This question paper contains 4+1 printed pages

Roll No.

S. No. of Question Paper 1169

Unique Paper Code 237604 G

Name of the Paper Bio-Statistics

Name of the Course B.Sc. (Hons.) Statistics

Semester M

Duration: 3 Hours Maximum Marks : 75

(Write your Roll No. on the top

immediaely on receipi of this question paper.

Attempt five questions in all,

selecting wo from Sections B and C each

Section A is
compulsory.
Use of simple calculator is allowed.

Section A

(a) Define death density function, survival function and

hazard function when survival time follows

lognomal
distribution. Find mean survival time and variation in
survival time.

P.T.O
 3 16
(6) Consider the following two Weibull distributions space as Section 1
Survival models
(a) Detine type-ll random censoring. Under this censoring
(0 Scale parameter 1, shape parameter0.5 scheme estimate mean survival time
assuming that the

()Scale parameter 0.5, shape parameter2

survival time follows exponential distribution.
For each distribution, find
(6) Estimate crude probability of deatlh when the joint
() The mean and variance
distribution of d. dd
) Nature of hazard function.
and ,. gven

, follows multinomial. Also find

ECQ). Var(Q,)
() For the following survival data, compute estimated

Survival function, probability density function and hazard

CovO9)6
(a) Define crude and
function net probability (type-A and type-B) of

death. Stating assumptions, establish

Year of follow-up Number Alive at the Number Dying in the the
inter-relationship
between them.
Beginning interval

of interval ) Suppose that two risks R and R, (o * ) are operating

0-1 in the population such that

2000 Q Show that

S60 105
1300 (a) Explain all the phases of clinical
drug trials.

1S0 00 6) What is duration of an

epidemic ? Obtain mean

1050 0 duration of epidemic

an under simple stochastic
960 epidemic model.

744 P.T.O.
 1169 169

Section C

()Explain life table method to estimate the survival function,

also compute the variance of the estimate for survival

function.
19
(6) Define partially crude probability of death and show

ar the deathintensitiescormesponding
1

()
that
to risks R R. Rgs B respectively then the

1- 4 -,) probability of dying due to risk R { 1 , 2, )iSs

a/A), where - 2A
() Consider the following tumor free time (Gin days) of l0 rats

on a low fat diet. Calculate Kaplan Meier estimate of S) Define survival function, death density function and hazard

for all the rats and S.E. of S) at 1 84. Tunction. Find

Rat No. Tumor Free time (0 St and ) when o) c

(i) S ) and K ) when o ) = ae.

(b Find death density function when competing risks are

dependent. Also obtain the death density function for a

bivariate dependent risk model when po, o 5,10p

1169
This question paper contains 42 printed pages)

Roll No.

S. No. of Question Paper 1167

Unique Paper Code 237602

Name of the Paper Design of Experiments

Name of the Course B.Sc. (H) Statistics

semester V

Maximum Marks: 75
Duration:3Hours
Write yvour Roll No. on the top immediately on receipt of this question paper)

Attempt f e questions. in all,

selecting ihree from Section I and nwo from Section 1

Section

1.(a Explain the terms treatments, blocks, experimental error,

fertility gradient. uniformity trials and fertility contour

map along with their utility in Design of Experiments.

(6) How do the size and shape of the plots and blocks

aflect the results of tield experiments 2 Explain.

P.TO.
 2 3 167
1167
( In a randomized block design there b) A randomised block experiment
has been carried out
are
only two
blocks. treatments A, B. C. D and E. The
Let & be the on 4 blocks with 5
number of treatments and
treatment A in block 2 is
to
and ya be the average
yields reading corresponding
of two blocks. Show
can be used
that the between blocks missing Explain how ANOCOVA techinique
sum of
squares can be
oft the plot. Also show
to estimate the missing yield
oressed as )
73.5 same as that obtained by Yate's missing
that it is the
2 () Derive the ANOVA table for an L.S.D. How iot technique.
would
you test For a symmetric BisD with parameters v, A, and A.
(a)

0 The define Complementary and Residual designs and obtain

hypothesis of
equality of all
treatment for the followings
their parameters. Also construct these
etfects, nd
BIBDs.
BIBD and verify if the resultant designs are
(i The hypothesis of
equality of two
specific
treatment effects?

