AHK/KW/19/1395

Master of Science ([Link].) Semester—II (CBCS) (Statistics) Examination

LINEAR MODELS AND DESIGNS OF EXPERIMENTS
Paper—4
Paper—VIII
Time : Three Hours] [Maximum Marks : 80
Note :— All questions are compulsory and carry equal marks.
EITHER

1. (A) Explain elementary concepts of analysis of variance for one way and two way classified data.

(B) Explain missing plot technique in R.B.D. when exactly two observations are missing. Find the
estimates of the missing observations. 8+8

OR

(C) Given 3 independent variates y1, y2 and y3 having common variance σ 2 such that
E(y1) = θ1 + θ2, E(y2) = θ1 + θ3 and E(y3) = θ1 + θ2. Show that l1θ1 + l 2θ2 + l 3θ3 is estimable
iff l1 = l 2 + l 3 . Find unbiased estimator of σ 2.

(D) Give complete statistical analysis of L.S.D. 8+8

EITHER

2. (A) Define B.I.B.D. State and prove the necessary conditions on the parameters of B.I.B.D.

(B) Define Youden Square Design. Give its complete statistical analysis. 8+8

OR

(C) Write the mathematical model of B.I.B.D. Give one example. Explain the construction of B.I.B.D.
by developing initial blocks.

(D) Explain RBIBD. Prove that for a RBIBD, rank N ≤ b – (r – 1). (Symbols carry usual meanings).
8+8

EITHER

3. (A) What is split plot design ? Explain the procedure of randomisation in split plot design. Obtain
variance of the estimates of treatment contrasts in this design.

(B) Carry out the analysis of 23 factorial experiment arranged in R.B.D. 8+8

OR

(C) Explain the technique of analysis of co-variance for two way classified data.

(D) Define 24 factorial experiment. Also explain main effects and interaction effects in 24 factorial
experiment. 8+8

CC—3209 1 (Contd.)
 EITHER
4. (A) Explain the need of confounding in factorial experiments and describe in detail complete confounding
and partial confounding.
(B) Give the structure of blocks in a 24 factorial experiment to confound :
(i) AB, CD, ABCD in first replication
(ii) AC, BD, ABCD in second replication
(iii) AD, ABC, BCD in third replication
(iv) BC, ABD, ACD in fourth replication. 8+8
OR
(C) Explain the concept of generalized interaction. Mention the problems created due to it.
(D) Explain the concepts of double confounding and simultaneous confounding with an example.
8+8
5. (A) Give the assumptions and applications of L.S.D. in field experimentations.
(B) Define PBIBD. Is it always connected ? State the parametric relations in PBIBD.
(C) Define factorial experiments. Why are these considered better than experiments in which factors
are tried one by one.
(D) Explain ‘orthogonality’. Show that in a 23 factorial experiment, the main effects and interactions
are mutually orthogonal. 4×4

CC—3209 2 AHK/KW/19/1395
 NIR/KW/18/2662
Master of Science ([Link].) Semester–II (CBCS) (Statistics) Examination
LINEAR MODELS AND DESIGNS OF EXPERIMENTS
Paper–4
Paper–VIII
Time : Three Hours] [Maximum Marks : 80
Note :— All questions are compulsory and carry equal marks.
EITHER
1. (a) State and prove Gauss– Markov theorem.
(b) Explain the concept of RBD. Explain the method of estimating missing values in RBD when
exactly one observation is missing. 8+8
OR
(c) Define LSD. Describe in detail the analysis of LSD.
(d) Obtain the estimates of two missing values in RBD considering all the possibilities. 8+8
EITHER
2. (a) Define BIBD. Explain all its parameters. Give complete analysis of BIBD. 16
OR
(b) Write the model of SBIBD. When is BIBD said to be SBIBD ?
(c) Show that if b = v and v is even then (r–λ) will be perfect square (symbols carry usual meanings).
(d) For a resolvable BIBD, show that b ≥ v + r – 1. 4+6+6
EITHER
3. (a) Considering analysis of covariance model in an RBD, obtain a test for testing H0 : r = 0 and
obtain an unbiased estimate of tj – tj' , j ≠ j' and hence find the confidence interval.
(b) Give a mathematical model for split plot design. State the assumptions of the parameters involved.
Discuss advantages of split plot design. 10+6
OR
(c) Find the efficiency of split plot design relative to RBD.
(d) Describe Yate's method for computing 24 factorial effect total. 8+8
EITHER
4. (a) Describe the concept of partial and complete confounding considering 24 factorial experiment.
Give suitable examples of both.
(b) Discuss advantages and disadvantages of confounding. 10+6
OR
(c) A 23 experiment with factors A, B, C is to be conducted in 4 replicates consisting of two blocks
having four plots each. Two experimentors conduct such four replicate experiments in two
different forms. In one experiment ABC is totally confounded and in other AB, AC, BC and ABC
are partially confounded. Give complete analysis of the above. 16
5. (a) Define estimability. Show that all linear parametric functions are estimable iff the number of
unknown parameters equals the rank of the matrix of known coefficients.
(b) Prove or disprove : PBIBD is a grouped divisible design.
(c) What is 23 factorial experiment ? Give its main and interaction effects.
(d) Explain the concept of double confounding. 4+4+4+4

