Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
2 views

Topic5 Lab TwoDimVariable

The document outlines various exercises related to statistics, focusing on independence and correlation in two-dimensional random variables, stock price changes, service quality assessments, ethical issue identification, and relationships between behaviors and outcomes. Each exercise requires calculations of marginal and conditional distributions, correlation coefficients, and hypothesis testing. The exercises utilize data from different scenarios, including stock market analysis, client service ratings, and ethical issue assessments, to explore statistical relationships and dependencies.

Uploaded by

linhchiemilycb
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
2 views

Topic5 Lab TwoDimVariable

The document outlines various exercises related to statistics, focusing on independence and correlation in two-dimensional random variables, stock price changes, service quality assessments, ethical issue identification, and relationships between behaviors and outcomes. Each exercise requires calculations of marginal and conditional distributions, correlation coefficients, and hypothesis testing. The exercises utilize data from different scenarios, including stock market analysis, client service ratings, and ethical issue assessments, to explore statistical relationships and dependencies.

Uploaded by

linhchiemilycb
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 14

STATISTICS

Laboratory
Topic 5 – Laboratory:
Independence & Correlation
Radosław Antczak
Paweł Strzelecki
Marta Styrc
Lucas Van der Velde

Warsaw School of Economics


Spring Semester 2023
1. TWO DIMENSIONAL RANDOM
VARIABLE AND INDEPENDENCE
Exercise 1.1
Table below presents the distribution of two-dimensional random
variable (X,Y).
Using this distribution please determine:
1) The value of c
2) Marginal distribution functions of X and Y.
3) EX, DX, EY, DY
4) Conditional distributions: X|Y=1 and Y|X=5
5) What conclusions about stochastic relationship can be drawn on
the basis of the calculations above?

Y 1 2 3
X
0 0,1 0,1 c
5 0,2 0,1 0
10 0 0,2 0,1
Exercise 1.2
The investor observed changes of the stock prices of companies just after publication of
the quarterly reports. He wanted to test a hypotesis that the messages from reports in
comparison to analyst’s expectations (results worse than expected, equal to expected or
better than expected) influence the prices. The table below contains information for a
random sample of n=100 companies to:
1) Calculate the marginal distributions of stock prices and messages from the reports in
this exercise.
2) Calculate the conditional distribution of stock price changes under the condition that the
messages from quarterly report suggest that results were better than expected by
analysts.
3) Verify the hypothesis of the investor by performing a statistical test at significance level
0.05.
4) Calculate the measure of the relationship between messages from reports and changes
in the stock prices. What is the interpretation of that measure?
Message from quarterly reports:
Results Results Results
Change in better than equal to worse than
share price: expected expected expected
Increase 12 8 4
No changes 16 20 16
Decrease 6 8 10
Exercise 1.3
A large brokerage firm wants to determine whether the service it
provides to affluent clients differs from the service provided to
lower income clients. A sample of 500 clients are selected and each
client is asked to rate his or her brokers on a scale from 1
(outstanding) to 3 (poor).
a) Compute probabilities of ranking the service as outstanding
conditioned on the level of client affluence. Plot these
probabilities on a bar plot.
b) Do the conditional probabilities computed in a) suggest that the
quality of service depends on customer’s affluence?
c) Assess the scale of dependence between customer income and
the quality of service in the sample using V-Cramer coefficient.
d) Test to determine whether the relationship you found in c) is
significant.
 see file BROKERAGE.xls
Source: J.T. Mc Clave, P.G. Benson,T. Sincich: Statistics for Business and Economics, 9th
Edition, 2005
Exercise 1.4

Source: J.T. Mc Clave, P.G. Benson,T.


Sincich: Statistics for Business and
Economics, 9th Edition, 2005
Exercise 1.5

a) Compute the probabilities of identifying an ethical issue under the condition that
the ethical issue was a) moderate, b) severe. Plot the probabilities.
b) Compare the probabilities computed in a). Does it seem that the severity of ethical
issue affects whether the issue is identified by auditors?
c) Compute the V-Cramer coefficient to verify the scale of dependence in the sample.
d) Perform a test to verify whether the severity of ethical issue affects whether the
issue is identified by auditors.
e) Keeping the same sample size change the numbers in contingency table so that the
answer you would get for the question posed in d) changes.
Exercise 1.6
A teacher wanted to investigate the relationship between cheating during
exams at school and an incidence of conflicts with law in adult life.
According to his survey (n=200) 30% of students admitted that they
cheated during exams and 40 of those students had problems with law
in adult life. In the same survey 100 persons were honest during exams
and experienced no problems with law in their adult life.

On the basis of these descriptions:


1) Create a contingency table
2) Calculate the measure of the stochastic relationship between the
occurrence of cheating during exams and conflicts with law in the future.
3) Check the hypothesis that the two events, cheating during exams and
experience of conflicts with law in adult life, are independent
2. LINEAR AND RANK CORRELATION
Exercise 2.1

Source: J.T. Mc Clave, P.G. Benson,T. Sincich: Statistics for Business and Economics, 9th Edition,
2005
Exercise 2.2
Refer to the DIAMONDS_sample.xls file. This file contains data on the
number of carats for a sample of 30 diamonds sold in the open
market and their prices.
a) Construct a scatterplot for these data. Interpret it.
b) Calculate the coefficient of correlation. What is the relationship
between the number of carats and price of the diamonds in the
analysed sample?
c) Is this correlation significant?

 see file DIAMONDS_sample.xls


Exercise 2.3

d. Is the correlation found in b significant at α=0.05?

 see file FORBES400.xls


Source: J.T. Mc Clave, P.G. Benson,T. Sincich: Statistics for Business and Economics, 9th Edition,
2005
Exercise 2.4
Forbes magazine reported financial standing of each team in the
National Football League (NFL). The table in file NFLVALUE.xls
reports the current team value (without deduction for debt) and
operating income for each team for 2004.
a) Construct a scatterplot for these data. Interpret it.
b) Calculate the coefficient of correlation. What was the relationship
between the current team value and operating income in 2004?
Was this correlation significant?
c) Identify the outliers in the dataset. Construct another scatterplot,
excluding the outliers and compute the coefficent of correlation
for this „new”dataset. Test it for the significance. What has
changed? Why?
 see file NFLVALUE.xls
Source: J.T. Mc Clave, P.G. Benson,T. Sincich: Statistics for Business and Economics, 9th Edition,
2005
Exercise 2.5
The representative group of the target population of advertising campaign (focus
group)
assessed the 10 proposals of pictures for the campaign „juicy fruits”. Two rankings
are the
results of this experiment:
1) The ranking of the intensity of colours of the 10 pictures Picture no. 5
2) The ranking of the attractiveness of 10 pictures.
Ranking of Ranking of
colours attractiveness
Picture 1 8 1
Picture 2 7 3
Picture 3 2 4
Picture 4 1 2
Picture 5 3 5
Picture 6 4 8
Picture 7 10 7
Picture 8 6 6
Picture 9 5 9
Picture 10 9 10

Do these results suggest the relationship between colour intensity and the
assessment of the attractiveness of ads? Use this two rankings to calculate the
measure of correlation. What are the extreme values of this measure?

You might also like