Professional Documents
Culture Documents
Faculty of Computer and Mathematical Sciences Bachelor of Science (Hons.) Actuarial Science
Faculty of Computer and Mathematical Sciences Bachelor of Science (Hons.) Actuarial Science
Faculty of Computer and Mathematical Sciences Bachelor of Science (Hons.) Actuarial Science
PREPARED FOR:
PREPARED BY:
Turtle Pizza, Sdn Bhd, is a regional franchisor of home delivery pizza outlets. To better effects.
To assess the effects of price, advertising, and income on pizza demand, the company has
conducted and compiled the following survey of monthly sales and operating data at 15 of the
company’s outlet in Malaysia:
Advertising
Disposable Income
Quantity Sold Price Expenditures
per Household
per Month
i. Objective
The regression analysis is trying to establish a functional relationship between Quantity
Sold (Yt) with Price (X1), Advertising Expenditure (X2) and Disposable Income (X3).
iii. Economic Theory: All signs of coefficient conform to the economic theory
̂
𝑌̂𝑡 = −33301.728 ̂ . 𝑋1 + 1.454
− 4041.534 ̂ . 𝑋2 + 1.528
̂ . 𝑋3
v. R-square
R² = 0.950
It shows that 95.0% of the changes in the dependent variable (Quantity Sold) can be
explained by the independent variables in the equation, i.e Price, Advertising
Expenditure and Disposable Income. But the other 5.0% cannot be explained by the
regression due to the omission of some other important independent variables.
vii. t-statistics
Hypothesis testing:
H₀: The variable is not significant in explaining the quantity of pizza sold.
df = n − k − 1
= 15 − 3 − 1
= 11
t-critical value:
SEE = 3333.092
= Ŷ ± [(t₁2,₀.₀₅) × SEE]
= Ŷ ± (2.179 × 3333.092)
= Ŷ ± (7262.8075)
ix. F-statistics
Hypothesis testing:
H₀: There is no significant relationship between the quantity sold with the price,
advertising expenditure and disposable income.
H₁: There is a significant relationship between the quantity sold with the price,
advertising expenditure and disposable income.
F-critical value:
F₂,₁₄,₀.₀₅ = 3.74
Conclusion:
Since F-statistic value = 69.175 is greater than F-critical value = 3.74, there is a
significant relationship between the quantity sold with the price, advertising
expenditure and disposable income.
x. Durbin Watson
Since the value of Durbin-Watson is not in between the interval 1.8 to 2.2, it shows that
there is an autocorrelation problem that can arise in the regression analysis with time
series data. There are 3 possibilities where autocorrelation or multicollinearity
problems can arise in the regression analysis, i.e. when independent variables are
interrelated or duplicated, where independent variables have been miss-specified, and
where important independent variables are found missing.
b) Based on the analysis, do you think the above estimated equation can be acceptable?
Yes, the above estimated equation is acceptable because all independent variables within
the equation conform with the economic theory. The independent variables also have
significant relationships with dependent variables, in which price (X₁), advertising
expenditure (X₂), and disposable income (X3) are significant in explaining the changes in
the quantity sold (Yₜ).
QUESTION 2
Given the data on electric power consumption (in billions of kilowatt-hours). GNP (in billions
of dollars), and electricity prices (in cents per kilowatt-hour) for period 1998 – 2012 are shown
here.
i. The objective
Yₜ = c + b₁X₁ + b₂X₂
v. R-square
R² = 0.945
It shows that 94.5 % of the changes in the dependent variable (Electric Power
Consumption) can be explained by the independent variables in the equation, i.e GNP
and Electricity Price. But the other 5.5% cannot be explained by the regression due to
the omission of some other important independent variables.
Hypothesis testing:
H₀: The variable is not significant in explaining the electric power consumption.
df = n − k − 1
= 15 − 2 − 1
= 12
t-critical value:
SEE = 28.6651
= Ŷ ± [(t₁₃,₀.₀₅) × SEE]
= Ŷ ± (2.16 × 28.6651)
= Ŷ ± (61.9166)
ix. F-statistics
Hypothesis testing:
H₀: There is no significant relationship between the electric power consumption with
the GNP and electricity price.
H₁: There is a significant relationship between the electric power consumption with the
GNP and electricity price.
F-critical value:
F₂,₁₄,₀.₀₅ = 3.74
Conclusion:
Since F-statistic value = 103.276 is greater than F-critical value = 3.74, there is a
significant relationship between the electric power consumption with the GNP and
electricity price.
x. Durbin Watson
b) Based on the analysis, do you think the above estimated equation can be acceptable?
Yes, the above estimated equation is acceptable because all independent variables within
the equation conforms with the economic theory. The independent variables also have
significant relationships with dependent variables, in which GNP (X₁) and electricity price
(X₂) are significant in explaining the changes in the electric power consumption (Yₜ).
c) In 2013, GNP was $3,661.3 billion and the price of electricity was 7.16 cents per
kilowatt-hour. Use the estimating equation from part (a) to predict electricity
consumption for 2013.
Ŷ = 861.60898
Hence, the predicted electricity consumption for 2013 is 861.609 billion kilowatt-hours.
APPENDICES