BKR—4581
 NRT/KS/19/2662
Master of Science ([Link].) Semester-II (CBCS) (Statistics) Examination
LINEAR MODELS AND DESIGNS OF EXPERIMENTS
Paper—4
Paper—VIII
Time : Three Hours] [Maximum Marks : 80
N.B. :— All questions are compulsory and carry equal marks.
EITHER
1. (A) Describe basic principles of designs of experiments. Explain how they are met in R.B.D.
(B) Describe L.S.D. and give complete analysis of L.S.D. when one observation is missing.
6+10
OR
(C) State and prove Gauss-Markov Theorem. 16
EITHER
2. (A) Write the mathematical model of BIBD and obtain the least square estimates of treatment
effects without recovery of inter-block information.
(B) Explain the difference between Youden Square Design and symmetric B.I.B.D. Find the
efficiency of Youden Square Design over corresponding symmetric B.I.B.D. 8+8
OR
(C) Define resolvable B.I.B.D. and affine resolvable B.I.B.D. (ARBIBD).
For RBIBD, prove that
b≥ ν + r-1
State the relationship if the design is ARBIBD.
(D) Define Youden Square Design. State its mathematical model. Carry out its analysis. 8+8
EITHER
3. (A) Discuss split-plot design with an example. Give the ANOVA table for the whole plot
analysis and sub-plot analysis.
(B) Explain the technique of analysis of covariance for one way classified data. 8+8
OR
(C) Carry out the analysis of 23 factorial experiment arranged in R.B.D.
(D) Explain the technique of analysis of covariance for two way classified data. 8+8
EITHER
4. (A) What is meant by confounding in factorial experiments ? State its advantages. Explain the
concept of double confounding.
(B) Explain the concept of generalised interaction with an example. Also state the problems
created due to generalised interaction. 8+8
OR
(C) Differentiate between partial confounding and complete confounding with an example in
24 factorial experiment.
(D) Give the layout of 23 factorial experiment with 4 replicates such that in replicates 1, 2, 3,
4 effects A, B, AB and ABC respectively are confounded. 8+8
5. (A) Explain elementary concept of analysis of variance for one way classified data.
(B) Obtain the efficiency of B.I.B.D. relative to R.B.D.
(C) Differentiate between simple experiments and factorial experiments with an example.
(D) Write a note on simultaneous confounding. 4×4=16

CLS—239
 NRJ/KW/17/3820
Master of Science ([Link].) Semester—II (CBCS) (Statistics) Examination
LINEAR MODELS AND DESIGNS OF EXPERIMENTS
Paper—4
Paper—VIII
Time : Three Hours] [Maximum Marks : 80

Note :— ALL questions are compulsory and carry equal marks.

EITHER

1. (A) Explain missing plot technique in R.B.D. when exactly one observation is missing. Estimate it
and give the subsequent analysis.

