ASSIGNMENT 4:
REGRESSION ANALYSIS
DUE DATE IS APRIL 17, 2024, AT 11:59 PM.
THE EXAM IS ASYNCHRONOUS.
THE EXAM IS AVAILABLE APRIL 7, 2024 AT 11:59 PM.
IT IS AN INDIVIDUAL ASSIGNMENT, WORTH 40% OF YOUR OVERALL
GRADE. DOWNLOAD THE ASSIGNMENT QUESTIONS AND THE DATASETS
FORM THE ASSIGNMENT SECTION ON LEARN@SENECA (BLACKBOARD).
GETTING ASSISTANCE ON THIS ASSESSMENT CONSTITUTES ACADEMIC
MISCONDUCT.
IF YOU DO NOT COMPLETE THE ASSIGNMENT BY THE DUE DATE, OR YOU
EXCEED THE MAXIMUM ALLOTTED TIME, YOU WILL GET ZERO.
YOU SHOULD SUBMIT YOUR EXAM ON LEARN@SENECA (BLACKBOARD)
UNDER THE ASSIGNMENT SECTION.
THERE ARE THREE PROBLEMS. THERE IS NO PENALTY FOR INCORRECT
ANSWERS.
IF YOU HAVE QUESTIONS, YOU CAN SEND AN EMAIL TO ME.
ANALYSIS SHOULD BE DONE WITH SAS.
Page 2 of 8
Problem 1 (20 marks) File: bikes_sharing.csv
This dataset was generated by a bicycle sharing system which is part of the Capital Bikeshare
program in Washington, D.C. These systems allow people to rent a bike from one location and
return it to a different location on an as-needed basis. The process of membership, rental, and bike
return is automated via a network of kiosks located throughout the city. The dataset contains
information such as the duration of travel, departure location, arrival location, the time elapsed, etc.
Attribute Information:
- datetime : date
- season : season (1:winter, 2:spring, 3:summer, 4:fall)
- holiday : weather day is holiday or not (extracted from [Web Link])
- workingday : if day is neither weekend nor holiday is 1, otherwise is 0.
- weather :
o 1: Clear, Few clouds, Partly cloudy, Partly cloudy
o 2: Mist + Cloudy, Mist + Broken clouds, Mist + Few clouds, Mist
o 3: Light Snow, Light Rain + Thunderstorm + Scattered clouds, Light Rain + Scattered
clouds
o 4: Heavy Rain + Ice Pallets + Thunderstorm + Mist, Snow + Fog
- temp : Normalized temperature in Celsius. The values are derived via (t-t_min)/(t_max-
t_min), t_min=-8, t_max=+39 (only in hourly scale)
- atemp: Normalized feeling temperature in Celsius. The values are derived via (t-
t_min)/(t_max-t_min), t_min=-16, t_max=+50 (only in hourly scale)
- humidity: Normalized humidity. The values are divided to 100 (max)
- windspeed: Normalized wind speed. The values are divided to 67 (max)
- casual: count of casual users
- registered: count of registered users
- count: count of total rental bikes including both casual and registered
Page 3 of 8
Goal: to predict the number of bikes based on other variables
a. Find a multiple regression model for the data.
b. Interpret the values of the coefficients in the model.
c. Test whether the model as a whole is significant. At the 0.05 level of significance,
what is your conclusion?
d. Plot the residuals versus the actual values. Do you think that the model does a
good job of predicting number of bikes? Why or why not?
e. Find and interpret the value of R2 for this model.
f. Do you think that this model will be useful in helping the planners? Why or why
not?
g. Test the individual regression coefficients. At the 0.05 level of significance, what
are your conclusions?
h. If you were going to drop just one variable from the model, which one would you
choose? Why?
i. Use stepwise regression to find the best model for the data.
j. Analyze the model you have identified to determine whether it has any problems.
k. Write a memo reporting your findings to your boss. Identify the strengths and
weaknesses of the model you have chosen
Page 4 of 8
Problem 2 (14 marks) File: Titanic. xlsx
The data set contains personal information for 891 passengers, including an indicator variable for
their survival, and the objective is to predict survival, or probability thereof, from the other
characteristics. The survival data for all passengers is stored in the binary variable called Survived.
