to control the overall mosquito population. It
remains to be seen if this novel approach has unforeseen consequences.

Which one of the following is the correct logical inference based on the
information in the above passage?

(A) Using chemicals to kill mosquitoes is better than using genetically modified
mosquitoes because genetic engineering is dangerous

(B) Using genetically modified mosquitoes is better than using chemicals to kill
mosquitoes because they do not have any side effects

(C) Both using genetically modified mosquitoes and chemicals have undesired
consequences and can be dangerous

(D) Using chemicals to kill mosquitoes may have undesired consequences but it is
not clear if using genetically modified mosquitoes has any negative
consequence

Page 5
 Consider the following inequalities.
Q.7
(i) 2𝑥 − 1 > 7
(ii) 2𝑥 − 9 < 1

Which one of the following expressions below satisfies the above two
inequalities?

(A) 𝑥 ≤ −4

(B) −4 < 𝑥 ≤ 4

(C) 4 < 𝑥 < 5

(D) 𝑥 ≥ 5

Four points P(0, 1), Q(0, −3), R(−2, −1), and S(2, −1) represent the vertices
Q.8
of a quadrilateral.

What is the area enclosed by the quadrilateral?

(A) 4

(B) 4 2

(C) 8

(D) 8√2

Page 6
Q.9 In a class of five students P, Q, R, S and T, only one student is known to have
copied in the exam. The disciplinary committee has investigated the situation
and recorded the statements from the students as given below.

Statement of P: R has copied in the exam.

Statement of Q: S has copied in the exam.

Statement of R: P did not copy in the exam.

Statement of S: Only one of us is telling the truth.

Statement of T: R is telling the truth.

The investigating team had authentic information that S never lies.

Based on the information given above, the person who has copied in the exam is

(A) R

(B) P

(C) Q

(D) T

Page 7
Q.10 Consider the following square with the four corners and the
center marked as P, Q, R, S and T respectively.

Let X, Y and Z represent the following operations:

X: rotation of the square by 180 degree with respect to the

S-Q axis.

Y: rotation of the square by 180 degree with respect to the P-R axis.

Z: rotation of the square by 90 degree clockwise with respect to the axis

perpendicular, going into the screen and passing through the point T.

Consider the following three distinct sequences of operation (which are applied
in the left to right order).

(1) XYZZ
(2) XY
(3) ZZZZ

Which one of the following statements is correct as per the information

provided above?

(A) The sequence of operations (1) and (2) are equivalent

(B) The sequence of operations (1) and (3) are equivalent

(C) The sequence of operations (2) and (3) are equivalent

(D) The sequence of operations (1), (2) and (3) are equivalent

Page 8
GATE 2022 Statistics (ST)

Q.11 – Q.35 Carry ONE mark Each

Q.11 Let 𝑴 be a 2 × 2 real matrix such that (𝑰 + 𝑴) = 𝑰 − 𝛼𝑴, where 𝛼 is a non-zero

real number and 𝑰 is the 2 × 2 identity matrix. If the trace of the matrix 𝑴 is 3, then
the value of 𝛼 is

(A) 3
4

(B) 1
3

(C) 1
2

(D) 1
4

Page 9
GATE 2022 Statistics (ST)

Q.12 Let {𝑋(𝑡)} be a linear pure death process with death rate
𝜇 = 5𝑖, 𝑖 = 0, 1, … , 𝑁, 𝑁 ≥ 1 . Suppose that 𝑝 (𝑡) = 𝑃(𝑋(𝑡) = 𝑖). Then the
system of forward Kolmogorov’s equations is

(A) 𝑑𝑝 (𝑡) 𝑑𝑝 (𝑡)

= 5(𝑖 + 1)𝑝 (𝑡) + 5𝑖 𝑝 (𝑡) and = 5𝑁𝑝 (𝑡)
𝑑𝑡 𝑑𝑡
for 𝑖 = 0, 1, 2, … , 𝑁 − 1 with initial conditions 𝑝 (0) = 0 for 𝑖 ≠ 𝑁, and 𝑝 (0) = 1

(B) 𝑑𝑝 (𝑡) 𝑑𝑝 (𝑡)

= 5(𝑖 + 1)𝑝 (𝑡) − 5𝑖 𝑝 (𝑡) and = −5𝑁𝑝 (𝑡)
𝑑𝑡 𝑑𝑡
for 𝑖 = 0, 1, 2, … , 𝑁 − 1 with initial conditions 𝑝 (0) = 0 for 𝑖 ≠ 𝑁, and 𝑝 (0) = 1

(C) 𝑑𝑝 (𝑡) 𝑑𝑝 (𝑡)

= 5(𝑖 + 1)𝑝 (𝑡) + 5𝑖 𝑝 (𝑡) and = 5𝑁𝑝 (𝑡)
𝑑𝑡 𝑑𝑡
for 𝑖 = 0, 1, 2, … , 𝑁 − 1 with initial conditions 𝑝 (0) = 1 for 𝑖 ≠ 𝑁, and 𝑝 (0) = 0

(D) 𝑑𝑝 (𝑡) 𝑑𝑝 (𝑡)

= 5(𝑖 + 1)𝑝 (𝑡) − 5𝑖 𝑝 (𝑡) and = −5𝑁𝑝 (𝑡)
𝑑𝑡 𝑑𝑡
for 𝑖 = 0, 1, 2, … , 𝑁 − 1 with initial conditions 𝑝 (0) = 1 for 𝑖 ≠ 𝑁, and 𝑝 (0) = 0

Page 10
GATE 2022 Statistics (ST)

Q.13 Let 𝑆 be the variance of a random sample of size 𝑛 > 1 from a normal population
with an unknown mean 𝜇 and an unknown finite variance 𝜎 > 0. Consider the
following statements:

(I) 𝑆 is an unbiased estimator of 𝜎 , and 𝑆 is an unbiased estimator of 𝜎.

(II) 𝑆 is a maximum likelihood estimator of 𝜎 , and 𝑆 is a
maximum likelihood estimator of 𝜎.

Which of the above statements is/are true?

(A) (I) only

(B) (II) only

(C) Both (I) and (II)

(D) Neither (I) nor (II)

Page 11
GATE 2022 Statistics (ST)

Q.14 Let 𝑓: ℝ → ℝ be a function defined by

𝑥 𝑦
, (𝑥, 𝑦) ≠ (0,0),
𝑓 (𝑥, 𝑦) = 𝑥 +𝑦
0, (𝑥, 𝑦) = (0,0).

Then which one of the following statements is true?

(A) 𝑓 is bounded and is unbounded on ℝ

(B) 𝑓 is unbounded and is bounded on ℝ

(C) Both 𝑓 and are unbounded on ℝ

(D) Both 𝑓 and are bounded on ℝ

Page 12
GATE 2022 Statistics (ST)

Q.15 Let 𝑋 , 𝑋 , … , 𝑋 be a random sample from a distribution with cumulative

distribution function 𝐹 (𝑥 ). Let the empirical distribution function of the sample be
𝐹 (𝑥 ). The classical Kolmogorov-Smirnov goodness of fit test statistic is given by

𝑇 = √𝑛 𝐷 = √𝑛 sup |𝐹 (𝑥) − 𝐹(𝑥)|.

Consider the following statements:

(I) The distribution of 𝑇 is the same for all continuous underlying

distribution functions 𝐹(𝑥).
(II) 𝐷 converges to 0 almost surely, as 𝑛 → ∞.

Which of the above statements is/are true?

(A) (I) only

(B) (II) only

(C) Both (I) and (II)

(D) Neither (I) nor (II)

Page 13
GATE 2022 Statistics (ST)

Q.16 Consider the following transition matrices 𝑷𝟏 and 𝑷𝟐 of two Markov chains:

1 0 0 1/6 1/3 1/2

𝑷𝟏 = 1/3 1/2 1/6 and 𝑷𝟐 = 1/4 0 3/4 .
0 0 1 0 1 0
Then which one of the following statements is true?

(A) Both 𝑷𝟏 and 𝑷𝟐 have unique stationary distributions

(B) 𝑷𝟏 has a unique stationary distribution, but 𝑷𝟐 has infinitely many stationary
distributions

(C) 𝑷𝟏 has infinitely many stationary distributions, but 𝑷𝟐 has a unique stationary
distribution

(D) Neither 𝑷𝟏 nor 𝑷𝟐 has unique stationary distribution

Page 14
GATE 2022 Statistics (ST)

Q.17 Let 𝑿𝟏 , 𝑿𝟐 , … , 𝑿𝟐𝟎 be a random sample of size 20 from Ν (𝝁, 𝚺), with
det(𝚺) ≠ 0, and suppose both 𝝁 and 𝚺 are unknown. Let

1 1
𝑿= 𝑿 and 𝑺= (𝑿 − 𝑿) (𝑿 − 𝑿) .
20 19

Consider the following two statements:

(I) The distribution of 19 𝑺 is 𝑊 (19, 𝚺) (Wishart distribution of order 6

with 19 degrees of freedom).
(II) The distribution of (𝑿𝟑 − 𝝁) 𝑺 (𝑿𝟑 − 𝝁) is 𝜒 (Chi-square
distribution with 6 degrees of freedom).

Then which of the above statements is/are true?

(A) (I) only

(B) (II) only

(C) Both (I) and (II)

(D) Neither (I) nor (II)

Page 15
GATE 2022 Statistics (ST)

Q.18 Let 𝑋 , 𝑋 , … , 𝑋 be a random sample from the distribution

2𝑥 /
𝑒 , 𝑥 > 0,
𝑓(𝑥; 𝜃) = 𝜃
0, 𝑥 ≤ 0.

Let 𝜒 , denote the value of a Chi-square random variable 𝑌 with 𝑛 degrees of

freedom such that 𝑃 𝑌 > 𝜒 , = 𝛼. If 𝑥 , 𝑥 , … , 𝑥 is a realization of this random
sample, then, based on the sufficient statistic ∑ 𝑋 , which one of the following
is a 98% confidence interval for 𝜃?

(A) 2∑ 𝑥 2∑ 𝑥
,
𝜒 . , 𝜒. ,

(B) 2∑ 𝑥 2∑ 𝑥
,
𝜒 . , 𝜒. ,

(C) ∑ 𝑥 ∑ 𝑥
,
𝜒 . , 𝜒 . ,

(D) ∑ 𝑥 ∑ 𝑥
,
𝜒 . , 𝜒 . ,

Page 16
GATE 2022 Statistics (ST)

Q.19 Let 𝑋 , 𝑋 , … , 𝑋 be a random sample from a population 𝑓(𝑥; 𝜃), where 𝜃 is a

parameter. Then which one of the following statements is NOT true?

(A) ∑ 𝑋 is a complete and sufficient statistic for 𝜃, if

𝑒 𝜃
𝑓 (𝑥; 𝜃 ) = , 𝑥 = 0, 1, 2, … , and 𝜃 > 0
𝑥!