(B) Define L.S.D. Give its layout and analysis for 5 treatments. 8+8

OR

(C) Define C.R.D. and R.B.D. with their advantages and disadvantages. State the situation when
R.B.D. is more appropriate than C.R.D.

(D) Give the statement of Gauss-Markov theorem stating the assumptions clearly. Comment on the
significance of these assumptions. 8+8

EITHER

2. (A) Define B.I.B.D. In what sense is it incomplete and balanced ? State the parameters of BIBD and
prove that bk = Vr and r(k – 1) = λ(v – 1).

(B) State and prove Fisher’s inequality. 8+8

OR

(C) What is S.B.I.B.D. ? Show that if the no. of treatments common between any two blocks is
constant then b = V(with usual notations)

(D) Explain Youden square design and its linear model. State its advantages over L.S.D. 8+8

EITHER

3. (A) Explain factorial experiments with its advantages. Give Yates method to calculate factorial effect
totals for 23 factorial experiment.

(B) Explain the technique of analysis of covariance for two-way classified data. 8+8

OR

(C) What is split-plot design ? State its advantages. Write the ANOVA table for whole plot analysis
and sub-plot analysis.

(D) Explain the technique of analysis of covariance for one-way classified data. 8+8

POY—25406 1 (Contd.)
 EITHER
4. (A) Explain the need of confounding in factorial experiments and describe in detail complete and
partial confounding.
(B) Given the structure of blocks in a 24 factorial experiment to confound :
(i) AB, CD, ABCD in first replication
(ii) AC, BD, ABCD in 2nd replication
(iii) AD, ABC, BCD in 3rd replication
(iv) BC, ABD, ACD in 4th replication. 8+8
OR
(C) Explain the concept of generalised interaction with an example. State the problems created due to
generalised interaction.
(D) Differentiate between partial confounding and complete confounding with an example in 23 factorial
experiment. 8+8
5. (a) Give the layout of RBD with 5 treatments A, B, C, D, E with 4 replications. Derive the efficiency
of R.B.D. over C.R.D.
(b) Prove OR disprove that a SBIBD with the following parameters is possible :
v = 36, b = 42, r = 7, K = 6, λ = 2.
(c) Explain 24 factorial experiment with illustration.
(d) Explain the concept of double confounding. 4+4+4+4

POY—25406 2 NRJ/KW/17/3820
 NJR/KS/18/3820
Master of Science ([Link].) Semester-II (C.B.C.S.) (Statistics) Examination
LINEAR MODELS AND DESIGNS OF EXPERIMENTS
Paper—4
Paper—VIII
Time : Three Hours] [Maximum Marks : 80
N.B. :— All questions are compulsory and carry equal marks.
EITHER
1. (A) Prove that treatment mean sum of squares and error mean sum of squares are unbiased

estimates of σ2e in a C.R.D. and R.B.D.

(B) Explain missing plot technique in R.B.D. when exactly one observation is missing. 8+8
OR
(C) Derive an expression for expected value of sums of squares due to rows, columns, treatment
and error in a L.S.D.
(D) Explain missing plot technique in L.S.D. when exactly one observation is missing.
8+8
EITHER
2. (A) Define B.I.B.D. In what sense is it incomplete and balanced ? State the parameters of
B.I.B.D. and prove that bk = vr and r(k – 1) = λ(v – 1). Where notations have their usual
meanings.
(B) State and prove Fisher’s inequality. 8+8
OR
(C) What is S.B.I.B.D. ? Show that if the number of treatments common between any two
blocks is constant, then b = v (in usual notations).
(D) Explain Youden square design and its linear model. State its advantages over L.S.D.
8+8
EITHER
3. (A) What is split-plot design ? State its advantages. Write the ANOVA table for whole plot
analysis and sub-plot analysis.
(B) Explain the technique of analysis of covariance for one way classified data. 8+8
OR
(C) Explain factorial experiments with its advantages. Give Yate’s method to calculate factorial
effect totals for 23 factorial experiment.
(D) Explain the technique of analysis of co-variance for two-way classified data. 8+8