The predictors include Sex (modeled with male/female dummy variables), Age (and additional
dummy variables for ranges), Class (first, second, or third, modeled with dummy
variables), SiblingSpouse (number of siblings and spouses accompanying the passenger, and
corresponding dummy variables), ParentChild (number of parents and children accompanying the
passenger, and corresponding dummy variables), and Embarked (ports of Cherbourg,
QueensTown, and Southampton, modeled by dummy variables)
a. Write the logistic regression equation relating Age and Survived.
b. For the Titanic data, use SAS to compute the estimated logistic regression equation.
c. Estimate the probability of surviving the passenger with the average Age 30.
d. Suppose we want to check who have a 0.50 or higher probability of surviving. What is the
average age to achieve this level of probability?
e. What is the estimated odds ratio? What is the interpretation?
Page 5 of 8
Problem 3 (6 marks): Capital punishment
I ran two models between race and capital punishment. However, I recoded race two different
ways:
Model 1: White defendants coded as 0, Black defendants coded as 1
Model 2: Black defendants coded as 0, White defendants coded as 1
This gave me the following results:
Model 1 Model 2
Coefficient -1.081 1.081
Odds for white defendants 2.472 2.472
Odds for black defendants 0.838 0.838
Odds ratio 0.34 2.95
a. Why the odds ratios are different? Explain it
b. Show the relation between the odd ratios and coefficient .
Page 6 of 8
Exepectations 4 (Exemplary) 3 (Accomplished) 2 (Developing)
Explanation and There is a single focal question There is a single focal question There is a single focal
justification of that is testable given the data. that is testable given the data. question that is testable given
question(s) There are additional questions The student has described the the data. The student has
that are subsets or follow ups of rationale behind the question, described the rationale
the focal question. These providing context for how they behind the question.
questions will also be testable came up with this question.
given the data. The student has
described the rationale behind
the question, providing context
for how they came up with this
question.
Description of workflow The student provides a verbal The student provides a verbal The student provides a verbal
description of the workflow description of the workflow description of the workflow
used to answer their question. used to answer their question. of the analyses used to
What steps did they take to What steps did they take to answer their question.
answer their question. This can answer their question? This can
include everything from data include everything from data
tidying to visualization to tidying to visualization to
analysis. Justification for the analysis.
workflow is included in this
description.
Selection and The statistical methods used are The statistical methods used are The statistical methods used
justification of statistical appropriate and answer the appropriate and answer the are appropriate and answer
methods question posed. A justification question posed. The student has the question posed.
of statistical method choice clearly described the
includes a clear statement of the assumptions of the test or tests
underlying model used, with a that they used and provided
description of the data support that these assumptions
generating process and error have been met (e.g., verbal
generating process. The student descriptions, additional
has clearly described the statistical tests, data
assumptions of the test or tests visualisations).
that they used and provided
support that these assumptions
have been met (e.g., verbal
descriptions, additional
Page 7 of 8
Exepectations 4 (Exemplary) 3 (Accomplished) 2 (Developing)
statistical tests, data
visualisations).
Code quality The SAS code is easy to read NA NA.
and well commented, allowing
an external reviewer to
understand why certain steps are
being done.
Discussion/evaluation The conclusions are derived The conclusions are derived The conclusions are derived
of results logically from the results and logically from the results and logically from the results and
data visualisations. The student data visualisations. The student data visualisations.
has examined and evaluated has multiple conclusions drawn
limitations in the analysis and from analyses, but does not
has proposed ways to overcome bring them together into a
these limitations. The student is single strong point.
able to synthesize multiple
different results into a single
strong conclusion.
Quality of writing Spelling and grammar count. Spelling and grammar count. Spelling and
Sections written so that they can Sections written so that they grammar count.
be clearly understood by the can be clearly understood by Sections written
reader. Student’s prose flow the reader. Clarity of so that they can
cleanly and clearly. Sentences communication is key. be clearly
are complete, organized, and are understood by
easy to understand. Clarity of the reader.
communication is key.
Page 8 of 8