Sunteți pe pagina 1din 5

Assignment 2

Course: QTMS
Submitted By: Zoya palijo (8211)
Submitted to: Dr. Arsalan Hashmi

1|Page
Question 1: The following regression results are based on a study of labour turnover in the
United States for the period 1950Q1 to 1979Q4.

Where
Y= quit rate in manufacturing, defined as the number of people leaving jobs voluntarily per 100
employees.
X2= an instrumental or proxy variable for adult male unemployment rate
X3= percentage of employees younger than 25
X4= ratio of manufacturing employment in quarter (t-1) to that in quarter (t-4)
X5= percentage of women employees
X6= time trend (1950Q1=1)
Part a
Interpret the regression coefficients and comment on their statistical significance.
Interpretation of constant:
The estimated value of intercept is 4.47. This means that when all the IVs are 0. The number of
people leaving job voluntary per 100 employees will be 4.47.
Interpretation of slope coefficients:
(Part b) Is the negative relationship between the logs of Y and X2 justifiable? Explain your
answer.
For X2: the estimated value of slope coefficient is -0.34. This value implies that unemployment
increases by 1% the Quit rate will decrease by -0.34%, holding other variable constant. This
makes sense.
For X3: the estimated value of slope coefficient is 1.22. This value implies that if the number of
young employees increases by 1% the quit rate will increase by 1.22%, holding other factors
constant. This makes sense because the young people tend to look for new opportunities.
For X4: the estimated value of slope is 1.22. This value implies that if the manufacturing
employment rate increases by 1% the quit rate will increase by 1.22%, holding other factors
constant. This makes sense if there is the boom in the economy quit rate can be increased, so that
people look for new opportunities. But if there is recession the quit rate can decrease.

2|Page
For X5: the estimated value of slope 0.80. This value implies that if the percentage of women
employees will increase be 1% the quit rate will also increase by 0.80% holding other variables
constant. It Makes sense.
For X6:
Part c
Comment about the goodness of fit of the model.
Goodness of fit of model:
The model has the 0.5370 value of R squared. That means 53.70% variation of dependent
variable has been explained by the model. Model has a normal good fit.
Part d
Estimate the standard errors of the regression coefficients.
For X2
T= -5.31 B2= -0.34
SeB2= -0.34/-5.31 = 0.064
For X3
T= 3.64 B3= 1.22
SeB3= 1.22/3.64 = 0.335
For X4
T= 3.10 B4= 1.22
SeB4= 1.22/3.10 = 0.393
For X5
T= 1.10 B5= 0.80
SeB5= 0.80/1.10 = 0.727
For X6
T= -3.09 B6= 0.005
SeB6= 0.005/-3.09 = -1.77x10^-3

Part e
Estimate the value of r-squared and calculate the F-statistic. What can you conclude about
the overall significance of the model?
Adjusted R-Squared = 1-(1-R2) (n-1/n-k)
0.5370= 1-(1-R2) (119/114)
0.5370= 1- (1-R2) (1.0438)
0.5370= 1- 1.0438 + 1.0438R2
0.5370 + 1.0438 -1 = 1.0438R2
0.5808/1.0438 = R2
R2= 0.5564
F- statistics= R2/(k-1) ÷ (1-R2)/(n-k)
= 0.5564/5 ÷ (1-0.5564)/(114)
= 0.11128/ (0.4436)/114

3|Page
= 0.1128/ 0.0638
= 29.28
We reject the null hypothesis at 5% level of significance because the critical value is 2.29. It
shows that this model is overall significant.

Question 2: The data from 46 states of the United States was used to estimate the following
statistical results.

Where,
C= cigarette consumption, packs per year
P= real price per pack
Y= real disposable income per capita

Part a
Interpret the regression results. Do they make logical sense?
Interpretation of coefficient:
For P:
The estimated value is -1.34. This means if the price per pack increases by 1% the cigarette
consumption will decrease by 1.34% holding other factors constant. It makes logical sense as the
price will increase, consumer will buy less.
For Y:
The value is 0.17. This means if the income per capita will increase by 1% consumption will also
increase by 0.17% holding other factors constant. It makes logical sense as the income will
increase consumer’s buying power will also increase and they will consume more.
Part b
Are the regression coefficients statistically different from one?
T- Test For P (df= n-k = 46-3 =43, level of significance= 5%)
T = B2-B0/SeB2
= -1.34/0.32
= -4.1875
Critical value= + 2.021, -2.021
Hence -4.1875 lie in the Rejection region so it means they are statistically significant.
T- Test For Y (df= n-k = 46-3 =43, level of significance= 5%)
T = B3-B0/SeB3

4|Page
= 0.17/0.20
= 0.85
Critical value= + 2.021, -2.021
Hence 0.85 lie in the Acceptance region so it means they are not statistically significant.

Part c
Compute the value of r-squared and test the overall significance of the model.
Adjusted R-Squared = 1-(1-R2) (n-1/n-k) F- statistics= R2/(k-1) ÷ (1-R2)/(n-k)
0.27= 1- (1-R2) (46-1)(46-3) = 0.2141/2 ÷ 0.7859/43
0.27= 1- (1-R2)(1.046) = 0.10705/0.0182
0.27= 1- 1.046 + 1.046R2 =5.881
0.27= 0.046 + 1.046R2 N= 2, D= 43, α= 5%
0.27 - 0.046 = 1.046R2 Critical value= 3.23
0.224/1.046 = R2 Hence the null hypothesis is rejected at 5%
R2= 0.2141 level of significance because the value 5.881
lie in the rejection region and it shows that the
overall model is significant.

5|Page

S-ar putea să vă placă și