ECON 216 Final Project
David Clingingsmith
07 April, 2022
Objective
The objective of the final project for this class is for each group to produce a visualization-based data
analysis. The analysis will demonstrate the students’ familiarity with the concepts learned throughout the
course. The analysis will use data of interest to the group. Students will explore the data and see what can
be learned from it. The visualization-based analysis will report what the group has learned.
The final report will center on a persuasive argument that uses visualizations and accom-
panying text to establish a set of facts that can be learned from the data chosen. This
visualization-based argument is something you could share with others to explain your anal-
ysis and its findings. Note that most of my class presentations are structured around building and
visualizing datasets with this goal in mind. You are therefore being exposed to models for making this kind
of persuasive, visualization-based argument throughout the semester. Since every topic and dataset has its
own peculiarities, it isn’t possible to formulate general rules for how to make such an argument. Learning
by example is best.
In preparing for making that persuasive arugment, the students will conduct and present an exploratory
data analysis that will produce a larger set of visualizations than included in the argument that characterize
their dataset, explore different avenues of analysis, some of which may not pan out, and from which the
finished analysis will draw.
The end product to be handed in will be report whose format is described below. Along the way to producing
the report, groups will complete two check-in assignments and a presentation of their analysis.
Project Steps and Deliverables
There are four deliverables: two check-ins, a presentation, and a final report. These efforts build on each
other. The check-ins deliberately have low point value to allow for improvement toward the
final report. Because of the cumulative nature of the work, the quality of the final, high value outputs of
presentation and report depend a lot on the work done at the check-ins, so you should put more effort in
than the point value suggests.
Feedback
You will get feedback after each check-in to help you improve. Please read it. I expect corrections made at
these stages to be addressed. I will also make suggestions which may be helpful and I will indicate whether
these are things you must do or are optional.
1
Check-in 1: Topic and Dataset (3 points)
The first task each group has is to find a dataset to analyse. A first step might be to have a discussion of
topics that interest each group member to narrow the search. We will discuss places to look for data in a
later class. When looking at a dataset, pay attention to the unit of observation and the variables included
and think about what those variables might tell you, both on their own and in relation to each other. In
class and on DataCamp we learning techniques that will help you.
This is an important step since it determines the rest of the project so should be undertaken with some care.
The check-in is a simple paragraph submitted to Canvas that includes:
1. The name of the dataset and a link to it.
2. The general topic you propose to use the dataset to investigate. Be as specific as you can. You may
change your focus, but you will need to have identified at least one topic at this stage.
3. The unit of observation of the dataset and the variables within it that you think will be useful. Describe
what the variables measure.
This check-in is due February 24th.
Check-in 2: Exploratory Data Analysis (EDA) (6 points)
This is the most time-consuming and coding-intensive part of the project. It contributes material both to
this check-in and to the final report.
The goal of the EDA is to explore the dataset to develop an understanding of the variables of interest and
how they may be related. The findings that are most interesting and that can be used to establish some
facts about your topic will go on the contribute to the persuasive argument in the final report.
The EDA will also be part of the final report, so will be graded twice. You can think of this check-in as
the draft EDA on which you will get feedback and the EDA in the final report as the final version.
Format
Your EDA should have the following format. It should be constructed as an RMarkdown document. Both
the markdown and a PDF version should be submitted. All code used to conduct the analysis should be
included.
I. Introduction
What is the topic under study and what do you think you can learn about it using the data?
II. Background
What data are you using? What structure does it have, including the level of observation and variables
included? What contextual information about the topic we will need to understand the EDA?
III. Data Wrangling
Show the data wrangling code required to get the data into the format(s) you use in section IV.
Discuss the main steps in words. This includes any new variables that must be generated, pivots or joins
made, recoding of variable categories, selection of a subset of the observations, etc.
2
IV. Exploratory Analysis
The exploratory analysis should include but is not limited to:
1. A histogram or density plot (continuous variables) or a bar plot (factors) of all variables used in the
analysis. Also a pairs plot of the variables you use. These should be accompanied by a discussion of
any salient features of the data that will be important for understanding the data.
2. The full set of visualizations you used to explore the dataset with the goal of learning about your topic.
This should include everything you tried to look at, even the dead ends. Make brief comments about
the visualizations that describe your thought process. Remember that the goal of this analysis
is to move you toward the eventual production a persuasive argument that uses the data
to establish a set of facts about your topic. That means that some of the visualizations you
produce should be appropriate for that purpose. When you feel you have the basis for that argument,
you’ve done enough exploring.
All visualizations should be properly labeled with plain English descriptions of variables on the relevant axes
or facet labels.
This check-in is due April 1st.
Presentation (5 points)
Groups will present their projects to the class on April 14, April 19, or April 21. Presentation slides are to
be uploaded to Canvas. Each group will have 10 minutes to present and three minutes for questions.
The presentation will focus on sections I, II, V, and VI of the final report as detailed below. You should
think of this as an opportunity to present your persuasive argument publicly and get feedback.
Final Report (15 points)
The final report will have the following general format. You are free to draw on material previously created
to construct it. Pay careful attention to differences in requirements.
I. Introduction
What is the goal of the project? What is the phenomenon under study? Give a brief summary of the EDA
and persuasive argument from sections IV and V below.
II. Background
What data are you using? What structure does it have, including the level of observation and variables
included? What contextual information we will need to understand the analysis?
III. Data Wrangling
Show and discuss the data wrangling required to get the data into the formats you use sections IV and V.
IV. Exploratory Analysis
This is the same EDA from the check-in with corrections and suggestions implemented or discussed if not
implemented.
3
V. Finished Analysis
This section makes a persuasive argument using visualizations of the data. You should repeat the visual-
izations you use in your argument again even if they are the same as some used in the exploratory analysis.
Visualizations should be well constructed and labelled and of publication quality.
Text accompanying the visualizations explains to the reader what they show and what facts about the
phenomenon under study you would like the reader to take note of and why.
You should also explain what limitations the data might present to understanding the phenomenon you are
investigating.
VI. Conclusion
What did you learn about the phenomenon you decided to study that you did not know before undertaking
it? What are directions you would take this project with more time or additional data?
VII. Appendix
All of the code you use should be made visible to at the end of the document by adding the following code
as the final lines.
## Appendix
‘‘‘ {r ref.label=knitr::all_labels(), echo=TRUE, eval=FALSE}
‘‘‘
Final Report Formatting Tips
Choose selectively what R code and output you want to appear in your final document text by including the
following options in the headers for code chunks.
1. Use echo=FALSE if you want to output to be displayed, such as a visualization, but not the code that
created it.
2. Use include=FALSE if you want neither the output nor the code to be displayed.
The resulting header would look like this:
‘‘‘ {r echo=FALSE}

‘‘‘
or
‘‘‘ {r include=FALSE}

‘‘‘
Knit your document to check the results.
Generally, you should not display the code or output with the text unless you discuss it in the text. The
appendix (see above) will contain all of your code so it can be checked.
4
Peer evaluation (5 points)
Group members will evaluate each other’s contributions of effort to their project. Note that I say effort and
not value. Members will have different strengths and sometimes good ideas will turn out to be dead ends.
The purpose of this evaluation is to deter the free riding of some group members on the efforts of others.
This isn’t a common situation in this class, but when it happens it’s a negative experience for all parties.
Data
Groups are encouraged but not required to find their own data related to a topic of personal interest.
Techniques for finding data will be discussed early in the semester.
Groups may also use Qualtrics software to conduct a survey for use in analysis.
A small number of interesting datasets will be made available to groups not wishing to find their own data.
5