Scribd
Upload
7 views

Course10 Exercises

Uploaded by

wtjqkgbn85
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
7 views

Course10 Exercises

Uploaded by

wtjqkgbn85
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 4

Theme 8 : Simple Linear Regression

Exercises

Exercise 1
We are performing a study on household debt in a certain region. The debt is measured by the total
balance of all loans (from banks, credit unions, credit cards, stores, etc.). To explain the total balance of
the loans (Y variable), we will proceed with a regression analysis, using weekly household income as the
independent variable X. Following a survey of 20 households in the region, we gather the following data:

Househol Balanc Weekly


d e Income
Number Y X
1 0 380
2 3 384 735
3 3 998 489
4 5 975 640
5 32 665 1 250
6 8 380 707
7 4 307 450
8 4 016 560
9 5 670 454
10 12 925 509
11 16 975 670
12 1 222 301
13 15 125 1 037
14 908 300
15 0 435
16 7 211 574
17 9 300 574
18 3 574 495
19 4 000 516
20 16 000 1 660

a) Based on this table, use Excel to calculate the following quantities:

∑ xi =
∑ yi =
∑ xi 2 =
∑ yi 2 =
∑ xiyi =
b) Using the formulas seen in student material, calculate

SST =
SSR =
SSE=

c) At a significance level of 0.01, can we confirm that the variable X significantly explains Y?

d) If a household has a weekly revenue of 100$ more than some other chosen household, by how
much would we expect that the total balance of their loans will be higher or lower, on average, in
relation to the other household?
e) If we survey a household other than one of the 20 in the sample, and we find that this household’s
weekly revenue is 800$, estimate the total balance of the loans of this household using a 95%
confidence interval.

Exercise 2

The manager of a bank wishes to determine whether there is a relationship between the annual income of
a family (independent variable X) and the amount of money that they allocate to savings (dependent
variable Y). In a random sample of 10 families, we obtain the following results:

Slope of the regression line = 0.132


y-intercept of the regression line = -0.854
Standard error on B0: sb0 = 0.371
Standard error on B1: sb1 = 0.026
Standard error of residues: se = 0.20777
Mean income of the sample: x = 13 800
1. Construct a 95% confidence interval for the savings of a new family (i.e. one which was not part of
the sample) with an income of 11 000 $.
2. Test whether the regression is significant at a level of α = 5%.
Exercise 3
A company has logged data concerning the daily demand for their product “Y” (in thousands), as well as
the unit price “X” in hundreds of dollars over a period of 11 days. The information collected is
summarized below:
X = 154 X2 = 2 586
Y = 451 Y2 = 18 901
XY = 5 930 SSE = 67.08

a) What is the regression line estimated by the method of least squares?

b) Calculate the coefficient of determination.

c) Test whether the slope is significantly different from zero at the level  = 0.05.

d) Is it possible to have a demand of zero? If so, give a pointwise estimate of the price at which this
occurs.

Exercise 4
The following data outlines the results of a regression analysis of spending on advertising (X)
(represented in % of the total operational spending) and the net profit of operations (Y) (stated in % of
total sales), of a random sample of 8 sporting goods stores:
Descriptive Statistics
Standard
Mean Deviation N
Profit 2.8500 1.36382 8
Spending on advertising 1.4625 0.91798 8

se = 0.24159
r2 = 0.973
b0 = 0.707
b1 = 1.466
sb0 = 0.169
sb1 = 0.099

a) What is the equation of the regression line?


b) What is the profit (expressed in % of its sales) that a store of this type can hope to make if it devotes
1.2 % of its total operational spending to advertising?
c) Construct a 95 % confidence interval for the coefficients 0 and 1.
d) Test at a level  = 0.05 whether spending on advertising significantly explain the profits generated by
this type of sporting goods store.
e) Use a 95 % confidence interval to estimate the mean profit (in %) that a store of this type which
devotes 1.8 % of its expenditures to advertising can hope to make.
f) Using a 95% confidence interval, estimate the average profit (in %) that we can expect a specific
sporting goods store of this type will make, if it is a new store (and thus not included in the sample)
and it dedicates 1.8% of its spending to advertising.
Exercise 5
Say that we gather the following data concerning 2 variables from a sample of 15 different four-person
families:
X: net weekly income (in $)
Y: spending on food (in $)
and we obtain the following results:

n
∑ ( y i− ^yi )2=625 , 44
i=1
a) Determine the equation of the regression line for spending on food as a function of the net weekly
income.
b) What is the proportion of the variation of spending on food that is explained by the net weekly
income?
c) Is the regression signification at a level of 5 % ?
d) If the net weekly income were to increase by 25$, by how much would this affect spending on
food?
e) Using a 95% confidence interval, estimate the average spending on food of a family whose net
weekly revenue is 500$.
f) Calculate the coefficient of correlation and interpret it in the context of the problem.

You might also like