(B) (∑ 𝑋 , ∑ 𝑋 ) is a complete and sufficient statistic for 𝜃, if

1 ( )
𝑓(𝑥; 𝜃) = 𝑒 , −∞ < 𝑥 < ∞, 𝜃 > 0
√2𝜋 𝜃

(C) 𝑓 (𝑥; 𝜃 ) = 𝜃𝑥 , 0 < 𝑥 < 1, 𝜃 > 0 has monotone likelihood ratio property in
∏ 𝑋

(D) 𝑋( ) − 𝑋( ) is ancillary statistic for 𝜃 if 𝑓(𝑥; 𝜃) = 1, 0 < 𝜃 < 𝑥 < 𝜃 + 1, where

𝑋( ) = min{𝑋 , 𝑋 , … , 𝑋 } and 𝑋( ) = max{𝑋 , 𝑋 , … , 𝑋 }

Page 17
GATE 2022 Statistics (ST)

Q.20 A random sample 𝑋 , 𝑋 , … , 𝑋 of size 6 is taken from a Bernoulli distribution with

the parameter 𝜃. The null hypothesis 𝐻 : 𝜃 = is to be tested against the alternative
hypothesis 𝐻 : 𝜃 > , based on the statistic 𝑌 = ∑ 𝑋 . If the value of 𝑌
corresponding to the observed sample values is 4, then the 𝑝-value of the test
statistic is

(A) 21
32

(B) 9
64

(C) 11
32

(D) 7
64

Page 18
GATE 2022 Statistics (ST)

Q.21 Let {𝑎 } be a sequence of positive real numbers satisfying

8 7 𝑎
= + , 𝑛≥1
𝑎 𝑎 343

with 𝑎 = 3 and 𝑎 < 7 for all 𝑛 ≥ 2.

Consider the following statements:

(I) {𝑎 } is monotonically increasing.

(II) {𝑎 } converges to a value in the interval [3, 7].

Then which of the above statements is/are true?

(A) (I) only

(B) (II) only

(C) Both (I) and (II)

(D) Neither (I) nor (II)

Q.22 Let 𝑴 be any square matrix of arbitrary order 𝑛 such that 𝑴 = 𝟎 and the nullity
of 𝑴 is 6. Then the maximum possible value of 𝑛 (in integer) is ______

Page 19
GATE 2022 Statistics (ST)

Q.23 Consider the usual inner product in ℝ . Let 𝒖 ∈ ℝ be a unit vector orthogonal to
the subspace

𝑆 = {(𝑥 , 𝑥 , 𝑥 , 𝑥 ) ∈ ℝ | 𝑥 + 𝑥 + 𝑥 + 𝑥 = 0}.

If 𝒗 = (1, −2, 1, 1) , and the vectors 𝒖 and 𝒗 − 𝛼𝒖, 𝛼 ∈ ℝ, are orthogonal, then the
value of 𝛼 (rounded off to two decimal places) is equal to ______

Q.24 Let {𝐵(𝑡)} be a standard Brownian motion and let Φ(⋅) be the cumulative
distribution function of the standard normal distribution. If
1
𝑃 𝐵(2) + 2𝐵(3) > 1 = 1 − Φ , 𝛼 > 0,
√𝛼
then the value of 𝛼 (in integer) is equal to ______

Q.25 Let 𝑋 and 𝑌 be two independent exponential random variables with 𝐸(𝑋 ) = and
𝐸(𝑌 ) = . Then 𝑃(𝑋 < 2𝑌) (rounded off to two decimal places) is equal
to ______

Q.26
Let 𝑋 be a random variable with the probability mass function 𝑝 (𝑥 ) = ,
𝑥 = 1, 2, 3, … . Then the value of

𝑃(𝑛 < 𝑋 ≤ 𝑛 + 3)

(rounded off to two decimal places) is equal to _______

Page 20
GATE 2022 Statistics (ST)

Q.27 Let 𝑋 , 𝑖 = 1, 2, … , 𝑛, be 𝑖. 𝑖. 𝑑. random variables from a normal distribution with

mean 1 and variance 4. Let 𝑆 = 𝑋 + 𝑋 + ⋯ + 𝑋 . If 𝑉𝑎𝑟(𝑆 ) denotes the
variance of 𝑆 , then the value of

𝑉𝑎𝑟(𝑆 ) 𝐸(𝑆 )
lim −
→ 𝑛 𝑛

(in integer) is equal to _______

Q.28 At a telephone exchange, telephone calls arrive independently at an average rate of

1 call per minute, and the number of telephone calls follows a Poisson distribution.
Five time intervals, each of duration 2 minutes, are chosen at random. Let 𝑝 denote
the probability that in each of the five time intervals at most 1 call arrives at the
telephone exchange. Then 𝑒 𝑝 (in integer) is equal to _______

Q.29 Let 𝑋 be a random variable with the probability density function

𝑐(𝑥 − [𝑥]), 0 < 𝑥 < 3,

𝑓(𝑥 ) =
0, elsewhere,

where 𝑐 is a constant and [𝑥] denotes the greatest integer less than or equal to 𝑥. If
𝐴 = , 2 , then 𝑃(𝑋 ∈ 𝐴) (rounded off to two decimal places) is equal to ______

Page 21
GATE 2022 Statistics (ST)

Q.30 Let 𝑋 and 𝑌 be two random variables such that the moment generating function of
𝑋 is 𝑀(𝑡 ) and the moment generating function of 𝑌 is

3 1
𝐻(𝑡 ) = 𝑒 + 𝑀(𝑡 ),
4 4

where 𝑡 ∈ (−ℎ, ℎ), ℎ > 0. If the mean and the variance of 𝑋 are and ,
respectively, then the variance of 𝑌 (in integer) is equal to _______

Q.31 Let 𝑋 , 𝑖 = 1, 2, … 𝑛, be 𝑖. 𝑖. 𝑑. random variables with the probability density

function

/ /
𝑥 𝑒 , 0 < 𝑥 < ∞,
√
𝑓 (𝑥 ) =
0, elsewhere,

where Γ(⋅) denotes the gamma function. Also, let 𝑋 = (𝑋 + 𝑋 + ⋯ + 𝑋 ). If

√ 𝑛 𝑋 (3 − 𝑋 ) − converges to Ν(0, 𝜎 ) in distribution, then 𝜎 (rounded

off to two decimal places) is equal to _______

Q.32 Consider a Poisson process {𝑋(𝑡), 𝑡 ≥ 0}. The probability mass function of 𝑋(𝑡) is
given by

𝑒 (4𝑡)
𝑓 (𝑡) = , 𝑛 = 0, 1, 2, …
𝑛!
If 𝐶 (𝑡 , 𝑡 ) is the covariance function of the Poisson process, then the value of
𝐶(5, 3) (in integer) is equal to ______

Page 22
GATE 2022 Statistics (ST)

Q.33 A random sample of size 4 is taken from the distribution with the probability density
function

2(𝜃 − 𝑥)
𝑓(𝑥; 𝜃) = , 0 < 𝑥 < 𝜃,
𝜃
0, elsewhere.

If the observed sample values are 6, 5, 3, 6, then the method of moments estimate
(in integer) of the parameter 𝜃, based on these observations, is ______

Q.34 A company sometimes stops payments of quarterly dividends. If the company pays
the quarterly dividend, the probability that the next one will be paid is 0.7. If the
company stops the quarterly dividend, the probability that the next quarterly
dividend will not be paid is 0.5. Then the probability (rounded off to three
decimal places) that the company will not pay quarterly dividend in the long run
is ______

Q.35 Let 𝑋 , 𝑋 , … , 𝑋 be a random sample taken from a distribution with the probability
density function

𝑥
, 0 < 𝑥 < 4,
𝑓 (𝑥) = 8
0, elsewhere.

Let 𝐹 (𝑥 ) be the empirical distribution function of the sample. If 𝛼 is the variance

of 𝐹 (2), then 128𝛼 (in integer) is equal to ______

Page 23
GATE 2022 Statistics (ST)

Q.36 – Q.65 Carry TWO marks Each

Q.36 Let 𝑴 be a 3 × 3 real symmetric matrix with eigenvalues −1, 1, 2 and the
corresponding unit eigenvectors 𝒖, 𝒗, 𝒘, respectively. Let 𝒙 and 𝒚 be two vectors in
ℝ such that

𝑴𝒙 = 𝒖 + 2(𝒗 + 𝒘) and 𝑴 𝒚 = 𝒖 − (𝒗 + 2𝒘).

Considering the usual inner product in ℝ , the value of |𝒙 + 𝒚| , where |𝒙 + 𝒚| is

the length of the vector 𝒙 + 𝒚, is

(A) 1.25

(B) 0.25

(C) 0.75

(D) 1

Page 24
GATE 2022 Statistics (ST)

Q.37 Consider the following infinite series:

𝑛
𝑆 := (−1) and 𝑆 := (−1) 𝑛 +1 −𝑛 .
𝑛 +4

Which of the above series is/are conditionally convergent?

(A) 𝑆 only

(B) 𝑆 only

(C) Both 𝑆 and 𝑆

(D) Neither 𝑆 nor 𝑆

Q.38 Let (3, 6) , (4, 4) , (5, 7) and (4, 7) be four independent observations from a
bivariate normal distribution with the mean vector 𝝁 and the covariance matrix 𝚺. Let
𝝁 and 𝚺 be the maximum likelihood estimates of 𝝁 and 𝚺, respectively, based on
these observations. Then 𝚺𝝁 is equal to

(A) 3.5
10

(B) 7.5
4

(C) 4
13.5

(D) 10
3.5

Page 25
GATE 2022 Statistics (ST)

Q.39 𝑋 2 4 −1 1
Let 𝑿 = 𝑋 follow Ν (𝝁, 𝚺) with 𝝁 = −3 and 𝚺 = −1 2 𝑎 , where
𝑋 2 1 𝑎 2
𝑎 ∈ ℝ. Suppose that the partial correlation coefficient between 𝑋 and 𝑋 , keeping
𝑋 fixed, is . Then 𝑎 is equal to

(A) 1

(B) 3
2

(C) 2

(D) 1
2

Page 26
GATE 2022 Statistics (ST)

Q.40 If the line 𝑦 = 𝛼𝑥, 𝛼 ≥ √2, divides the area of the region

𝑅: = {(𝑥, 𝑦) ∈ ℝ | 0 ≤ 𝑥 ≤ 𝑦 , 0 ≤ 𝑦 ≤ 2}

into two equal parts, then the value of 𝛼 is equal to

(A) 3
√2

(B) 2√2

(C) √2

(D) 5
2√2

Page 27
GATE 2022 Statistics (ST)

Q.41 Let (𝑋, 𝑌, 𝑍) be a random vector with the joint probability density function

1
(2𝑥 + 3𝑦 + 𝑧), 0 < 𝑥 < 1, 0 < 𝑦 < 1, 0 < 𝑧 < 1,
𝑓 , , (𝑥, 𝑦, 𝑧) = 3
0, elsewhere.

