(a) The following regression equation is estimated as a production function:
. The sample size is 60. Test the following hypotheses at the 5% level of significance:
(ii) There are constant returns to scale.
(b) The following estimated equation was obtained by ordinary least squares regression using quarterly data for 1980 to 1999 inclusive (T = 80):
Standard errors are in parentheses, the explained sum of squares was 122.5, and the residual sum of squares was 20.5.
QUESTION TWO (21 MARKS)
Suppose that you are given two sets of samples with the following information:
(a) Estimate a linear regression equation for each sample separately and for the pooled sample.
(b) State the assumptions under which estimation of the pooled regression is valid. (c) Explain how you will test the validity of these assumptions using the data provided.
QUESTION THREE (9 MARKS)
(a) A study on unemployment in the British interwar period produced the following regression equation:
Sample period 1920 – 1938 (n = 19).
U = unemployment rate
B/W = ratio of unemployment benefits to average wage
Q = actual output
Q* = trend predicted output
log Q - log Q* = captures unexpected changes in aggregate demand.
The authors (Benjamin and Kochin, 1979) conclude that the high benefit levels are partly responsible for the high rates of unemployment. Critics of this study argued that, when the single observation for 1920 is dropped, the results change dramatically (see Ormerod and Worswick, 1982). The equation is now
Sample period 1921-1938 (n = 18).
Test whether the results are significantly different from each other.
Test at 5% level of significance. [5 marks]
(b) An econometrician wishes to test heteroscedasticity in the data he has collected for estimating a regression model. He decides to use Goldfeld-Quandt test and generates the following residual sum of squares (RSS).
RSS1 = Based on the first 30 observations
= 55, df = 25
RSS2 = Based on the last 30 observations
= 140, df = 25
Using a 5% level of significance, determine whether or not heteroscedasticity is present. State the alternatives, decision rule, and conclusion. [4 marks]
QUESTION FOUR (16 MARKS)
Explain the following tests for homoscedasticity:
(a) Goldfeld and Quandt’s test.
(b) Breusch – Pagan - Godfrey test.
Illustrate each of these tests with the data in the Table1 below:
N |
Price |
WL |
DA |
D75 |
A |
MO |
1 |
4556 |
1.00 |
12.1 |
4.9 |
36.0 |
34 |
2 |
4236 |
1.00 |
12.1 |
4.9 |
38.2 |
31 |
3 |
45952 |
1.00 |
12.0 |
4.9 |
21.0 |
16 |
4 |
6000 |
0.00 |
16.0 |
1.2 |
40.0 |
45 |
5 |
2750 |
0.00 |
15.5 |
3.2 |
40.0 |
44 |
6 |
6000 |
0.00 |
13.7 |
3.2 |
20.0 |
26 |
7 |
4952 |
0.00 |
14.5 |
2.5 |
21.0 |
25 |
8 |
1009 |
0.00 |
16.1 |
0.1 |
656.0 |
20 |
9 |
1583 |
1.00 |
15.2 |
3.0 |
60.0 |
19 |
10 |
1449 |
0.00 |
15.5 |
1.0 |
156.0 |
19 |
11 |
2500 |
0.50 |
15.2 |
2.0 |
40.0 |
3 |
12 |
2000 |
0.00 |
15.5 |
3.2 |
13.0 |
3 |
13 |
3704 |
0.00 |
13.5 |
2.5 |
27.0 |
3 |
14 |
3500 |
0.00 |
15.5 |
1.0 |
10.0 |
3 |
15 |
3500 |
