Scribd
Upload
47 views

Econometrics Assignment 2

The document contains an econometrics assignment analyzing the relationship between house price, bedrooms, square footage, and age. [1] Regressing price on bedrooms alone yielded a coefficient of $39,700 per extra bedroom, but adding square footage lowered the coefficient significantly to $700 per bedroom. [2] Adding age and its quadratic term to a price regression with square footage improved the R^2 to 0.72. [3] A hypothesis test confirmed that at least one of the age coefficients is statistically significant.

Uploaded by

Jovias Kelvinsun
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
47 views

Econometrics Assignment 2

The document contains an econometrics assignment analyzing the relationship between house price, bedrooms, square footage, and age. [1] Regressing price on bedrooms alone yielded a coefficient of $39,700 per extra bedroom, but adding square footage lowered the coefficient significantly to $700 per bedroom. [2] Adding age and its quadratic term to a price regression with square footage improved the R^2 to 0.72. [3] A hypothesis test confirmed that at least one of the age coefficients is statistically significant.

Uploaded by

Jovias Kelvinsun
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

Econometrics Assignment 2

Jovias Kelvinsun 29438403

a) (A1)

price=3.622+0.397 bed R2=0.30


^

se = (0.658) (0.0203)

b) holding all other regressor constant, with an increase in 1 unit of bedrooms, the price will increase
by (0.397*100) $39 700. The sign of the slope coefficient is expected since usually properties with
more bedrooms are typically more expensive.

c) yes, because houses with more bedrooms means that there are more sqft needed to make those
bedrooms, in other words they are positively correlated. The consequence of not omitting sqft into
the model, would mean that the model is biased and b 1 would be taking the credit of all the
increased sqft of a due to the number of bedrooms in a house.

d) (A2)

price=3.882+0.007 bed+ 0.043 sqft R2=0.637


^

se = (0.048) (0.0199) (0.001)

e) holding all other regressor constant, with an increase in 1 unit of bedrooms, the price will increase
by (0.007*100) $700. This result is now much more different from the first result which was $39 700,
a 39 000 difference.
f) The coefficients model A1&2

A1 is b1= 3.622, b2 = 0.397

A2 is b1= 3.622, b2 = 0.007 and b3 = 0.043

Model A1 is upwardly biased as we can see from including the sqft into the model, b 1’s coefficient
dropped quite significantly. This is due to the fact that more number of bedrooms typically mean a
larger sqft of the house, and most of the price will increase the price much more than the number of
bedrooms.

g) the Pearson correlation coefficient for bedrooms and sqft is 0.6768, which means that it is a
strong positive correlation. Yes, the sign is what is expected, as the more bedrooms there are the
bigger space needed for a house.

h) yes, the correlation proves that model is upwardly biased. From the calculation for model A2, the
R2 is also higher at 0.637 which means that a higher proportion of variance in price can be explained
by the explanatory variables.

Question 2

A) the reason we included is because we want to see whether age has any influence on the size of
the house and hence the price of the house. The quadratic age is the non-linear version

b)

price=4.2003+0.0396 sqft−0.0187 age +0.0002 age2 R2=0. 72


^

se = (0.03) (0.001) (0.001) (2.44E-05)

c)
(restricted model)

1) hypothesis:

H 0=β 3= β 4=0
H 1=β 3 ≠0∨β 4 ≠ 0∨both ≠0

( SSER −SSEU ) / j
2) F= Fdf 1= j ,df 2=(N− K)
SSE U /(N −K)

SSEU =65.68

SSER =85.27

3) rejection region: reject H 0 if F calc > F crit

∝=0.05
F crit =F 0.95,2,896

F crit =3.006

(85.27−65.68)/2
4) test statistic F= =133.62
65.68 /(900−4)

5) conclusion: since F calc > F crit , we reject H0 and conclude that at least one of the β ' s is not zero.

You might also like