Then which one of the following points is on the regression surface of 𝑋 on (𝑌, 𝑍) ?

(A) 4 1 1
, ,
7 3 3

(B) 6 2 2
, ,
7 3 3

(C) 1 1 2
, ,
2 3 3

(D) 1 2 1
, ,
2 3 3

Page 28
GATE 2022 Statistics (ST)

Q.42 A random sample 𝑋 of size one is taken from a distribution with the probability
density function

2𝑥
, 0 < 𝑥 < 𝜃,
𝑓(𝑥; 𝜃) = 𝜃
0, elsewhere.

If is used as a pivot for obtaining the confidence interval for 𝜃, then which one of
the following is an 80% confidence interval (confidence limits rounded off to three
decimal places) for 𝜃 based on the observed sample value 𝑥 = 10 ?

(A) (10.541, 31.623)

(B) (10.987, 31.126)

(C) (11.345, 30.524)

(D) (11.267, 30.542)

Page 29
GATE 2022 Statistics (ST)

Q.43 Let 𝑋 , 𝑋 , …, 𝑋 be a random sample from a normal population with mean 0 and
variance 𝜃 > 0. Let

𝑋 +𝑋
𝐾= .
𝑋 +𝑋 +⋯+𝑋

Consider the following statements:

(I) The statistics 𝐾 and 𝑋 + 𝑋 + ⋯ + 𝑋 are independent.

(II) has an 𝐹-distribution with 2 and 7 degrees of freedom.

(III) 𝐸 (𝐾 ) = .

Then which of the above statements is/are true?

(A) (I) and (II) only

(B) (I) and (III) only

(C) (II) and (III) only

(D) (I) only

Page 30
GATE 2022 Statistics (ST)

Q.44 Consider the following statements:

(I) Let a random variable 𝑋 have the probability density function

1 | |
𝑓 (𝑥) = 𝑒 , −∞ < 𝑥 < ∞.
2

Then there exist 𝑖. 𝑖. 𝑑. random variables 𝑋 and 𝑋 such that 𝑋 and

𝑋 − 𝑋 have the same distribution.

(II) Let a random variable 𝑌 have the probability density function

1
, − 2 < 𝑦 < 2,
𝑓 (𝑦) = 4
0, elsewhere.

Then there exist 𝑖. 𝑖. 𝑑. random variables 𝑌 and 𝑌 such that 𝑌 and

𝑌 − 𝑌 have the same distribution.

Then which of the above statements is/are true?

(A) (I) only

(B) (II) only

(C) Both (I) and (II)

(D) Neither (I) nor (II)

Page 31
GATE 2022 Statistics (ST)

Q.45 Suppose 𝑋 , 𝑋 , … , 𝑋 , … are independent exponential random variables with the

mean . Let the notation 𝑖. 𝑜. denote ‘infinitely often’. Then which of the following
is/are true?

(A) 𝑃 𝑋 > log 𝑛 𝑖. 𝑜. = 1 for 0 < 𝜖 ≤ 1

(B) 𝑃 𝑋 < log 𝑛 𝑖. 𝑜. = 1 for 0 < 𝜖 ≤ 1

(C) 𝑃 𝑋 > log 𝑛 𝑖. 𝑜. = 1 for ϵ > 1

(D) 𝑃 𝑋 < log 𝑛 𝑖. 𝑜. = 1 for ϵ > 1

Page 32
GATE 2022 Statistics (ST)

Q.46 Let {𝑋 }, 𝑛 ≥ 1, be a sequence of random variables with the probability mass

functions
𝑛
⎧𝑛 + 1 , 𝑥 = 0,
⎪
𝑝 (𝑥) = 1
⎨𝑛 + 1 , 𝑥 = 𝑛,
⎪
⎩ 0, elsewhere.

Let 𝑋 be a random variable with 𝑃(𝑋 = 0) = 1. Then which of the following

statements is/are true?

(A) 𝑋 converges to 𝑋 in distribution

(B) 𝑋 converges to 𝑋 in probability

(C) 𝐸(𝑋 ) ⎯ 𝐸(𝑋)

(D) There exists a subsequence {𝑋 } of {𝑋 } such that 𝑋 converges to 𝑋 almost

surely

Page 33
GATE 2022 Statistics (ST)

Q.47 Let 𝑴 be any 3 × 3 symmetric matrix with eigenvalues 1, 2 and 3. Let 𝑵 be any
3 × 3 matrix with real eigenvalues such that 𝑴𝑵 + 𝑵 𝑴 = 3𝑰, where 𝑰 is the
3 × 3 identity matrix. Then which of the following cannot be eigenvalue(s) of the
matrix 𝑵 ?

(A) 1
4

3
(B)
4
1
(C)
2
7
(D)
4

Q.38 Let 𝑴 be a 3 × 2 real matrix having a singular value decomposition as 𝑴 = 𝑼𝑺𝑽 ,

where the matrix 𝑺 = √3 0 0 , 𝑼 is a 3 × 3 orthogonal matrix, and 𝑽 is a

0 1 0
2 × 2 orthogonal matrix. Then which of the following statements is/are true?

(A) The rank of the matrix 𝑴 is 1

(B) The trace of the matrix 𝑴 𝑴 is 4

(C) The largest singular value of the matrix (𝑴 𝑴) 𝑴 is 1

(D) The nullity of the matrix 𝑴 is 1

Page 34
GATE 2022 Statistics (ST)

Q.49 Let 𝑋 be a random variable such that

𝑎
𝑃 𝑋 ∈ ℤ = 1, 𝑎 > 0,
2𝜋
where ℤ denotes the set of all integers. If 𝜙 (𝑡), 𝑡 ∈ ℝ, denotes the characteristic
function of 𝑋, then which of the following is/are true?

(A) 𝜙 (𝑎) = 1

(B) 𝜙 (⋅) is periodic with period 𝑎

(C) |𝜙 (𝑡)| < 1 for all 𝑡 ≠ 𝑎

(D) ∫ 𝑒 𝜙 (𝑡)𝑑𝑡 = 𝜋 𝑃 𝑋 = , 𝑛 ∈ ℤ, 𝑖 = √−1

Q.50 Which of the following real valued functions is/are uniformly continuous on
[0, ∞)?

(A) sin 𝑥

(B) 𝑥 sin 𝑥

(C) sin(sin 𝑥)

(D) sin(𝑥 sin 𝑥)

Page 35
GATE 2022 Statistics (ST)

Q.51 Two independent random samples, each of size 7, from two populations yield the
following values:

Population 1 18 20 16 20 17 18 14

Population 2 17 18 14 20 14 13 16

If Mann-Whitney 𝑈 test is performed at 5% level of significance to test the null

hypothesis 𝐻 : Distributions of the populations are same, against the alternative
hypothesis 𝐻 : Distributions of the populations are not same, then the value of the
test statistic 𝑈 (in integer) for the given data, is ______

Q.52 Consider the multiple regression model

𝑌 = 𝛽 + 𝛽 𝑋 + 𝛽 𝑋 + 𝛽 𝑋 + 𝜖,

where 𝜖 is normally distributed with mean 0 and variance 𝜎 > 0, and

𝛽 , 𝛽 , 𝛽 , 𝛽 are unknown parameters. Suppose 52 observations of (𝑌, 𝑋 , 𝑋 , 𝑋 )
yield sum of squares due to regression as 18.6 and total sum of squares as 79.23.
Then, for testing the null hypothesis 𝐻 : 𝛽 = 𝛽 = 𝛽 = 0 against the alternative
hypothesis 𝐻 : 𝛽 ≠ 0 for some 𝑖 = 1, 2, 3, the value of the test statistic (rounded
off to three decimal places), based on one way analysis of variance, is _______

Page 36
GATE 2022 Statistics (ST)

Q.53 Suppose a random sample of size 3 is taken from a distribution with the probability
density function

2𝑥, 0 < 𝑥 < 1,

𝑓(𝑥 ) =
0, elsewhere.

If 𝑝 is the probability that the largest sample observation is at least twice the smallest
sample observation, then the value of 𝑝 (rounded off to three decimal places)
is ______

Q.54 Let a linear model 𝑌 = 𝛽 + 𝛽 𝑋 + 𝜖 be fitted to the following data, where 𝜖 is

normally distributed with mean 0 and unknown variance 𝜎 > 0.

𝑥 0 1 2 3 4

𝑦 3 4 5 6 7

Let 𝑌 denote the ordinary least-square estimator of 𝑌 at 𝑋 = 6, and the variance of

of 𝑌 = 𝑐𝜎 . Then the value of the real constant 𝑐 (rounded off to one decimal
place) is equal to ______

Q.55 Let 0, 1, 1, 2, 0 be five observations of a random variable 𝑋 which follows a Poisson

distribution with the parameter 𝜃 > 0. Let the minimum variance unbiased estimate
of 𝑃(𝑋 ≤ 1), based on this data, be 𝛼. Then 5 𝛼 (in integer) is equal to ______

Page 37
GATE 2022 Statistics (ST)

Q.56 While calculating Spearman’s rank correlation coefficient, based on 𝑛 observations

{(𝑥 , 𝑦 ), 𝑖 = 1,2, … , 𝑛} from a paired data, it is found that 𝑥 are distinct for all
𝑖 ≥ 2, 𝑥 = 𝑥 , and ∑ 𝑑 = 19.5, where 𝑑 = rank(𝑥 ) − rank(𝑦 ). Then the
minimum possible value of 𝑛 − 𝑛 (in integer) is ______

Q.57 In a laboratory experiment, the behavior of cats are studied for a particular food
preference between two foods A and B. For an experiment, 70% of the cats that
had food A will prefer food A, and 50% of the cats that had food B will prefer food
A. The experiment is repeated under identical conditions. If 40% of the cats had
food A in the first experiment, then the percentage (rounded off to one decimal
place) of cats those will prefer food A in the third experiment, is ______

Q.58 A random sample of size 5 is taken from a distribution with the probability density
function

3𝑥
, 0 < 𝑥 < 𝜃,
𝑓 (𝑥; 𝜃 ) = 𝜃
0, elsewhere,

where 𝜃 is an unknown parameter. If the observed values of the random sample are
3, 6, 4, 7, 5, then the maximum likelihood estimate of the th quantile of the
distribution (rounded off to one decimal place) is ______

Page 38
GATE 2022 Statistics (ST)

Q.59 Consider a gamma distribution with the probability density function

1 /
𝑥 𝑒 , 𝑥 > 0,
𝑓(𝑥; 𝛽) = 24 𝛽
0, elsewhere,

with 𝛽 > 0. Then, for 𝛽 = 2, the value of the Cramer-Rao lower bound (rounded
off to one decimal place) for the variance of any unbiased estimator of 𝛽 , based
on a random sample of size 8 from this distribution, is ______