RQA—268 1 (Contd.)
 EITHER
4. (A) Explain the concept of generalised interaction with an example. State the problems created
due to generalised interaction.
(B) Differentiate between partial confounding and complete confounding with an example in
23 factorial experiment. 8+8
OR
(C) Give the layout of 23 factorial experiment with 3 replicates such that in replicates 1, 2, 3
effects AB, BC, ABC are respectively confounded.
(D) Explain double confounding in detail. 10+6

5. (A) If θ̂ is BLUE for θ then show that it is a p-variate normal with mean θ and co-variance
(A'A)–1 σ2 .
(B) Obtain an efficiency of B.I.B.D. relative to R.B.D.
(C) Explain 24 factorial experiment with an example.
(D) Explain the concept of simultaneous confounding in factorial experiments. 4+4+4+4

RQA—268 2 NJR/KS/18/3820
 KNT/KW/16/5833

Master of Science ([Link].) Semester–II Choice Based Credit System (C.B.C.S.)

(Statistics) Examination
LINEAR MODELS AND DESIGNS OF EXPERIMENTS
Paper–4
Time : Three Hours] [Maximum Marks : 80
N.B. :— All questions are compulsory and carry equal marks.
EITHER
1. (a) Describe the analysis of variance of randomized block design.
(b) Determine the standard error of the difference between means of two varieties in R.B.D.
10+6
OR
(c) State and prove Gauss-Markov Theorem.
(d) Define estimability and show that all linear parametric functions are estimable iff the number of
unknown parameters equals the rank of the matrix of known coefficients. 8+8
EITHER

r2 − 1
2. (a) Define BIBD with parameters v, b, r, k and λ. Show that for BIBD b ≤ and
λ
b ≥ v + r – k.
(b) State and prove Shrikhande’s inequality. 8+8
OR
(c) When is BIBD said to be RBIBD. State the parametric relation of RBIBD.
(d) Write down SBIBD. State the difference between BIBD and SBIBD. Give suitable example of
SBIBD. 8+8
EITHER
3. (a) Explain split plot design. Write down the experimental layout for a split-plot design where whole
plots are laid in a RBD with 3 blocks (replicates) and two factors ‘a’ and ‘b’. Give complete
analysis of it.
(b) Give complete analysis of covariance in one way. 10+6

NVM—5715 (Contd.)
 OR
(c) Explain Factorial Experiment (FE). What are the advantages of F.E. ? Describe Yate’s method
of computing factorial effect for 24 F.E.
(d) Write down the ANOVA table for 24 F.E. 10+6
EITHER
4. (a) Describe partial confounding and complete confounding in 24 F.E. What are the advantages of
partial and complete confounding ? How is the information on partial confounded interaction
obtained ? 16
OR
(b) In 23 experiment with factors a, b, c is to be conducted in 4 replicates consisting of two 4-plot
blocks. Two experimenters conduct such 4-replicate experiments into different farms–in one
experiment ABC is totally confounded and in other AB, BC, AC and ABC are partially confounded.
How will you make a combined analysis of this experiment ? 16
5. (a) What is Gracolatin square design ? Write down 5×5 Gracolatin square design.
(b) Obtain efficiency of BIBD relative to RBD.
(c) Explain the model of two-way analysis of co-variance.
(d) Explain the concept of generalized interaction. 4×4=16

NVM—5715
 NKT/KS/17/5833

Master of Science ([Link].) Semester–II (C.B.C.S.) (Statistics) Examination

LINEAR MODELS AND DESIGNS OF EXPERIMENTS
Paper–4
Paper–VIII

Time : Three Hours] [Maximum Marks : 80

N.B. :— All questions are compulsory and carry equal marks.
EITHER

1. (A) State and prove Gauss-Markov Theorem. 16

OR
(B) Explain the basic principles of design of experiments.

(C) Give complete analysis of RBD when one observation is missing. 6+10

EITHER
2. (A) Define :

(i) BIBD

(ii) PBIBD
(iii) SBIBD

(iv) RBIBD.

(B) Define Youden square design. State its advantages. State the mathematical model of Youden
square design and corresponding normal equations. 8+8
OR

(C) Write the mathematical model of BIBD and obtain the least square estimates of treatment effects
without recovery of inter-block information.

