M.Sc. (Semester IV) (CBCS) Examination Nov/Dec-2018
Statistics
DISCRETE DATA ANALYSIS
Time: 2½ Hours Max. Marks: 70
Instructions: Attempt five questions.
Q. No. 1 and Q. No. 2 are compulsory.
Attempt any three questions from Q. 3 to 7.
Figures to the right indicate full marks.
Q.1 Select correct alternative: 05
If a response variable in GLM follows Poisson distribution then following
link function is suitable.

None of these
The degrees of freedom associated with the U12 term in a log linear
model for an I × J table are
I J I − 1 J
I − 1 J − 1 IJ
For a 2 × 2 table, the cross product ratio always lies between
0,½
∞
Which one of the following measures is used to test the goodness of fit of
a GLM?
t Statistic Deviance
F ratio None of these
G2 statistic is distributed as
Asymptotically Chi-square Normal
Binomial Gamma
Fill in the blanks: 05
Logistic regression model is appropriate when the response variable is

The cross product ratio in log-linear model is also known as
The logistic regression model on y on x
In regression analysis when outcome variable is dichotomous lies
in to
In a log-linear model U23 is a higher order relative of and
State whether the following statement are True or False: 04
If the state of A is independent of state of B then
2 − − − − I
2 − − − − I
For a 2 × 2 table, log-linear model, − is used as measure of
association.
A GLM with as identity link is the classical linear model.
A Poisson regression model is also appropriate for describing an I × J
table.
Page 2 of 2
SLR-VR-496
Q.2 Explain the terms: 06
Analysis of deviance
ROC curve
Write short notes on the following: 08
Hierarchical family of models
Over dispersion problem in Poisson regression
Q.3 Explain the following terms: 07
One parameter exponential family of distribution.
GLM and link function.
Derive Nelder and Wedderburn's weighed least squares estimator of the
parameters of a GLM.
07
Q.4 State and establish a condition for existence of direct estimates of
elementary cell frequencies in an × J × Table.
07
Explain: 07
Multinomial and Poisson sampling scheme.
Birch's results
Q.5 Derive the maximum likelihood estimates of the parameters involved in the
logistic regression model
07
Explain the Poisson regression model. Give an appropriate example. Derive
the score equation for the same.
07
Q.6 Describe log-odds ratio with reference logistic regression model with single
covariate. Discuss its computational procedure.
07
Explain variable selection problem. State AIC and BIC criteria. How they
can used for selection of variables.
07
Q.7 Explain the following terms: 07
Relative risk
Non-comprehensive model
Configuration
Describe Pearson's Chi-square test for goodness of fit of a logistic
regression.