Q.60 Let 𝑋 , 𝑋 , 𝑋 , 𝑋 be a random sample of size four from a Bernoulli distribution

with the parameter 𝜃, 0 < 𝜃 < 1. Consider the null hypothesis 𝐻 : 𝜃 = against
the alternative hypothesis 𝐻 : 𝜃 > . Suppose 𝐻 is rejected if and only if
𝑋 + 𝑋 + 𝑋 + 𝑋 > 2. If 𝛼 is the probability of Type I error for the test and 𝛾(𝜃)
is the power function of the test, then the value of 16𝛼 + 7 𝛾 (in integer) is
equal to ______

Q.61 Given that Φ(1.645) = 0.95 and Φ(2.33) = 0.99, where Φ(⋅) denotes the
cumulative distribution function of a standard normal random variable. For a
random sample 𝑋 , 𝑋 , … , 𝑋 from a normal population Ν(𝜇, 2 ), where 𝜇 is
unknown, the null hypothesis 𝐻 : 𝜇 = 10 is to be tested against the alternative
hypothesis 𝐻 : 𝜇 = 12. Suppose that a test that rejects 𝐻 if the sample mean 𝑋 is
large, is used. Then the smallest value of 𝑛 (in integer) such that Type I error is
0.05 and Type II error is at most 0.01, is ______

Page 39
GATE 2022 Statistics (ST)

Q.62 Let 𝑌 < 𝑌 < ⋯ < 𝑌 be the order statistics of a random sample of size 𝑛 from
a continuous distribution, which is symmetric about its mean 𝜇. Then the smallest
value of 𝑛 (in integer) such that 𝑃(𝑌 < 𝜇 < 𝑌 ) ≥ 0.99, is ______

Q.63 If 𝑃(𝑥, 𝑦, 𝑧) is a point which is nearest to the origin and lies on the intersection of
the surfaces 𝑧 = 𝑥𝑦 + 5 and 𝑥 + 𝑦 + 𝑧 = 1. Then the distance (in integer)
between the origin and the point 𝑃 is ______

Q.64 Let 𝑋 and 𝑌 be random variables such that 𝑋 is uniformly distributed over (0, 4),
and the conditional distribution of 𝑌 given 𝑋 = 𝑥 is uniformly distributed over
0, . Then 𝐸(𝑌 ) (rounded off to three decimal places) is equal to _____

Q.65 Let 𝑿 = (𝑋 , 𝑋 , 𝑋 ) be a random vector with the distribution Ν (𝝁, 𝚺), where

3 4 2 1
𝝁= 2 and 𝚺= 2 3 0 .
4 1 0 2
Then 𝐸(𝑋 |(𝑋 = 4, 𝑋 = 7)) (in integer) is equal to ______

Page 40
 Statistics (ST)

General Aptitude (GA)

Q.1 – Q.5 Multiple Choice Question (MCQ), carry ONE mark each (for each wrong
answer: – 1/3).

Q.1 The current population of a city is 11,02,500. If it has been increasing at the
rate of 5% per annum, what was its population 2 years ago?

(A) 9,92,500

(B) 9,95,006

(C) 10,00,000

(D) 12,51,506

Q.2 p q
p and q are positive integers and + = 3,
q p

p2 q2
then, + =
q2 p2

(A) 3

(B) 7

(C) 9

(D) 11

ST - Copyright © GATE 2021 Page 1 of 27

 Statistics (ST)

Q.3

The least number of squares that must be added so that the line P-Q
becomes the line of symmetry is ________

(A) 4

(B) 3

(C) 6

(D) 7

Q.4 Nostalgia is to anticipation as _______ is to ________

Which one of the following options maintains a similar logical relation in the
above sentence?

(A) Present, past

(B) Future, past

(C) Past, future

(D) Future, present

ST - Copyright © GATE 2021 Page 2 of 27

 Statistics (ST)

Q.5 Consider the following sentences:

(i) I woke up from sleep.
(ii) I woked up from sleep.
(iii) I was woken up from sleep.
(iv) I was wokened up from sleep.
Which of the above sentences are grammatically CORRECT?

(A) (i) and (ii)

(B) (i) and (iii)

(C) (ii) and (iii)

(D) (i) and (iv)

ST - Copyright © GATE 2021 Page 3 of 27

 Statistics (ST)

Q. 6 – Q. 10 Multiple Choice Question (MCQ), carry TWO marks each (for each wrong
answer: – 2/3).

Q.6 Given below are two statements and two conclusions.

Statement 1: All purple are green.
Statement 2: All black are green.
Conclusion I: Some black are purple.
Conclusion II: No black is purple.
Based on the above statements and conclusions, which one of the following
options is logically CORRECT?

(A) Only conclusion I is correct.

(B) Only conclusion II is correct.

(C) Either conclusion I or II is correct.

(D) Both conclusion I and II are correct.

ST - Copyright © GATE 2021 Page 4 of 27

 Statistics (ST)

Q.7 Computers are ubiquitous. They are used to improve efficiency in almost all
fields from agriculture to space exploration. Artificial intelligence (AI) is
currently a hot topic. AI enables computers to learn, given enough training
data. For humans, sitting in front of a computer for long hours can lead to
health issues.
Which of the following can be deduced from the above passage?
(i) Nowadays, computers are present in almost all places.
(ii) Computers cannot be used for solving problems in engineering.
(iii) For humans, there are both positive and negative effects of using
computers.
(iv) Artificial intelligence can be done without data.

(A) (ii) and (iii)

(B) (ii) and (iv)

(C) (i), (iii) and (iv)

(D) (i) and (iii)

Q.8 Consider a square sheet of side 1 unit. In the first step, it is cut along the
main diagonal to get two triangles. In the next step, one of the cut triangles
is revolved about its short edge to form a solid cone. The volume of the
resulting cone, in cubic units, is ________

(A) 𝜋
3
(B) 2𝜋
3
(C) 3𝜋
2
(D) 3π

ST - Copyright © GATE 2021 Page 5 of 27

 Statistics (ST)

Q.9
65
Sunday 55

50
Saturday 60

35
Friday 20

55
Thursday 60

50
Wednesday 60

65
Tuesday 55

70
Monday 45

0 10 20 30 40 50 60 70 80
Y X

The number of minutes spent by two students, X and Y, exercising every day
in a given week are shown in the bar chart above.
The number of days in the given week in which one of the students spent a
minimum of 10% more than the other student, on a given day, is

(A) 4

(B) 5

(C) 6

(D) 7

ST - Copyright © GATE 2021 Page 6 of 27

 Statistics (ST)

Q.10

Corners are cut from an equilateral triangle to produce a regular convex

hexagon as shown in the figure above.
The ratio of the area of the regular convex hexagon to the area of the original
equilateral triangle is

(A) 2:3

(B) 3:4

(C) 4:5

(D) 5:6

ST - Copyright © GATE 2021 Page 7 of 27

 Statistics (ST)

Statistics (ST)

Q.1 – Q.9 Multiple Choice Question (MCQ), carry ONE mark each (for each wrong
answer: – 1/3).

Q.1 Let 𝑿 be a non-constant positive random variable such that 𝑬(𝑿) = 𝟗.

Then which one of the following statements is true?

(A) 𝐸 ( 1 ) > 0.1 and 𝑃(𝑋 ≥ 10) ≤ 0.9

𝑋+1

(B) 𝐸 ( 1 ) < 0.1 and 𝑃(𝑋 ≥ 10) ≤ 0.9

𝑋+1

(C ) 𝐸 ( 1 ) > 0.1 and 𝑃(𝑋 ≥ 10) > 0.9

𝑋+1

(D) 𝐸 ( 1 ) < 0.1 and 𝑃(𝑋 ≥ 10) > 0.9

𝑋+1

Q.2 Let {𝑾(𝒕)}𝒕≥𝟎 be a standard Brownian motion. Then the variance of

𝑾(𝟏)𝑾(𝟐) equals

(A) 1

(B) 2

(C) 3

(D) 4

ST - Copyright © GATE 2021 Page 8 of 27

 Statistics (ST)

Q.3 Let 𝑿𝟏 , 𝑿𝟐 , … , 𝑿𝒏 be a random sample of size 𝒏 (≥ 𝟐) from a distribution

having the probability density function
1 − 𝑥−𝜃
𝑓(𝑥; 𝜃) = { 𝜃 𝑒
𝜃 , 𝑥 > 𝜃,
0, otherwise,
where 𝜃 ∈ (0, ∞). Then the method of moments estimator of 𝜃 equals
𝑛
(A) 1
∑ 𝑋𝑖
2𝑛
𝑖=1

𝑛
(B) 2
∑ 𝑋𝑖
𝑛
𝑖=1

𝑛
(C) 1
∑ 𝑋𝑖
𝑛
𝑖=1

(D) 𝑛
∑𝑛𝑖=1 𝑋𝑖

Q.4 Let {𝒙𝟏 , 𝒙𝟐 , … , 𝒙𝒏 } be a realization of a random sample of size 𝒏 (≥ 𝟐)

from a 𝑵(𝝁, 𝝈𝟐 ) distribution, where −∞ < 𝝁 < ∞ and 𝝈 > 𝟎. Which of
the following statements is/are true?
P : 95% confidence interval of 𝜇 based on {𝑥1 , 𝑥2 , … , 𝑥𝑛 } is unique when
𝜎 is known.
Q : 95% confidence interval of 𝜇 based on {𝑥1 , 𝑥2 , … , 𝑥𝑛 } is NOT unique
when 𝜎 is unknown.