(D) State and prove Fisher’s inequality. 8+8

NXO—20875 1 (Contd.)
 EITHER
3. (A) What is split plot design ? State its advantages. Give the ANOVA table for the whole plot analysis
and sub-plot analysis.

(B) Explain the technique of analysis of covariance for one way classified data. 8+8

OR
(C) Carry out the analysis of 23 factorial experiment arranged in R.B.D.

(D) Explain the technique of analysis of covariance for two-way classified data. 8+8

EITHER
4. (A) What is meant by confounding in factorial experiments ? State its advantages. Explain the concept
of double confounding.

(B) Explain the concept of generalised interaction with an example. Also state the problems created
due to generalised interaction. 8+8

OR
(C) Differentiate between partial confounding and complete confounding with an example.

(D) Give the layout of 23 factorial experiment with 3 replicates such that in replicates 1, 2, 3 effects
A, B, AB respectively are confounded. 8+8

5. (A) Derive an estimate of one missing value in LSD.

(B) Obtain the efficiency of B.I.B.D. relative to R.B.D.

(C) Differentiate between simple experiments and factorial experiments with an example.

(D) Write a note on simultaneous confounding. 4+4+4+4

NXO—21875 2 NKT/KS/17/5833
 TKN/KS/16/6154

Master of Science ([Link].) (Statistics)

Semester–II (C.B.S.) Examination
LINEAR MODELS AND DESIGNS OF
EXPERIMENTS
Compulsory Paper—4 MST-204

Time—Three Hours] [Maximum Marks—100

N.B. :— All questions are compulsory and carry equal
marks.

EITHER
1. (a) Give complete statistical analysis of ANOVA for
one way classification.
(b) Explain missing plot technique in RBD for two missing
observations. 10+10
OR
(c) State and prove Gauss-Markov Theorem.
(d) Give the layout of LSD for 4×4 treatments. State
the advantages of LSD over RBD. 12+8
EITHER
2. (a) Define :
(i) Incomplete Block Design
(ii) BIBD
(iii) SBIBD
(iv) RBIBD
(b) Give complete analysis of BIBD. 10+10
MXP—M—3560 1 Contd.
 OR (c) Explain the concept of simultaneous confounding in
(c) Explain the parameters of BIBD. Prove that : 24 EE.

(i) νr = bk (d) Give the layout of 22 factorial experiment with

3 replicates such that in replicates 1, 2, 3 effects A,
(ii) λ(ν – 1) = r (k – 1)
B, AB respectively are confounded. 10+10
where notations have their usual meanings.
5. (a) Explain the missing plot technique in LSD with one
(d) Give complete statistical analysis of Youden Square missing value.
Design. 10+10
(b) Define PBIBD. Explain its parameters.
EITHER
(c) Explain Yate's method for computing factorial effect
3. (a) Explain ANOCOVA for one way classification. total in 23 factorial experiment.
(b) Explain the terms main effects and interaction effects (d) Explain the concept of double confounding in factorial
in factorial experiments. Differentiate between simple experiments. 5×4=20
experiment and factorial experiments. 10+10
OR
(c) Explain ANOCOVA for two-way classification.
(d) Give statistical analysis of split plot design. 10+10
EITHER
4. (a) Explain the concept of confounding in factorial
experiments. Differentiate between partial confounding
and complete confounding.
(b) Discuss the concept of generalized interaction with
an example. 12+8

OR
MXP—M—3560 2 Contd. MXP—M—3560 3 125
 TKN/KS/16/6521

Master of Science ([Link].) (Statistics) (C.B.C.S.)

Semester–II Examination
LINEAR MODELS AND DESIGNS OF
EXPERIMENTS (MST-204)
Paper—4/Paper–VIII

Time—Three Hours] [Maximum Marks—80

N.B. :— ALL questions are compulsory and carry
equal marks.

