Assignment 2 Stat222
Assignment 2 Stat222
Department of Economics
1.
(50
marks)
Download
the
Excel
file
A2Q1.xlsx
from
Moodle.
The
spreadsheet
contains
observations
for
each
of
five
variables
for
90
American
cities:
Variable
Variable
Name
Description
Y
HouseValue
Mean
market
price
for
houses
X1
CommProp
Percentage
of
municipal
tax
revenue
from
commercial
property
X2
HouseSize
Mean
number
of
rooms
in
houses
X3
TaxRate
Tax
rate
per
thousand
dollars
of
assessed
value
for
houses
X4
Income
Mean
household
income
a. Use
Excel
to
estimate
the
following
relationship:
Yi = β0 + β1 X1i + β2 X 2i + β3 X3i + β 4 X 4i + εi
Attach
the
Excel
output.
(4
marks)
b. Use
EViews
to
estimate
the
same
relationship.
Attach
the
EViews
output.
(4
marks)
c. The
two
outputs
should
be
the
same.
Write
out
the
estimated
model,
including
the
standard
errors
in
parentheses
under
the
estimated
coefficients,
n,
R 2
and
RSS
(use
Question
1
from
Tutorial
3
as
a
template).
(3
marks)
2
d. Do
the
coefficients
have
the
signs
that
you
expected?
Make
sure
to
explain
if
you
felt
any
of
the
signs
could
have
been
ambiguous.
(8
marks)
e. By
comparing
just
the
estimated
coefficients,
could
you
say
that
the
mean
number
of
rooms
has
a
stronger
influence
on
the
mean
market
price
for
houses
than
the
percentage
of
municipal
tax
revenue
that
comes
from
commercial
property?
(4
marks)
3
f. Predict
the
mean
market
price
for
houses
for
the
following
two
scenarios
and
comment
on
the
difference.
(4
marks)
Scenario
1:
X1 = 0.225 ,
X 2 = 5.718 ,
X3 = 0.036
and
X 4 = 3678
Scenario
2:
X1 = 0.099 ,
X 2 = 5.237 ,
X3 = 0.018
and
X 4 = 3044
g. Develop
and
test
appropriate
hypotheses
about
the
individual
slope
coefficients
at
the
10
percent
significance
level.
(10
marks)
4
h. Develop
and
test
an
appropriate
hypothesis
about
the
overall
fit
of
the
model
at
the
5
percent
significance
level.
(5
marks)
i. Suppose
that
all
the
“classical
assumptions”
hold.
Does
this
mean
that
the
true
marginal
effect
of
a
one-‐unit
increase
in
the
mean
number
of
rooms
in
houses
is
50
(ie,
$50
000)?
Explain.
(4
marks)
j. The
value
of
R 2
is
relatively
low.
Does
this
mean
that
the
regression
model
cannot
be
used
to
predict
mean
house
prices
for
a
given
city?
Why
or
why
not?
(4
marks)
5
2.
(30
marks)
An
agent
wants
to
find
the
relationship
between
the
number
of
weeks
houses
are
on
the
market
prior
to
sale,
Y,
and
the
difference
between
the
asking
price
and
the
municipal
taxable
value
(in
thousands
of
dollars),
X.
The
following
basic
statistics
based
were
calculated
for
14
observations:
∑Y = 168
i ∑ X = 1624
i ∑Y i
2
= 2436.02
∑ X = 211856
i
2
∑ X Y = 22131.7
i i ∑e 2
i = 122.255
a. Calculate
the
estimated
intercept
coefficient,
β̂0 ,
and
slope
coefficient,
β̂1 .
(6
marks)
b. Write
down
the
estimated
regression
equation
and
interpret
the
value
of
the
slope
coefficient.
(4
marks)
c. Calculate
the
estimated
error
variance,
σˆ 2 .
(3
marks)
6
d. Calculate
the
coefficient
of
determination,
R 2 .
Briefly
explain
what
it
means.
(5
marks)
e. Calculate
the
estimated
standard
errors
for
the
intercept
coefficient,
sβ̂ ,
and
the
slope
0
1
coefficient,
sβ̂ .
(Hint:
sβ̂ = sβ̂ ∑ Xi2 .)
(4
marks)
1 0 1
n
f. Develop
and
test
an
appropriate
hypothesis
about
the
slope
coefficient
at
the
5
percent
significance
level.
(4
marks)
g. Develop
a
95-‐percent
confidence
interval
for
the
true
value
of
β1 .
(4
marks)
7
3.
(20
marks)
The
following
table
comes
from
a
multiple
regression
model
that
was
run
on
Excel.
An
avionics
manufacturer
wants
to
establish
the
relationship
between
the
monthly
sales
of
navigation
units,
Y,
and
the
number
of
ads
placed
in
monthly
magazines,
X1,
and
the
interest
rate
charged
on
credit
customers,
X2.
a. Write down the formulae for each of the highlighted cells.
(8
marks)
8
b. Calculate the values for each of the highlighted cells. (8 marks)