(6) Obtain the standard

error
of difference between two
treatment means for R.B.D., one of which involves a

missing plot
(a) Discuss how the 0
efficiency of an experiment can be
increased by using local control and by increasing the
number of
replications. Obtain the etliciency of a L.S.D.
relative to R.B.D. taking rows as blocks and efficiency
of R.B.D. relative to C.R.D. stating clearly
the
assumptions used in the derivation. P.T.O.
 167 I 167

factors A. B and C is
with
Find the variance of the estimated elementary treatment factorial experiment
() (c) A 3
3 blocks each. If
some

contrast of a BIBD. Compare it with that of RBD. 9,6 arranged in a single replicate of

of the blocks are given as

of the elements of one
Section I1
12
Replicate : 000, 011,
5. (a) What are factorial experiments ? Derive the contrasts
blocks. Also, list
Write down the contents of all the

due to main and interaction effects in a tactorial

etfects confounded.
9,3.3
all the

experiment with twO Tactors at three levels each without in factorial experiments
(a) What is meant by confounding

tabie. between partial and total confounding

using Yates's table and plus and minus signs Distinguish
2
Construct a system of partial confounding for a
Also show that these contrasts are mutually orthogonal.

blocks of size
factorial experiment in 3 replications with
(6) If a 2" factorial experiment is conducted in 2 plots per
infomation can be obtained
4. so that at least partial
DiOcK, then
and
wo and three factor interaction components
about

(0 how many treatment combinations can be allocated

the main effects.
full information about

independently, in the principal block tactorial designs

mean by Iractional
(6) What
do you

(i) how many effects will be confounded in total per

Define the terns

replication? How many of them are independent ) Defining relation

P.T.O.
 This question paper contains 4 printed pages
167 Roll No. |

n) Principal fraction and S. No.

of Question Paper 1166

(in Aliases Unique Paper Code 237601

Name
7. (a) Define Resolution 1, V and V designs. Construct of the Paper Statistical Inference l1

Name of the Course B.Se. (H) Statisties

23 design with defining relationsI ACE and
Semester
I- ABCD. Write down the alias structure of
the design.y
What is the resolution of this
Duration:3 Hours Maximum Marks: 75
design ?
(Write your Roll No. on the top immediately on
receipt of this question paper)
(6) Construct a
2 factorial experiment, the
confounded Attempt six questions in all,
selecting
effects being ABC and ADE.
three from each section

Section
a) Define *Type" I error and power of the test. If
x2I is the critical region for testing H,:0=2 against
H, :8=I on the basis of a single observation from the
population with p.d.f. f(x. 0)=6e x>0, 0 <0<
Obtain the values of type I and
type ll erors and the power
of the test.

b) What are
simple and composite hypotheses 2 State and prove
Neyman-Pearson lemma for testing a
simple hypothesis
against a simple alternative.
5.7%
PTO.
1167
 1166

4 Discuss the method of construction of likelihood ratio test

2 () If w is an MP region for testing H, :0=, against
for
and state its properties, Construct likelihood ratio test
H,:0 , . then prove that it is necessarily unbiased.
testing the cquality of the means of two normal populations
Also prove that the same holds good if W is an UMP
with common unknown variance. 12%

region. Section 11

S.P.R.1.. its OC and ASN functions.

Let be random sampie Describe
(6) X, X2 a
from
Construct S.P.R.T. for testing H,:0=0 against
NG4, a) where g is known. Obtain the B.C.R. of size
the basis of a random sample
H,:0-6,(0 < , <6,) on

a for testing H,:o o against H, :o oj(o Poisson distribution with parameter 8 Also
drawn trom a

Hence obtain the power function of the test. 6%6

obtain its OC and ASN functions

3. ( Given a random sample X. X2 Trom are the advantages of non-parametric tests Define a
What
distribution with p.d.f. of a run. Describe the r est detail
run and the length
the equality of the awo populations.
S a . 8) = e , x 2 0 , 8 > 0 .
for testing
Discuss the Mann-Whitney-Wilcoxon test for testing
for (a)
Obtain the UMP critical region of size a
testing
two are drawnfrom the same
whether samples
H, 0 - 0 , against H 0 > 8 o in terms of
How is the test carried out for
continuous population.
chi-square statistic, Also obtain the power function o t
Also discuss the case of ties.
large samples
test for testing the null
ne test.
() Discuss the Kruskal-Wallis

same contanuous
(6) Consider n Bermoullian trials with probability of success hypothesis that K-samples come from

befhaviour ot he
P tor each trial. Derive the likelihood ratio test for population. Explain the large sample

St.
6,6%
testing Ho :P= Po against H:P> Po 6%6 P.T.O
 4
1166

(a) For the S.P.R.T. of strength:

(B,) and for given A- and B P
I-ax
prove that

(i) a +B sa+B.
(6) Describe the sign test, stating clearly the assumptions
involved. Also discuss the paired sample
sign
test.
5,7%

WPSOffice

1166