DFSE-T-STT

'-
STATISTICS

PAPER-Ill

I Time Allowed : Three Hours I Maximum Marks 200 I

QUESTION PAPER SPECIFIC INSTRUCTIONS

Please read each of the following instructions carefully

before attempting questions

There are EIGHT questions divided under TWO Sections.

Candidate has to attempt FIVE questions in all.

Both the questions in Section-A are compulsory.

Out of the SIX questions in Section-B, any THREE questions are to be attempted.

The number of marks carried by a question/part is indicated against it.

Unless otherwise mentioned, symbols and notations have their usual standard meanings.

Assume suitable data, if necessary, and indicate the same clearly.

Attempts of questions shall be counted in sequential order. Unless struck off, attempt of a
question shall be counted even if attempted partly.

Any page or portion of the page left blank in the Question-cum-Answer (QCA) Booklet
must be clearly struck off.

Answers must be written in ENGLISH only.

DFSE-T-STT/5 1 [P.T.O.

Examrace www.examrace.com
 SECTION-A

Both the questions are compulsory

1. (a) Describe the advantages of sampling versus complete enumeration. Write the
circumstances under which complete enumeration is preferred to sampling. 10

(b) In the regression model Yi = a + Pxi + ui, i = 1, 2, · · ·, n, if the sample mean x of x

is zero, show that cov (&., ~) = 0, where &. and ｾ＠ are the least square estimators
2
of a and p. Assume that u 1, u 2 , ···, un are independent and are from N(O, cr )
distribution. 15

(c) Define stationary time series process and autocovariance function. Show that
autocovariance function, denoted by y (h), is an even function, positive
semi-definite and uniformly continuous if it is continuous at h = 0. 15

2. (a) Assume that the correlation function of a continuous parameter process is

given by p (t) = ae-b It I, a, b > 0. Find the spectral density function for the
process. Comment on the density curve with respect to its parameter. (Here,
t is the time lag) 10

(b) Define simple random sampling with replacement and without replacement. In
simple random sampling without replacement, show that the sample mean y is
unbiased estimate of Y, the population mean. Derive its variance. Also, find the
standard error of the estimate of the population total. 15

(c) Describe the problem of heteroscedasticit y. How does one detect it? What
happens to OLS estimators if we introduce heteroscedasticit y? 15

DFSE-T-SIT/5 2

Examrace www.examrace.com
 SECTION-B

Answer any THREE questions of the SIX questions given below

3. (a) Explain the concept of interpenetrat ing sub-sampling . 10

(b) Define ratio estimator for the population ratio, R. Obtain its bias. In SRSWOR,
for large n, derive its variance. Deduce the variance of the estimator for
population total. 15

(c) Define Des Raj ordered estimator for population mean for the case when the
sample size is 2. Show that it is unbiased. Derive its variance. 15

4. (a) Consider an econometric model where two variables Y and X are jointly
determined by the following equations :

The Greek letters denote unknown parameters, U and V are model errors,
mutually uncorrelated , with zero mean, and Z is an exogenous variable,
independent of the errors. Show that OLS estimator of a is asymptotical ly
2
biased in general. 10

(b) Give some practical examples for reasons of lags in distributed lag models and
explain the Koyck approach to distributed lag models. 15

(c) Consider a regression function y=a+P 1x 1 +P 2 x 2 +u with u~IN(0, o- 2 ), which

relates annual salary (y) of employees for a large firm in terms of their years of
education (xi) and years of experience (x 2 ). The following data are obtained
from a sample of size n = 23

:X\ =10 .X2 =5 y =12

S11 =12 S12 = 8 s2 2 = 12

Sly= 10 S2y = 8 Syy = 10

DFSE-T-ST I/5 3 [P.T.O.

Examrace www.examrace.com
 (i) Compu te a, ~1 and ~ 2 . Presen t the regress ion equatio n.

sum of square s and multip le coeffic ient of

(ii) Determ ine residu al
determ ination .

(iii) Test the hypoth esis f:3 2 = 0 at the 5% signifi cance level.

(Symbo ls used have their usual interpr etation ) 15

PERCENTAGE POINTS OF THE t-DISTRIBUTION

Example :
Pr (t > 2.086) = 0.025
Pr (t > 1.725) = 0.05 for df = 20
Pr(ltl> 1.725) = 0.10
0 1.725

ｾ＠
0.25 0.10 0.05 0.025 0.01 0.005 O.OQ1
0.50 0.20 0.10 0.05 0.02 0.010 0.002

1 1.000 3.078 6.314 12.706 31.821 63.657 318.31

2 0.816 1.886 2.920 4.303 6.965 9.925 22.327
3 0.765 1.638 2.353 3.182 4.541 5.841 10.214
4 0.741 1.533 2.132 2.n6 3.747 4.604 7.173

0.727 1.476 2.015 2.571 3.365 4.032 5.893

5
6 0.718 1.440 1.943 2.447 3.143 3.707 5.208
7 0.711 1.415 1.895 2.365 2.998 3.499 4.785
8 0.706 1.397 1.860 2.306 2.896 3.355 4.501
9 0.703 1.383 1.833 2.262 2.821 3.250 4.297