(A) P only

(B) Q only

(C) Both P and Q

(D) Neither P nor Q

ST - Copyright © GATE 2021 Page 9 of 27

 Statistics (ST)

Q.5 Let 𝑿𝟏 , 𝑿𝟐 , … , 𝑿𝒏 be a random sample of size 𝒏 (≥ 𝟐) from a 𝑵(𝟎, 𝝈𝟐 )

distribution. For a given 𝝈 > 𝟎, let 𝒇𝝈 denote the joint probability density
function of (𝑿𝟏 , 𝑿𝟐 , … , 𝑿𝒏 ) and 𝑺 = {𝒇𝝈 : 𝝈 > 𝟎}. Let 𝑻𝟏 = ∑𝒏𝒊=𝟏 𝑿𝟐𝒊 and
𝟏 𝟐
𝑻𝟐 = (𝒏 ∑𝒏𝒊=𝟏 𝑿𝒊 ) . For any positive integer 𝝂 and any 𝜶 ∈ (𝟎, 𝟏), let 𝝌𝟐𝝂,𝜶
denote the (𝟏 − 𝜶)-th quantile of the central chi-square distribution with 𝝂
degrees of freedom. Consider testing 𝑯𝟎 : 𝝈 = 𝟏 against 𝑯𝟏 : 𝝈 > 𝟏 at level
𝜶. Then which one of the following statements is true?
2
(A) 𝑆 has a monotone likelihood ratio in 𝑇1 and 𝐻0 is rejected if 𝑇1 > 𝜒𝑛,𝛼
2
(B) 𝑆 has a monotone likelihood ratio in 𝑇1 and 𝐻0 is rejected if 𝑇1 > 𝜒𝑛,1−𝛼
2
(C) 𝑆 has a monotone likelihood ratio in 𝑇2 and 𝐻0 is rejected if 𝑇2 > 𝜒𝑛,𝛼
2
(D) 𝑆 has a monotone likelihood ratio in 𝑇2 and 𝐻0 is rejected if 𝑇2 > 𝜒𝑛,1−𝛼

ST - Copyright © GATE 2021 Page 10 of 27

 Statistics (ST)

Q.6 Let 𝑿 and 𝒀 be two random variables such that 𝒑𝟏𝟏 + 𝒑𝟏𝟎 + 𝒑𝟎𝟏 + 𝒑𝟎𝟎 =
𝟏, where 𝒑𝒊𝒋 = 𝑷(𝑿 = 𝒊, 𝒀 = 𝒋), 𝒊, 𝒋 = 𝟎, 𝟏. Suppose that a realization of a
random sample of size 𝟔𝟎 from the joint distribution of (𝑿, 𝒀) gives
𝒏𝟏𝟏 = 𝟏𝟎, 𝒏𝟏𝟎 = 𝟐𝟎, 𝒏𝟎𝟏 = 𝟐𝟎 𝐚𝐧𝐝 𝒏𝟎𝟎 = 𝟏𝟎,
where 𝒏𝒊𝒋 denotes the frequency of (𝒊, 𝒋) for 𝒊, 𝒋 = 𝟎, 𝟏. If the chi-square
test of independence is used to test
𝑯𝟎 : 𝒑𝒊𝒋 = 𝒑𝒊⋅ 𝒑⋅𝒋 for 𝒊, 𝒋 = 𝟎, 𝟏 against 𝑯𝟏 : 𝒑𝒊𝒋 ≠ 𝒑𝒊⋅ 𝒑⋅𝒋 for at least one pair
(𝒊, 𝒋),
where 𝒑𝒊⋅ = 𝒑𝒊𝟎 + 𝒑𝒊𝟏 and 𝒑⋅𝒋 = 𝒑𝟎𝒋 + 𝒑𝟏𝒋 , then which one of the following
statements is true?

(A) Under 𝐻0 , the test statistic follows central chi-square distribution with one
20
degree of freedom and the observed value of the test statistic is 3

(B) Under 𝐻0 , the test statistic follows central chi-square distribution with three
20
degrees of freedom and the observed value of the test statistic is 3

(C) Under 𝐻0 , the test statistic follows central chi-square distribution with one
16
degree of freedom and the observed value of the test statistic is 3

(D) Under 𝐻0 , the test statistic follows central chi-square distribution with three
16
degrees of freedom and the observed value of the test statistic is 3

Q.7 Let the joint distribution of (𝑿, 𝒀) be bivariate normal with mean vector
𝟎 𝟏 𝝆
( ) and variance-covariance matrix ( ), where −𝟏 < 𝝆 < 𝟏. Let
𝟎 𝝆 𝟏
𝚽𝝆 (𝟎, 𝟎) = 𝑷(𝑿 ≤ 𝟎, 𝒀 ≤ 𝟎). Then the Kendall’s 𝝉 coefficient between 𝑿
and 𝒀 equals

(A) 4Φ𝜌 (0,0) − 1

(B) 4Φ𝜌 (0,0)

(C) 4Φ𝜌 (0,0) + 1

(D) Φ𝜌 (0,0)

ST - Copyright © GATE 2021 Page 11 of 27

 Statistics (ST)

Q.8 Consider the simple linear regression model

𝒀𝒊 = 𝜷𝟎 + 𝜷𝟏 𝒙𝒊 + 𝝐𝒊 , 𝒊 = 𝟏, 𝟐, … , 𝒏 (𝒏 ≥ 𝟑),
where 𝜷𝟎 and 𝜷𝟏 are unknown parameters and 𝝐𝒊 ’s are independent and
identically distributed random variables with mean zero and finite variance
𝝈𝟐 > 𝟎. Suppose that 𝜷 ̂𝟎 and 𝜷̂𝟏 are the ordinary least squares estimators
𝟏
of 𝜷𝟎 and 𝜷𝟏 , respectively. Define 𝒙̅ = ∑𝒏𝒊=𝟏 𝒙𝒊 , 𝑺𝟏 = ∑𝒏𝒊=𝟏(𝒙𝒊 − 𝒙
̅)𝟐 and
𝒏
𝑺𝟐 = ∑𝒏𝒊=𝟏 𝒚𝒊 (𝒙𝒊 − 𝒙
̅), where 𝒚𝒊 is the observed value of 𝒀𝒊 , 𝒊 = 𝟏, 𝟐, … , 𝒏.
̂𝟎 + 𝒄 is
Then for a real constant 𝒄, the variance of 𝜷

(A) 2
1 𝑥̅ 2
𝜎 ( + )
𝑛 𝑆2

(B) 2
1 𝑥̅ 2
𝜎 ( + )
𝑛 𝑆1

(C) 𝜎 2
𝑛
(D) 1 𝑥̅ 2
𝜎2 ( + ) + 𝑐2
𝑛 𝑆2

Q.9 Let 𝑿𝟏 , 𝑿𝟐 , 𝑿𝟑 , 𝒀𝟏 , 𝒀𝟐 , 𝒀𝟑 and 𝒀𝟒 be independent random vectors such that

𝑿𝒊 follows 𝑵𝟒 (𝟎, 𝚺𝟏 ) distribution for 𝒊 = 𝟏, 𝟐, 𝟑, and 𝒀𝒋 follows 𝑵𝟒 (𝟎, 𝚺𝟐 )
distribution for 𝒋 = 𝟏, 𝟐, 𝟑, 𝟒, where 𝚺𝟏 and 𝚺𝟐 are positive definite
matrices. Further, let
𝒁 = 𝚺𝟏− ½ 𝑿𝑿𝑻 𝚺𝟏− ½ + 𝚺𝟐− ½ 𝒀𝒀𝑻 𝚺𝟐− ½ ,
where 𝑿 = [𝑿𝟏 𝑿𝟐 𝑿𝟑 ] is a 𝟒 × 𝟑 matrix, 𝒀 = [𝒀𝟏 𝒀𝟐 𝒀𝟑 𝒀𝟒 ] is a 𝟒 × 𝟒
matrix and 𝑿𝑻 and 𝒀𝑻 denote transposes of 𝑿 and 𝒀, respectively. If
𝑾𝒎 (𝒏, 𝚺) denotes a Wishart distribution of order 𝒎 with 𝒏 degrees of
freedom and variance-covariance matrix 𝚺 and 𝑰𝒏 denotes the 𝒏 × 𝒏
identity matrix, then which one of the following statements is true?

(A) 𝑍 follows 𝑊4 (7, 𝐼4 ) distribution

(B) 𝑍 follows 𝑊4 (4, 𝐼4 ) distribution

(C) 𝑍 follows 𝑊7 (4, 𝐼7 ) distribution

(D) 𝑍 follows 𝑊7 (7, 𝐼7 ) distribution

ST - Copyright © GATE 2021 Page 12 of 27

 Statistics (ST)

Q.10 – Q.25 Numerical Answer Type (NAT), carry ONE mark each (no negative
marks).

Q.10 𝟏
𝒏 𝟏
𝐥𝐢𝐦 (𝟐𝒏 + 𝒏𝟐𝒏 𝐬𝐢𝐧𝟐 ) 𝟐𝒏−𝒏 𝐜𝐨𝐬𝒏
𝒏→∞ 𝟐
equals __________

Q.11 Let
𝟏 𝒙
𝟏
𝑰=𝟒 ∫ ∫ 𝒅𝒚 𝒅𝒙
𝟏 √𝟏−𝒙𝟐
√𝒙𝟐 + 𝒚𝟐
√𝟐
Then the value of 𝒆𝑰+𝝅 equals __________ (round off to 𝟐 decimal
places).

Q.12 𝟎 𝟎 𝟏
Let 𝑨 = [𝟏 𝟎 𝟎] and 𝑰𝟑 be the 𝟑 × 𝟑 identity matrix. Then the
𝟎 𝟏 𝟎
nullity of 𝟓𝑨(𝑰𝟑 + 𝑨 + 𝑨𝟐 ) equals __________

Q.13 Let 𝑨 be the 𝟐 × 𝟐 real matrix having eigenvalues 𝟏 𝐚𝐧𝐝 − 𝟏, with

√𝟑 −𝟏
𝟐 𝟐
corresponding eigenvectors [ ] and [ ], respectively. If 𝑨𝟐𝟎𝟐𝟏 =
𝟏 √𝟑
𝟐 𝟐
𝒂 𝒃
[ ], then 𝒂 + 𝒃 + 𝒄 + 𝒅 equals __________ (round off to 𝟐 decimal
𝒄 𝒅
places).

Q.14 𝟑 𝟏
Let 𝑨 and 𝑩 be two events such that 𝑷(𝑩) = and 𝑷(𝑨 ∪ 𝑩𝒄 ) = .
𝟒 𝟐
If 𝑨 and 𝑩 are independent, then 𝑷(𝑨) equals __________ (round off to
𝟐 decimal places).

ST - Copyright © GATE 2021 Page 13 of 27

 Statistics (ST)

Q.15 A fair die is rolled twice independently. Let 𝑿 and 𝒀 denote the outcomes
of the first and second roll, respectively. Then 𝑬(𝑿 + 𝒀 | (𝑿 − 𝒀)𝟐 = 𝟏)
equals __________

Q.16 Let 𝑿 be a random variable having distribution function

𝟎, 𝒙 < 𝟏,
𝒂
, 𝟏 ≤ 𝒙 < 𝟐,
𝑭(𝒙) = 𝟐𝒄
, 𝟐 ≤ 𝒙 < 𝟑,
𝟔
{ 𝟏, 𝒙 ≥ 𝟑,
𝟏 𝟏
where 𝒂 and 𝒄 are appropriate constants. Let 𝑨𝒏 = [𝟏 + 𝒏 , 𝟑 − 𝒏], 𝒏 ≥
𝟏 𝟓
𝟏, and 𝑨 = ⋃∞
𝒊=𝟏 𝑨𝒊 . If 𝑷(𝑿 ≤ 𝟏) = 𝟐 and 𝑬(𝑿) = 𝟑, then 𝑷(𝑿 ∈ 𝑨)
equals __________ (round off to 𝟐 decimal places).