EITHER

1. (a) Define RBD. Write the model for RBD and perform
the analysis. State the advantages and disadvantages
of RBD. 16

OR

(b) Given 3 independent variates Y1, Y2 and Y3 having

common variance 62 such that E(y1 ) = Q1 + Q2 ,
E(y2) = Q1 + Q3 and E(y3) = Q2 + Q3. Show that
l1Q1 + l2Q2 + l3Q3 is estimable iff l1 = l 2 + l 3

(c) Derive the expression for relative efficiency of RBD

with respect to CRD. 8+8

MXP-M—3593 1 Contd.
 EITHER OR
2. (a) Define BIBD. Write the model of BIBD and perform (c) Explain the concept of double confounding,
its analysis. simultaneous confounding.
(b) Give the mathematical model of Youden Square Design (d) Give the layout of 22 factorial experiment with 3
and corresponding normal equations. 10+6 replicates such that in replicates 1, 2, 3 effects, A,
B, AB are respectively confounded. 8+8
OR
5. (a) Comment : "LSD is considered as incomplete three
(c) Describe SBIBD with its parameters along with an
way layout".
example. Give its analysis. 16
(b) State and prove Shah's inequality.
EITHER
(c) Give the advantages and disadvantages of confounding.
3. (a) Write down the Covariance model in RBD. Give its
(d) Explain split plot design. Give its layout.
analysis; obtain a test for Ho : v = 0. Obtain an
4+4+4+4
unbiased estimate of tj – tjl ' 16
OR
(b) Define mn factorial experiment. Also explain main
effect and interaction effect in factorial experiment.
(c) Explain the concept of concomitant variables in
ANOCOVA, with example. 10+6
EITHER
4. (a) Explain the terms 'complete confounding' and 'partial
confounding' with an illustration.
(b) Explain the concept of generalized interaction. Mention
the problems created due to it. 8+8

MXP-M—3593 2 Contd. MXP-M—3593 3 125

 TKN/KS/16/6521

Master of Science ([Link].) (Statistics) (C.B.C.S.)

Semester–II Examination
LINEAR MODELS AND DESIGNS OF
EXPERIMENTS (MST-204)
Paper—4/Paper–VIII

Time—Three Hours] [Maximum Marks—80

N.B. :— ALL questions are compulsory and carry
equal marks.

EITHER

1. (a) Define RBD. Write the model for RBD and perform
the analysis. State the advantages and disadvantages
of RBD. 16

OR

(b) Given 3 independent variates Y1, Y2 and Y3 having

common variance 62 such that E(y1 ) = Q1 + Q2 ,
E(y2) = Q1 + Q3 and E(y3) = Q2 + Q3. Show that
l1Q1 + l2Q2 + l3Q3 is estimable iff l1 = l 2 + l 3

(c) Derive the expression for relative efficiency of RBD

with respect to CRD. 8+8

MXP-M—3593 1 Contd.
 EITHER OR
2. (a) Define BIBD. Write the model of BIBD and perform (c) Explain the concept of double confounding,
its analysis. simultaneous confounding.
(b) Give the mathematical model of Youden Square Design (d) Give the layout of 22 factorial experiment with 3
and corresponding normal equations. 10+6 replicates such that in replicates 1, 2, 3 effects, A,
B, AB are respectively confounded. 8+8
OR
5. (a) Comment : "LSD is considered as incomplete three
(c) Describe SBIBD with its parameters along with an
way layout".
example. Give its analysis. 16
(b) State and prove Shah's inequality.
EITHER
(c) Give the advantages and disadvantages of confounding.
3. (a) Write down the Covariance model in RBD. Give its
(d) Explain split plot design. Give its layout.
analysis; obtain a test for Ho : v = 0. Obtain an
4+4+4+4
unbiased estimate of tj – tjl ' 16
OR
(b) Define mn factorial experiment. Also explain main
effect and interaction effect in factorial experiment.
(c) Explain the concept of concomitant variables in
ANOCOVA, with example. 10+6
EITHER
4. (a) Explain the terms 'complete confounding' and 'partial
confounding' with an illustration.
(b) Explain the concept of generalized interaction. Mention
the problems created due to it. 8+8

MXP-M—3593 2 Contd. MXP-M—3593 3 125