0.700 1.372 1.812 2.228 2.764 3.169 4.144

10
0.697 1.363 1.796 2.201 2.718 3.106 4.025
11
12 0.695 1.356 1.782 2.179 2.681 3.055 3.930
0.694 1.350 1.n1 2.160 2.650 3.012 3.652
13
0.692 1.345 1.761 2.145 2.624 2.9n 3.787
14
15 0.691 1.341 1.753 2.131 2.602 2.947 3.733
0.690 1.337 1.746 2.120 2.583 2.921 3.686
16
17 0.689 1.333 1.740 2.110 2.567 2.898 3.646
0.688 1.330 1.734 2.101 2.552 2.878 3.610
18
0.688 1.328 1.729 2.093 2.539 2.861 3.579
19
0.687 1.325 1.725 2.086 2.528 2.845 3.552
20
0.686 1.323 1.721 2.080 2.518 2.831 3.527
21
0.686 1.321 1.717 2.074 2.508 2.819 3.505
22
0.685 1.319 1.714 2.069 2.500 2.807 3.485
23
1.318 1.711 2.064 2.492 2.797 3.467
24 0.685
0.684 1.316 1.708 2.060 2.485 2.787 3.450
25
1.315 1.706 2.056 2.479 2.n9 3.435
26 0.684
1.314 1.703 2.052 2.473 2.n1 3.421
27 0.684
1.701 2.048 2.467 2.763 3.408
28 0.683 1.313
1.311 1.699 2.045 2.462 2.756 3.396
29 0.683
1.310 1.697 2.042 2.457 2.750 3.385
30 0.683
1.684 2.021 2.423 2.704 3.307
40 0.681 1.303
2.000 2.390 2.660 3.232
60 0.679 1.296 1.671
2.358 2.617 3.160
120 o.6n 1.289 1,658 1.980
2.576 3.090
00 0.674 1.282 1.645 1.960 2.326

area in one tail; the

Note : The smaller probability shown at the head of each column is the
larger probability is the area in both tails.

DFSE- T-STI /5 4

Examrace www.examrace.com
 5. (a) The prices and quant ities of four comm oditie s produ ced
by a manu factur ing
firm during the period 2015 and 2016 are given below
:

Quant ity (in kg) during Price (in ~) during

Comm odity
2015 2016 2015 2016
A 175 201 1,540 1,030
B 32 46 1,270 1,490
C 48 43 2,760 2,490
D 65 66 2,190 2,070

Calcu late-

(i) Laspe yres' index numb er;

(ii) Paasc he's index numb er;

(iii) Fisher 's index numb er

for 2016 assum ing 2015 as the base year.

10

(b) What is mean t by correl ogram in time series analys is?

Obtai n the genera l
expre ssions for correl ogram s of AR(l) and MA(q) model s.
15

(c) State ergodi city prope rty of a statio nary time series proce
ss. If {Xt, t E T} is a
statio nary time series proce ss with E(Xi) = µ and var(Xi )
= cr 2 for all t, define
ergodi city ofµ. Also, show that the proce ss whose covari ance
functi on y (h) ＠ 0
as h ＠ oo is ergodi c for µ .
15

6. (a) Distin guish betwe en two-st age sampl ing and two-p hase sampl
ing proce dures.
State the situat ions in which they are used.
10

DFSE -T-ST T/5 5 [P.T.O .

Examrace www.examrace.com
 size is an intege r
(b) Descr ibe system atic sampl ing proce dure when the popul ation
system atic sampl ing
multip le of sampl e size. Menti on its advan tages. Comp are
15
with stratif ied sampl ing and cluste r sampl ing.

on of the form
(c) Show that in a stratif ied sampl ing with a linear cost functi
L
C = c0 + L ch nh the varian ce of the estima ted mean Yst is a minim um for a
h=l
varian ce V(y 8 t), when
specif ied cost C and the cost is a minim um for a specif ied
nh = wh sh I ..jc;.
is the cost per unit
(Here, c 0 is the overh ead cost, Lis the numb er of strata , ch
weigh t of stratu m h
for stratu m h, nh is the sampl e size of stratu m h, Wh is the
15
and sl is the true varian ce of stratu m h)

are the metho ds of

7. (a) Define multic olline arity for the genera l linear model . What
of the regres sion
detect ing it? Also, state the effect on OLS estim ators
arity betwe en two
coeffi cients and their varian ces, if there is perfec t multic olline
10
explan atory variab les of the model Y =a.+13 1X 1 +13 2 X 2 +U.

conne ction with

(b) Expla in the proble m of identi ficatio n. Does it have any
of identif iabilit y. 15
multic olline arity? Devel op the rank and order condi tions

coeffi cients of the model Yi = 131+13 2 X 2 i +13 3 X 3 i +Ui

with
(c) Estim ate the
on a sampl e of
the restric tion on the coeffi cients as 13 2 +13 3 = 1, based
25 observ ations , which yielde d the follow ing inform ation

2 15
[Assu me that Ui ~ IN(0, cr )]

's law, derive the

8. (a) State Pareto 's law of incom e distrib ution. Using Pareto
a and b being the
proba bility densit y of incom e variab le X, where a< X < b, with
10
lowes t and highe st incom e, respec tively.

DFSE -T-ST I/5 6

Examrace www.examrace.com
 {b) State the general form of spectral density function for a stationary time series
process. Derive the expressions for spectral density functions of MA(l) and
AR(l) processes. 15

(c) Enunciate, in detail, the principal stages involved in setting up a Box-Jenkins

forecasting model. Also, write a note on Box-Pierce Q-statistic involved in
autocorrelation tests. 15

***

DFSE-T-STI/5 7 21BS-12180

Examrace www.examrace.com
Examrace www.examrace.com