Q.17 If the marginal probability density function of the 𝒌𝒕𝒉 order statistic of a
random sample of size 8 from a uniform distribution on [𝟎, 𝟐] is
𝟕 𝟔
(𝟐 − 𝒙), 𝟎 < 𝒙 < 𝟐,
𝒇(𝒙) = { 𝟑𝟐 𝒙
𝟎, 𝐨𝐭𝐡𝐞𝐫𝐰𝐢𝐬𝐞,
then 𝒌 equals __________

Q.18 (𝜶)
For 𝜶 > 𝟎, let {𝑿𝒏 } be a sequence of independent random variables
𝒏≥𝟏
such that
(𝜶) 𝟏 (𝜶)
𝑷(𝑿𝒏 = 𝟏) = = 𝟏 − 𝑷(𝑿𝒏 = 𝟎).
𝒏𝟐𝜶
(𝜶)
Let 𝑺 = {𝜶 > 𝟎 ∶ 𝑿𝒏 𝐜𝐨𝐧𝐯𝐞𝐫𝐠𝐞𝐬 𝐭𝐨 𝟎 𝐚𝐥𝐦𝐨𝐬𝐭 𝐬𝐮𝐫𝐞𝐥𝐲 𝐚𝐬 𝒏 → ∞}. Then
the infimum of 𝑺 equals __________ (round off to 𝟐 decimal places).

ST - Copyright © GATE 2021 Page 14 of 27

 Statistics (ST)

Q.19 Let {𝑿𝒏 }𝒏≥𝟏 be a sequence of independent and identically distributed

random variables each having uniform distribution on [𝟎, 𝟐]. For 𝒏 ≥ 𝟏,
let
𝟏
𝒏 𝒏
𝒁𝒏 = − 𝐥𝐨𝐠 𝒆 (∏(𝟐 − 𝑿𝒊 )) .
𝒊=𝟏

Then, as 𝒏 → ∞, the sequence {𝒁𝒏 }𝒏≥𝟏 converges almost surely to

_________ (round off to 𝟐 decimal places).

Q.20 Let {𝑿𝒏 }𝒏≥𝟎 be a time-homogeneous discrete time Markov chain with state
space {𝟎, 𝟏} and transition probability matrix
𝟎. 𝟐𝟓 𝟎. 𝟕𝟓
[ ].
𝟎. 𝟕𝟓 𝟎. 𝟐𝟓
If 𝑷(𝑿𝟎 = 𝟎) = 𝑷(𝑿𝟎 = 𝟏) = 𝟎. 𝟓, then
𝟏𝟎𝟎

∑ 𝑬[(𝑿𝟐𝒌 )𝟐𝒌 ]
𝒌=𝟏
equals __________

Q.21 Let {𝟎, 𝟐} be a realization of a random sample of size 𝟐 from a binomial

𝟏
distribution with parameters 𝟐 𝐚𝐧𝐝 𝒑, where 𝒑 ∈ (𝟎, 𝟏). To test 𝑯𝟎 : 𝒑 = 𝟐
𝟏
against 𝑯𝟏 : 𝒑 ≠ 𝟐, the observed value of the likelihood ratio test statistic
equals __________ (round off to 𝟐 decimal places).

Q.22 Let 𝑿 be a random variable having the probability density function

𝟑
(𝟏
𝒇(𝒙) = { 𝟏𝟑 − 𝒙)(𝟗 − 𝒙), 𝟎 < 𝒙 < 𝟏,
𝟎, 𝐨𝐭𝐡𝐞𝐫𝐰𝐢𝐬𝐞.
𝟒
Then 𝑬[𝑿(𝑿𝟐 − 𝟏𝟓𝑿 + 𝟐𝟕)] equals __________ (round off to 𝟐
𝟑
decimal places).

ST - Copyright © GATE 2021 Page 15 of 27

 Statistics (ST)

Q.23 𝟓
Let (𝒀, 𝑿𝟏 , 𝑿𝟐 ) be a random vector with mean vector (𝟐) and variance-
𝟎
𝟏𝟎 𝟎. 𝟓 −𝟎. 𝟓
covariance matrix [ 𝟎. 𝟓 𝟕 𝟏. 𝟓 ]. Then the value of the multiple
−𝟎. 𝟓 𝟏. 𝟓 𝟐
correlation coefficient between 𝒀 and its best linear predictor on 𝑿𝟏 and
𝑿𝟐 equals __________ (round off to 𝟐 decimal places).

Q.24 Let 𝑿𝟏 , 𝑿𝟐 and 𝑿𝟑 be a random sample from a bivariate normal

distribution with unknown mean vector 𝝁 and unknown variance-
covariance matrix 𝚺, which is a positive definite matrix. The 𝒑-value
corresponding to the likelihood ratio test for testing 𝑯𝟎 : 𝝁 = 𝟎 against
𝟏 𝟒 −𝟓
𝑯𝟏 : 𝝁 ≠ 𝟎 based on the realization {( ) , ( ) , ( )} of the random
𝟐 −𝟐 𝟎
sample equals __________ (round off to 𝟐 decimal places).

Q.25 Let 𝒀𝒊 = 𝜶 + 𝜷𝒙𝒊 + 𝝐𝒊 , 𝒊 = 𝟏, 𝟐, 𝟑, where 𝒙𝒊 ’s are fixed covariates, 𝜶 and

𝜷 are unknown parameters and 𝝐𝒊 ’s are independent and identically
distributed random variables with mean zero and finite variance. Let 𝜶 ̂ and
̂ be the ordinary least squares estimators of 𝜶 and 𝜷, respectively. Given
𝜷
the following observations

𝒚𝒊 𝟖. 𝟔𝟐 𝟐𝟔. 𝟖𝟔 𝟓𝟒. 𝟎𝟐

𝒙𝒊 𝟑. 𝟐𝟗 𝟐𝟏. 𝟓𝟑 𝟒𝟖. 𝟔𝟗

̂ equals __________ (round off to 𝟐 decimal places).

̂+𝜷
the value of 𝜶

ST - Copyright © GATE 2021 Page 16 of 27

 Statistics (ST)

Q.26 – Q.43 Multiple Choice Question (MCQ), carry TWO mark each (for each wrong
answer: – 2/3).

Q.26 Let 𝒇: ℝ → ℝ be defined by

𝒙𝟑 𝐬𝐢𝐧 𝒙, 𝒙 = 𝟎 𝐨𝐫 𝒙 𝐢𝐬 𝐢𝐫𝐫𝐚𝐭𝐢𝐨𝐧𝐚𝐥,
𝒇(𝒙) = { 𝟏 𝒑
𝟑
, 𝒙 = , 𝒑 ∈ ℤ ∖ {𝟎}, 𝒒 ∈ ℕ 𝐚𝐧𝐝 𝐠𝐜𝐝(𝒑, 𝒒) = 𝟏,
𝒒 𝒒
where ℝ denotes the set of all real numbers, ℤ denotes the set of all integers,
ℕ denotes the set of all positive integers and 𝐠𝐜𝐝(𝒑, 𝒒) denotes the greatest
common divisor of 𝒑 and 𝒒. Then which one of the following statements is
true?

(A) 𝑓 is not continuous at 0

(B) 𝑓 is not differentiable at 0

(C) 𝑓 is differentiable at 0 and the derivative of 𝑓 at 0 equals 0

(D) 𝑓 is differentiable at 0 and the derivative of 𝑓 at 0 equals 1

Q.27 Let 𝒇: [𝟎, ∞) → ℝ be a function, where ℝ denotes the set of all real
numbers. Then which one of the following statements is true?

(A) If 𝑓 is bounded and continuous, then 𝑓 is uniformly continuous

(B) If 𝑓 is uniformly continuous, then lim 𝑓(𝑥) exists

𝑥→∞

(C) If 𝑓 is uniformly continuous, then the function 𝑔(𝑥) = 𝑓(𝑥) sin 𝑥 is also
uniformly continuous

(D) If 𝑓 is continuous and lim 𝑓(𝑥) is finite, then 𝑓 is uniformly continuous

𝑥→∞

ST - Copyright © GATE 2021 Page 17 of 27

 Statistics (ST)

Q.28 Let 𝒇: ℝ → ℝ be a differentiable function such that 𝒇(𝟎) = 𝟎 and

𝒇′ (𝒙) + 𝟐𝒇(𝒙) > 𝟎 for all 𝒙 ∈ ℝ, where 𝒇′ denotes the derivative of 𝒇 and
ℝ denotes the set of all real numbers. Then which one of the following
statements is true?

(A) 𝑓(𝑥) > 0, for all 𝑥 > 0 and 𝑓(𝑥) < 0, for all 𝑥 < 0

(B) 𝑓(𝑥) < 0, for all 𝑥 ≠ 0

(C) 𝑓(𝑥) > 0, for all 𝑥 ≠ 0

(D) 𝑓(𝑥) < 0, for all 𝑥 > 0 and 𝑓(𝑥) > 0, for all 𝑥 < 0

Q.29 Let 𝑴 be the collection of all 𝟑 × 𝟑 real symmetric positive definite

matrices. Consider the set
𝟏
𝑺 = {𝑨 ∈ 𝑴 ∶ 𝑨𝟓𝟎 − 𝑨𝟒𝟖 = 𝟎},
𝟒
where 𝟎 denotes the 𝟑 × 𝟑 zero matrix. Then the number of elements in
𝑺 equals

(A) 0

(B) 1

(C) 8

(D) ∞

Q.30 Let 𝑨 be a 𝟑 × 𝟑 real matrix such that 𝑰𝟑 + 𝑨 is invertible and let

𝑩 = (𝑰𝟑 + 𝑨)−𝟏 (𝑰𝟑 − 𝑨), where 𝑰𝟑 denotes the 𝟑 × 𝟑 identity matrix. Then
which one of the following statements is true?

(A) If 𝐵 is orthogonal, then 𝐴 is invertible

(B) If 𝐵 is orthogonal, then all the eigenvalues of 𝐴 are real

(C) If 𝐵 is skew-symmetric, then 𝐴 is orthogonal

(D) If 𝐵 is skew-symmetric, then the determinant of 𝐴 equals −1

ST - Copyright © GATE 2021 Page 18 of 27

 Statistics (ST)

Q.31 Let 𝑿 be a random variable having Poisson distribution such that

𝑬(𝑿𝟐 ) = 𝟏𝟏𝟎. Then which one of the following statements is NOT true?

(A) 𝐸(𝑋 𝑛 ) = 10 𝐸[(𝑋 + 1)𝑛−1 ], for all 𝑛 = 1, 2, 3, …

(B) 1
𝑃(𝑋 is even) = 4 (1 + 𝑒 −20 )

(C) 𝑃(𝑋 = 𝑘) < 𝑃(𝑋 = 𝑘 + 1), for 𝑘 = 0, 1, … , 8

(D) 𝑃(𝑋 = 𝑘) > 𝑃(𝑋 = 𝑘 + 1), for 𝑘 = 10, 11, …

Q.32 𝝅 𝝅
Let 𝑿 be a random variable having uniform distribution on [− , 𝟐 ].
𝟐
Then which one of the following statements is NOT true?