0.00 |
17.5 |
5.4 |
20.0 |
38 |
16 |
4537 |
1.00 |
18.0 |
5.9 |
38.0 |
24 |
17 |
2700 |
0.00 |
17.2 |
5.1 |
5.0 |
3 |
18 |
1020 |
1.00 |
34.2 |
22.0 |
5.0 |
27 |
19 |
4000 |
0.00 |
11.1 |
5.1 |
3.5 |
13 |
20 |
3764 |
0.00 |
14.2 |
2.0 |
237.6 |
40 |
21 |
871 |
1.00 |
14.2 |
2.0 |
237.6 |
7 |
22 |
2500 |
1.00 |
11.1 |
3.1 |
20.0 |
41 |
23 |
14200 |
1.00 |
14.7 |
2.4 |
5.0 |
36 |
24 |
5767 |
0.00 |
12.1 |
4.1 |
30.0 |
22 |
25 |
15316 |
1.00 |
14.8 |
2.5 |
3.8 |
21 |
26 |
8873 |
1.00 |
14.8 |
2.5 |
7.9 |
17 |
27 |
5175 |
0.25 |
14.2 |
2.0 |
40.0 |
13 |
28 |
3977 |
0.00 |
11.4 |
2.9 |
8.8 |
10 |
29 |
5500 |
0.20 |
18.5 |
5.9 |
10.0 |
38 |
30 |
7500 |
0.00 |
16.5 |
3.9 |
8.0 |
42 |
31 |
4545 |
1.00 |
16.8 |
4.5 |
97.0 |
36 |
32 |
3765 |
0.72 |
18.7 |
6.4 |
178.0 |
29 |
33 |
5000 |
1.00 |
18.4 |
6.1 |
10.3 |
25 |
34 |
3300 |
0.00 |
16.2 |
4.0 |
525.7 |
25 |
35 |
5500 |
0.00 |
18.0 |
5.4 |
6.0 |
21 |
36 |
5172 |
0.00 |
15.0 |
2.4 |
29.0 |
20 |
37 |
3571 |
0.00 |
15.1 |
2.5 |
21.0 |
20 |
38 |
4000 |
0.00 |
18.2 |
6.0 |
10.0 |
15 |
39 |
4000 |
0.00 |
18.4 |
6.1 |
15.0 |
18 |
40 |
2625 |
0.00 |
15.5 |
2.9 |
80.0 |
10 |
41 |
2257 |
0.00 |
42.8 |
30.5 |
171.0 |
14 |
42 |
15504 |
0.00 |
4.0 |
4.5 |
38.7 |
39 |
43 |
5600 |
0.00 |
3.8 |
4.0 |
30.0 |
25 |
44 |
8000 |
0.00 |
3.5 |
4.2 |
30.0 |
27 |
45 |
7700 |
0.00 |
4.0 |
3.8 |
15.0 |
46 |
46 |
6187 |
0.00 |
4.2 |
5.0 |
69.5 |
15 |
47 |
7018 |
0.00 |
3.5 |
4.0 |
10.9 |
23 |
48 |
4821 |
0.60 |
7.8 |
2.7 |
224.0 |
33 |
49 |
6504 |
0.00 |
14.9 |
4.8 |
6.4 |
40 |
50 |
5225 |
0.00 |
16.2 |
5.0 |
10.0 |
40 |
The variables are as follows:
Price = observed land price per acre excluding improvements
WL = proportion of acreage that is wooded
DA = distance from parcel to Sarasota airport
D75 = distance from parcel to I-75
A = acreage of parcel
MO = month in which the parcel was sold.
QUESTION FIVE (5 MARKS)
A study on unemployment in the British interwar period produced the following data
given in the Table 2. below:
Year |
Weekly Wages, W (s.) |
Weekly Benefits, B (s.) |
Unemployment Rate, U (%) |
Benefits/ Wages |
NNP, Qa (£ million at 1938 factor cost) |
1920 |
73.8 |
11.3 |
3.9 |
0.15 |
3426 |
1921 |
70.6 |
16.83 |
17.0 |
0.24 |
3242 |
1922 |
59.1 |
22.00 |
14.3 |
0.37 |
3384 |
1923 |
55.5 |
22.00 |
11.7 |
0.40 |
3514 |
1924 |
56.0 |
23.67 |
10.3 |
0.42 |
3622 |
1925 |
56.4 |
27.00 |
11.3 |
0.48 |
3840 |
1926 |
55.8 |
27.00 |
12.5 |
0.48 |
3656 |
1927 |
56.2 |
27.00 |
9.7 |
0.48 |
3937 |
1928 |
55.7 |
27.67 |
10.8 |
0.50 |
4003 |
1929 |
55.8 |
28.00 |
10.4 |
0.50 |
4097 |
1930 |
55.7 |
29.50 |
16.1 |
0.53 |
4082 |
1931 |
54.9 |
29.54 |
21.3 |
0.54 |
3832 |
1932 |
54.0 |
27.25 |
22.1 |
0.50 |
3828 |
1933 |
53.7 |
27.25 |
19.9 |
0.51 |
3899 |
1934 |
54.3 |
28.6 |
16.7 |
0.53 |
4196 |
1935 |
55.0 |
30.3 |
15.5 |
0.55 |
4365 |
The assessment, CAT 1, focuses on applied econometrics and production/operations management concepts. The main objectives include:
Regression Analysis:
Estimating production functions using sample data.