(A) 𝑌 = cot 𝑋 follows standard Cauchy distribution

(B) 𝑌 = tan 𝑋 follows standard Cauchy distribution

(C) 𝑌 = − log ( 1 + 𝑋) has moment generating function 𝑀 (𝑡 ) = 1

𝑒 2 𝜋
, 𝑡<1
1−𝑡

(D) 𝑌 = −2 log ( 1 + 𝑋) follows central chi-square distribution with one degree of

𝑒 2 𝜋
freedom

Q.33 Let 𝛀 = {𝟏, 𝟐, 𝟑, … } represent the collection of all possible outcomes of a

random experiment with probabilities 𝑷({𝒏}) = 𝜶𝒏 for 𝒏 ∈ 𝛀. Then which
one of the following statements is NOT true?

(A) lim 𝛼𝑛 = 0
𝑛→∞

(B) ∑∞
𝑛=1 √𝛼𝑛 converges

(C) For any positive integer 𝑘, there exist 𝑘 disjoint events 𝐴1 , 𝐴2 , … , 𝐴𝑘 such
that 𝑃(⋃𝑘𝑖=1 𝐴𝑖 ) < 0.001

(D) There exists a sequence {𝐴𝑖 }𝑖≥1 of strictly increasing events such that
𝑃(⋃∞𝑖=1 𝐴𝑖 ) < 0.001

ST - Copyright © GATE 2021 Page 19 of 27

 Statistics (ST)

Q.34 Let (𝑿, 𝒀) have the joint probability density function

𝟒
, 𝒙 > 𝟏, 𝒚 > 𝟏,
𝒇𝑿,𝒀 (𝒙, 𝒚) = {(𝒙 + 𝒚)𝟑
𝟎, 𝐨𝐭𝐡𝐞𝐫𝐰𝐢𝐬𝐞.
Then which one of the following statements is NOT true?

(A) The probability density function of 𝑋 + 𝑌 is

4
(𝑧
𝑓𝑋+𝑌 (𝑧) = { 𝑧 3 − 2), 𝑧 > 2,
0, otherwise.

(B) 3
𝑃(𝑋 + 𝑌 > 4) =
4
(C) 𝐸(𝑋 + 𝑌) = 4 log 𝑒 2

(D) 𝐸(𝑌 | 𝑋 = 2) = 4

Q.35 Let 𝑿𝟏 , 𝑿𝟐 and 𝑿𝟑 be three uncorrelated random variables with common

variance 𝝈𝟐 < ∞. Let 𝒀𝟏 = 𝟐𝑿𝟏 + 𝑿𝟐 + 𝑿𝟑 , 𝒀𝟐 = 𝑿𝟏 + 𝟐𝑿𝟐 + 𝑿𝟑 and
𝒀𝟑 = 𝑿𝟏 + 𝑿𝟐 + 𝟐𝑿𝟑 . Then which of the following statements is/are true?
P : The sum of eigenvalues of the variance covariance matrix of
(𝒀𝟏 , 𝒀𝟐 , 𝒀𝟑 )
is 𝟏𝟖𝝈𝟐 .
Q : The correlation coefficient between 𝒀𝟏 𝐚𝐧𝐝 𝒀𝟐 equals that between
𝒀𝟐 𝐚𝐧𝐝 𝒀𝟑 .

(A) P only

(B) Q only

(C) Both P and Q

(D) Neither P nor Q

ST - Copyright © GATE 2021 Page 20 of 27

 Statistics (ST)

Q.36 Let {𝑿𝒏 }𝒏≥𝟎 be a time-homogeneous discrete time Markov chain with either
finite or countable state space 𝑺. Then which one of the following statements
is true?

(A) There is at least one recurrent state

(B) If there is an absorbing state, then there exists at least one stationary distribution

(C) If all the states are positive recurrent, then there exists a unique stationary
distribution

(D) If {𝑋𝑛 }𝑛≥0 is irreducible, 𝑆 = {1, 2} and [𝜋1 𝜋2 ] is a stationary distribution,

then lim 𝑃(𝑋𝑛 = 𝑖 | 𝑋0 = 𝑖) = 𝜋𝑖 for 𝑖 = 1, 2
𝑛→∞

Q.37 Let customers arrive at a departmental store according to a Poisson process

with rate 𝟏𝟎. Further, suppose that each arriving customer is either a male
𝟏
or a female with probability 𝟐 each, independent of all other arrivals. Let
𝑵(𝒕) denote the total number of customers who have arrived by time 𝒕.
Then which one of the following statements is NOT true?

(A) If 𝑆2 denotes the time of arrival of the second female customer, then
1
𝑃(𝑆2 ≤ 1) = 25 ∫0 𝑠𝑒 −5𝑠 𝑑𝑠

(B) If 𝑀(𝑡) denotes the number of male customers who have arrived by time 𝑡,
1 1
then 𝑃 (𝑀 (3) = 0 | 𝑀(1) = 1) = 3

(C) 𝐸 [(𝑁(𝑡))2 ] = 100𝑡 2 + 10𝑡

(D) 𝐸[𝑁(𝑡)𝑁(2𝑡)] = 200𝑡 2 + 10𝑡

ST - Copyright © GATE 2021 Page 21 of 27

 Statistics (ST)

Q.38 Let 𝑿(𝟏) < 𝑿(𝟐) < 𝑿(𝟑) < 𝑿(𝟒) < 𝑿(𝟓) be the order statistics corresponding
to a random sample of size 𝟓 from a uniform distribution on [𝟎, 𝜽], where
𝜽 ∈ (𝟎, ∞). Then which of the following statements is/are true?
P : 𝟑𝑿(𝟐) is an unbiased estimator of 𝜽.
Q : The variance of 𝑬[𝟐𝑿(𝟑) | 𝑿(𝟓) ] is less than or equal to the variance
of 𝟐𝑿(𝟑) .

(A) P only

(B) Q only

(C) Both P and Q

(D) Neither P nor Q

Q.39 Let 𝑿𝟏 , 𝑿𝟐 , … , 𝑿𝒏 be a random sample of size 𝒏 (≥ 𝟐) from a

distribution having the probability density function
𝟏 −𝒙
𝒇(𝒙; 𝜽) = {𝜽 𝒆 , 𝒙 > 𝟎,
𝜽

𝟎, 𝐨𝐭𝐡𝐞𝐫𝐰𝐢𝐬𝐞,
where 𝜽 ∈ (𝟎, ∞). Let 𝑿(𝟏) = 𝐦𝐢𝐧{ 𝑿𝟏 , 𝑿𝟐 , … , 𝑿𝒏 } and 𝑻 = ∑𝒏𝒊=𝟏 𝑿𝒊 .
Then 𝑬(𝑿(𝟏) | 𝑻 ) equals

(A) 𝑇
𝑛2
(B) 𝑇
𝑛
(C) (𝑛 + 1)𝑇
2𝑛
(D) (𝑛 + 1)2 𝑇
4𝑛2

ST - Copyright © GATE 2021 Page 22 of 27

 Statistics (ST)

Q.40 Let 𝑿𝟏 , 𝑿𝟐 , … , 𝑿𝒏 be a random sample of size 𝒏 (≥ 𝟐) from a uniform

distribution on [−𝜽, 𝜽], where 𝜽 ∈ (𝟎, ∞). Let 𝑿(𝟏) = 𝐦𝐢𝐧{ 𝑿𝟏 , 𝑿𝟐 , … , 𝑿𝒏 }
and 𝑿(𝒏) = 𝐦𝐚𝐱{ 𝑿𝟏 , 𝑿𝟐 , … , 𝑿𝒏 }. Then which of the following statements
is/are true?
P : (𝑿(𝟏), 𝑿(𝒏) ) is a complete statistic.
Q : 𝑿(𝒏) − 𝑿(𝟏) is an ancillary statistic.

(A) P only

(B) Q only

(C) Both P and Q

(D) Neither P nor Q

Q.41 Let {𝑿𝒏 }𝒏≥𝟏 be a sequence of independent and identically distributed

random variables having common distribution function 𝑭(⋅). Let 𝒂 < 𝒃 be
two real numbers such that 𝑭(𝒙) = 𝟎 for all 𝒙 ≤ 𝒂, 𝟎 < 𝑭(𝒙) < 𝟏 for all
𝒂 < 𝒙 < 𝒃 and 𝑭(𝒙) = 𝟏 for all 𝒙 ≥ 𝒃. Let 𝑺𝒏 (𝒙) be the empirical
distribution function at 𝒙 based on 𝑿𝟏 , 𝑿𝟐 , … , 𝑿𝒏 , 𝒏 ≥ 𝟏. Then which one
of the following statements is NOT true?

(A) 𝑃 [ lim sup |𝑆𝑛 (𝑥) − 𝐹(𝑥)| = 0] = 1

𝑛→∞ −∞<𝑥<∞

(B) For fixed 𝑥 ∈ (𝑎, 𝑏) and 𝑡 ∈ (− ∞, ∞),

√𝑛 |𝑆𝑛 (𝑥) − 𝐹(𝑥)|

lim 𝑃 ≤ 𝑡 = 𝑃(𝑍 ≤ 𝑡),
𝑛→∞
√𝑆 (𝑥)(1 − 𝑆𝑛 (𝑥))
[ 𝑛 ]
where 𝑍 is the standard normal random variable

(C) The covariance between 𝑆𝑛 (𝑥) and 𝑆𝑛 (𝑦) equals 1

𝐹(𝑥)(1 − 𝐹(𝑦)) for all
𝑛
𝑛 ≥ 2 and for fixed −∞ < 𝑥, 𝑦 < ∞

(D) If 𝑌 = 2
𝑛 sup (𝑆𝑛 (𝑥) − 𝐹(𝑥)) , then {4𝑛 𝑌𝑛 }𝑛≥1 converges in distribution to
−∞<𝑥<∞
a central chi-square random variable with 2 degrees of freedom

ST - Copyright © GATE 2021 Page 23 of 27

 Statistics (ST)

Q.42 Let the joint distribution of random variables 𝑿𝟏 , 𝑿𝟐 , 𝑿𝟑 and 𝑿𝟒 be

𝑵𝟒 ( 𝝁, 𝚺), where
𝟏 𝟏 𝟎. 𝟐 𝟎 𝟎
𝟎 𝟎. 𝟐 𝟐 𝟎 𝟎
𝝁=( ) 𝐚𝐧𝐝 𝚺=[ ].
𝟎 𝟎 𝟎 𝟐 𝟎. 𝟐
𝟏 𝟎 𝟎 𝟎. 𝟐 𝟏
Then which one of the following statements is true?

(A) 5
[(𝑋1 + 𝑋2 )2 + (𝑋3 + 𝑋4 − 1)2 ] follows a central chi-square
17
distribution with 2 degrees of freedom

(B) 1 [(𝑋 + 𝑋 − 1)2 + (𝑋 + 𝑋 − 1)2 ] follows a central chi-square

3 1 3 2 4
distribution with 2 degrees of freedom

(C) 𝑋1 +𝑋2 −1
𝐸 [√| | ] is NOT finite
𝑋3 +𝑋4 −1

(D) 𝑋1 +𝑋2 +𝑋3 +𝑋4 −2

𝐸[| 𝑋1 +𝑋2 −𝑋3 −𝑋4
| ] is NOT finite

Q.43 Let 𝒀 follow 𝑵𝟖 (𝟎, 𝑰𝟖 ) distribution, where 𝑰𝟖 is the 𝟖 × 𝟖 identity matrix.

Let 𝒀𝑻 𝚺𝟏 𝒀 and 𝒀𝑻 𝚺𝟐 𝒀 be independent and follow central chi-square
distributions with 𝟑 and 𝟒 degrees of freedom, respectively, where 𝚺𝟏 and
𝚺𝟐 are 𝟖 × 𝟖 matrices and 𝒀𝑻 denotes transpose of 𝒀. Then which of the
following statements is/are true?
P : 𝚺𝟏 and 𝚺𝟐 are idempotent.
Q : 𝚺𝟏 𝚺𝟐 = 𝟎, where 𝟎 is the 𝟖 × 𝟖 zero matrix.

(A) P only

(B) Q only

(C) Both P and Q

(D) Neither P nor Q

ST - Copyright © GATE 2021 Page 24 of 27

 Statistics (ST)

Q.44 – Q.55 Numerical Answer Type (NAT), carry TWO mark each (no negative marks).

Q.44 Let (𝑿, 𝒀) have a bivariate normal distribution with the joint probability
density function
𝟏 ( 𝟑 𝒙𝒚 − 𝟐𝟓 𝒙𝟐 − 𝟐 𝒚𝟐)
𝒇𝑿,𝒀 (𝒙, 𝒚) = 𝒆 𝟐 𝟑𝟐 , − ∞ < 𝒙, 𝒚 < ∞.
𝝅
Then 𝟖 𝑬(𝑿𝒀) equals __________

Q.45 Let 𝒇: ℝ × ℝ → ℝ be defined by 𝒇(𝒙, 𝒚) = 𝟖𝒙𝟐 − 𝟐𝒚, where ℝ denotes the

set of all real numbers. If 𝑴 and 𝒎 denote the maximum and minimum
values of 𝒇, respectively, on the set {(𝒙, 𝒚) ∈ ℝ × ℝ ∶ 𝒙𝟐 + 𝒚𝟐 = 𝟏}, then
𝑴 − 𝒎 equals __________ (round off to 𝟐 decimal places).

Q.46 Let 𝑨 = [𝒂 𝒖𝟏 𝒖𝟐 𝒖𝟑 ], 𝑩 = [𝒃 𝒖𝟏 𝒖𝟐 𝒖𝟑 ] and 𝑪 = [𝒖𝟐 𝒖𝟑 𝒖𝟏 𝒂 + 𝒃] be

three 𝟒 × 𝟒 real matrices, where 𝒂, 𝒃, 𝒖𝟏 , 𝒖𝟐 𝐚𝐧𝐝 𝒖𝟑 are 𝟒 × 𝟏 real column
vectors. Let 𝐝𝐞𝐭(𝑨), 𝐝𝐞𝐭(𝑩) and 𝐝𝐞𝐭(𝑪) denote the determinants of the
matrices 𝑨, 𝑩 and 𝑪, respectively. If 𝐝𝐞𝐭(𝑨) = 𝟔 and 𝐝𝐞𝐭(𝑩) = 𝟐, then
𝐝𝐞𝐭(𝑨 + 𝑩) − 𝐝𝐞𝐭(𝑪) equals __________

Q.47 Let 𝑿 be a random variable having the moment generating function

𝒆𝒕 − 𝟏
𝑴(𝒕) = , 𝒕 < 𝟏.
𝒕(𝟏 − 𝒕)
Then 𝑷(𝑿 > 𝟏) equals __________ (round off to 𝟐 decimal places).

Q.48 Let {𝑿𝒏 }𝒏≥𝟏 be a sequence of independent and identically distributed

random variables each having uniform distribution on [𝟎, 𝟑]. Let 𝒀 be a
random variable, independent of {𝑿𝒏 }𝒏≥𝟏 , having probability mass function
𝟏
, 𝒌 = 𝟏, 𝟐, … ,
𝑷(𝒀 = 𝒌) = {(𝒆 − 𝟏)𝒌!
𝟎, 𝐨𝐭𝐡𝐞𝐫𝐰𝐢𝐬𝐞.
Then 𝑷(𝐦𝐚𝐱{𝑿𝟏 , 𝑿𝟐 , … , 𝑿𝒀 } ≤ 𝟏) equals __________ (round off to 𝟐
decimal places).

ST - Copyright © GATE 2021 Page 25 of 27

 Statistics (ST)

Q.49 Let {𝑿𝒏 }𝒏≥𝟏 be a sequence of independent and identically distributed

random variables each having probability density function
𝒆−𝒙 , 𝒙 > 𝟎,
𝒇(𝒙) = {
𝟎, 𝐨𝐭𝐡𝐞𝐫𝐰𝐢𝐬𝐞.
Let 𝑿(𝒏) = 𝐦𝐚𝐱{𝑿𝟏 , 𝑿𝟐 , … , 𝑿𝒏 } for 𝒏 ≥ 𝟏. If 𝒁 is the random variable to
which {𝑿(𝒏) − 𝐥𝐨𝐠 𝒆 𝒏}𝒏≥𝟏 converges in distribution, as 𝒏 → ∞, then the
median of
𝒁 equals __________ (round off to 𝟐 decimal places).

Q.50 Consider an amusement park where visitors are arriving according to a

Poisson process with rate 𝟏. Upon arrival, a visitor spends a random amount
of time in the park and then departs. The time spent by the visitors are
independent of one another, as well as of the arrival process, and have
common probability density function
𝒆−𝒙 , 𝒙 > 𝟎,
𝒇(𝒙) = {
𝟎, 𝐨𝐭𝐡𝐞𝐫𝐰𝐢𝐬𝐞.
If at a given time point, there are 𝟏𝟎 visitors in the park and 𝒑 is the
probability that there will be exactly two more arrivals before the next
departure, then
𝟏
equals __________
𝒑

Q.51 Let {𝟎. 𝟗𝟎, 𝟎. 𝟓𝟎, 𝟎. 𝟎𝟏, 𝟎. 𝟗𝟓} be a realization of a random sample of size 4
from the probability density function
𝜽 (𝟐𝜽−𝟏)⁄(𝟏−𝜽)
𝒇(𝒙) = { 𝟏 − 𝜽 𝒙 , 𝟎 < 𝒙 < 𝟏,
𝟎, 𝐨𝐭𝐡𝐞𝐫𝐰𝐢𝐬𝐞,
where 𝟎. 𝟓 ≤ 𝜽 < 𝟏. Then the maximum likelihood estimate of 𝜽 based on
the observed sample equals __________ (round off to 𝟐 decimal places).

Q.52 Let a random sample of size 𝟏𝟎𝟎 from a normal population with unknown
mean 𝝁 and variance 𝟗 give the sample mean 𝟓. 𝟔𝟎𝟖. Let 𝚽(⋅) denote
the distribution function of the standard normal random variable.
If 𝚽(𝟏. 𝟗𝟔) = 𝟎. 𝟗𝟕𝟓, 𝚽(𝟏. 𝟔𝟒) = 𝟎. 𝟗𝟓 and the uniformly most powerful
unbiased test based on sample mean is used to test 𝑯𝟎 : 𝝁 = 𝟓. 𝟎𝟐 against
𝑯𝟏 : 𝝁 ≠ 𝟓. 𝟎𝟐, then the 𝒑-value equals __________ (round off to 𝟑 decimal
places).

ST - Copyright © GATE 2021 Page 26 of 27

 Statistics (ST)

Q.53 Let 𝑿 be a discrete random variable with probability mass function 𝒑 ∈

{𝒑𝟎 , 𝒑𝟏 }, where

𝒙 𝟕 𝟖 𝟗 𝟏𝟎

𝒑𝟏 (𝒙) 𝟎. 𝟔𝟗 𝟎. 𝟏𝟎 𝟎. 𝟏𝟔 𝟎. 𝟎𝟓

𝒑𝟎 (𝒙) 𝟎. 𝟗𝟎 𝟎. 𝟎𝟓 𝟎. 𝟎𝟒 𝟎. 𝟎𝟏

To test 𝑯𝟎 : 𝒑 = 𝒑𝟎 against 𝑯𝟏 : 𝒑 = 𝒑𝟏 , the power of the most powerful test

of size 𝟎. 𝟎𝟓, based on 𝑿, equals __________ (round off to 𝟐 decimal
places).

Q.54 Let 𝑿𝟏 , 𝑿𝟐 , … , 𝑿𝟏𝟎 be a random sample from a probability density function

𝒇𝜽 (𝒙) = 𝒇(𝒙 − 𝜽), −∞ < 𝒙 < ∞,
where −∞ < 𝜽 < ∞ and 𝒇(−𝒙) = 𝒇(𝒙) for −∞ < 𝒙 < ∞. For testing
𝑯𝟎 : 𝜽 = 𝟏. 𝟐 against 𝑯𝟏 : 𝜽 ≠ 𝟏. 𝟐, let 𝑻+ denote the Wilcoxson Signed-
rank test statistic. If 𝜼 denotes the probability of the event {𝑻+ < 𝟓𝟎} under
𝑯𝟎 , then 𝟑𝟐 𝜼 equals __________ (round off to 2 decimal places).

Q.55 Consider the multiple linear regression model

𝒀𝒊 = 𝜷𝟎 + 𝜷𝟏 𝒙𝟏,𝒊 + 𝜷𝟐 𝒙𝟐,𝒊 + ⋯ + 𝜷𝟐𝟐 𝒙𝟐𝟐,𝒊 + 𝝐𝒊 , 𝒊 = 𝟏, 𝟐, … , 𝟏𝟐𝟑,
where, for 𝒋 = 𝟎, 𝟏, 𝟐, … , 𝟐𝟐, 𝜷𝒋 ’s are unknown parameters and 𝝐𝒊 ’s are
independent and identically distributed 𝑵(𝟎, 𝝈𝟐 ), 𝝈 > 𝟎, random
variables.
If the sum of squares due to regression is 𝟑𝟑𝟖. 𝟗𝟐, the total sum of squares
is 𝟓𝟐𝟐. 𝟑𝟎 and 𝑹𝟐𝒂𝒅𝒋 denotes the value of adjusted 𝑹𝟐 , then 𝟏𝟎𝟎 𝑹𝟐𝒂𝒅𝒋
equals __________ (round off to 𝟐 decimal places).

END OF THE QUESTION PAPER

ST - Copyright © GATE 2021 Page 27 of 27