Testing hypotheses such as constant returns to scale.
Determining the significance of slope coefficients and calculating R2R^2R2 and adjusted R2R^2R2.
Pooled vs. Separate Regression:
Estimating regression equations for individual samples and pooled datasets.
Identifying assumptions required for pooled regressions and testing their validity.
Time Series & Unemployment Analysis:
Examining historical unemployment data to test differences in regression results.
Assessing the impact of specific variables like benefit levels on unemployment.
Heteroscedasticity Testing:
Applying Goldfeld-Quandt test for residual variance differences.
Understanding the Breusch-Pagan-Godfrey method for homoscedasticity assessment.
Serial Correlation:
Using Durbin-Watson test to detect autocorrelation in regression residuals.
The assessment tests the student’s ability to interpret, analyze, and apply econometric techniques in real-world and historical datasets.
The Academic mentor guided the student step by step, using a structured approach for each section:
Question One – Regression and Production Function:
Explained the formulation of the production function regression.
Guided the student in conducting hypothesis testing for constant returns to scale.
Showed the method to test slope coefficients for significance using t-tests.
Demonstrated calculation of R2R^2R2 and adjusted R2R^2R2 to evaluate model fit.
Question Two – Pooled and Separate Regression:
Taught the student how to estimate linear regressions for individual samples and pooled data.
Discussed assumptions like homogeneity of coefficients and error term independence.
Explained testing validity via F-tests and residual analysis.
Question Three – Unemployment Analysis:
Compared regression results for two periods (1920–1938 vs. 1921–1938).
Conducted a significance test to assess differences.
Explained the use of Goldfeld-Quandt test for heteroscedasticity, including stating null and alternative hypotheses, decision rules, and interpreting results.
Question Four – Homoscedasticity Tests:
Introduced Goldfeld-Quandt and Breusch-Pagan-Godfrey tests.
Walked through applying these tests to sample datasets to detect unequal variance in residuals.
Highlighted the practical implications for regression analysis and model reliability.
Question Five – Serial Correlation Test:
Explained the Durbin-Watson test for detecting autocorrelation in time-series data.
Demonstrated how to interpret the statistic to conclude presence or absence of serial correlation.
Through this guided approach, the student was able to:
Develop confidence in formulating and testing econometric models.
Apply regression analysis techniques to practical and historical datasets.
Understand assumptions behind pooled regressions, homoscedasticity, and serial correlation.
Learn stepwise interpretation of statistical tests, including t-tests, F-tests, Goldfeld-Quandt, Breusch-Pagan, and Durbin-Watson.
Achieve hands-on exposure in analyzing production, unemployment, and land price datasets.
Strengthen critical thinking to draw conclusions from statistical outcomes.
Final Outcome: The student successfully completed CAT 1, producing a comprehensive regression analysis, testing assumptions, detecting heteroscedasticity and serial correlation, and correctly interpreting all econometric results. The process also reinforced theoretical understanding and practical skills in applied econometrics.
Looking for guidance on your assignment? You can download our sample solution to understand the structure, approach, and methodology used by professional academic writers. Important: This sample is strictly for reference purposes. Submitting it as your own work may result in plagiarism issues.
Want a 100% original, plagiarism-free solution tailored to your requirements? Our team of experienced academic writers can create a fresh assignment solution for you, ensuring top-quality content, proper formatting, and in-depth research. Ordering a custom solution helps you:
© Copyright 2025 My Uni Papers – Student Hustle Made Hassle Free. All rights reserved.