The Journal of Economic Education

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/vece20

Learning Tableau: A data visualization tool

Steven Batt , Tara Grealis , Oskar Harmon & Paul Tomolonis

To cite this article: Steven Batt , Tara Grealis , Oskar Harmon & Paul Tomolonis (2020):
Learning Tableau: A data visualization tool, The Journal of Economic Education, DOI:
10.1080/00220485.2020.1804503

To link to this article: https://doi.org/10.1080/00220485.2020.1804503

Published online: 20 Aug 2020.

Submit your article to this journal

Article views: 49

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=vece20
THE JOURNAL OF ECONOMIC EDUCATION
https://doi.org/10.1080/00220485.2020.1804503

ECONOMIC INSTRUCTION

Learning Tableau: A data visualization tool

Steven Batta, Tara Grealisb, Oskar Harmonc, and Paul Tomolonisb
a
Research and Instruction Librarian, University of Connecticut, Storrs, CT, USA; bDepartment of Accounting
and Finance, Western New England University, Springfield, MA, USA; cDepartment of Economics, University of
Connecticut, Storrs, CT, USA

ABSTRACT KEYWORDS
“Doing economics” is an important theme of undergraduate economics Data literacy; data
programs. Capstone courses increasingly include instruction in “data liter- visualization; econometrics;
acy” and the STEM-related skills of quantitative and empirical methods. economic education
Because the professional discipline has moved in this direction and JEL CODE
because of greater employer demand for these skills, data visualization is a A2
key component of data literacy. Tableau is a free data visualization soft-
ware widely used in the data analytics industry. In this article, the authors
introduce an exercise that teaches the fundamental Tableau concepts and
commands needed to create charts, assemble them in a dashboard, and
tell a story of patterns observed in the data. The exercise assumes no prior
experience in Tableau and is appropriate for undergraduate upper-level
economics courses or an empirical methods course.

Literature review
Hoyt and McGoldrick (2019) survey the five decades starting from the 1970s and single out the
decline in the number of economics majors in the 1990s as a transition point in the economic
education literature. Before the 1990s, economic educators were concerned with incorporating
emerging technologies into instruction. Since the 1990s, the phrases “learning by doing,” and
“training to think like an economist” appear in seminal articles by Siegfried et al. (1991) and
Hansen (2001); these phrases have become the mantra for writers on economic education as
Hoyt and McGoldrick document.
“Doing economics” in undergraduate economics increasingly includes working with data. The
pedagogy of undergraduate econometrics is moving, albeit slowly, from the “Stone Age” emphasis
on statistical efficiency and functional form to the “Computer Age” of causal interpretation and
empirical examples (Griffith and McFall 2013; Angrist and Pischke 2017). Similarly, undergradu-
ate departments are turning toward increasing instruction in empirical and quantitative methods.
By 2018, economics degrees classified as STEM-eligible (econometrics and quantitative) com-
prised 13.2 percent of all undergraduate economics degrees conferred, up from just 1.2 percent in
2012 (Marshall and Underwood 2020). A key factor in the orientation toward developing empir-
ical quantitative methods is recognition of a greater demand by employers for these skills
(Angrist and Pischke 2017; Kassens 2019; Marshall and Underwood 2020).
Data literacy is emerging as an important component of “doing economics” (Halliday 2019).
Hoyt and McGoldrick (2017b) summarize what the literature emphasized that “doing economics”
was at that time, using analytical and presentation tools that students will be using after

CONTACT Oskar Harmon oskar.harmon@uconn.edu Department of Economics, University of Connecticut, Storrs, CT

06269-1063, USA.
ß 2020 Taylor & Francis Group, LLC
2 S. BATT ET AL.

graduation. They identify five goals of student research: (1) develop a worthwhile research ques-
tion; (2) connect it to the literature; (3) formulate a testable hypothesis; (4) collect, clean, and use
appropriate tools to analyze data to test the hypothesis; and (5) present the findings. The penulti-
mate goal defines what data literacy is, and is clearly foundational to the last goal of presenting
the findings. A recent symposium section of the Journal of Economic Education highlights efforts
in five top economics departments that have the goal of promoting economic literacy and a cul-
ture of undergraduate research (Hoyt and McGoldrick 2017a). Common to the department efforts
is a focus on developing data literacy and investing in infrastructure support from data librarians
as well as IT support specialists in software usage (e.g., Stata and GIS).1
Data visualization is a tool of data literacy: “Data visualization is the graphical representation
of information and data. By using visual elements like charts, graphs, and maps, data visualization
tools provide an accessible way to see and understand trends, outliers, and patterns in data”
(Tableau.com 2019, online). Data visualization helps the researcher identify relationships in data
and enhances a presentation in the same way as a thousand words while taking up less space
(Hennessey 2014). However, examples of modules published in the economic education literature
with exercises aimed at developing data literacy skills for undergraduate economics courses are
relatively few, although growing. In our search of the economic education literature, we uncov-
ered four such exercises, as discussed herein, none of which uses Tableau.
The four visualization tools that have been explored in the economics of education literature
are GIS, FRED, Infographics, and Excel. GIS (Geographic Information Systems) is a software that
has the specific purpose of linking geographic information with descriptive information (ESRI
2012, 6). Peterson (2000) provides four assignments, using data prepackaged for GIS, to create
thematic maps visualizing geo-spatial economic relationships for topics in economic development.
FRED (Federal Reserve Bank of St. Louis [FRBSL] 2019) combines charting and mapping tools
with an online database of economic time series. FRED specializes in the visualization of time
series data. Mendez-Carbajo (2015) and Mendez-Carbajo, Taylor, and Bayles (2017) provide sev-
eral assignments (using prepackaged data from the FRED online database) illustrating theoretical
macroeconomic relationships with time series data. Infographics is a visualization that combines
art and minimal text to make data easily understandable at a quick glance (Rouse 2012).
VanderMolen and Spivey (2017) describe an assignment to develop an infographic for a health
economics class. They discuss the steps of developing a research question and collecting relevant
data. They include links to resources that describe “best practices” for infographics and tutorials
on how to use several widely used infographics software. Excel is a workhorse software for many
tasks, including data “cleaning” and visualization. Halliday (2019) includes an assignment for an
economic development class that uses Excel to graph a Lorenz Curve. At first, students work
with prepackaged data, and then they progress to collecting and “cleaning” data from a pub-
lic site.
Learning a new data literacy tool has opportunity costs, so why Tableau? Because Tableau is a
general-purpose data visualization tool more widely used in the data analytics industry than GIS,
FRED, Infographics, or Excel.2 Whereas GIS specializes in thematic geographic mapping (ESRI
2012), FRED (Federal Reserve Bank of St. Louis [FRBSL] 2019) in plotting time series data, Excel
in data cleaning, and Infographics in directing focus to a specific statistic (Rouse 2012), all these
features are also available in Tableau. Hennessey (2014) has an excellent overview of the options
in Tableau that can be used to make thematic maps (like GIS), to manipulate data (like Excel), to
make time series plots (like FRED), and to make infographics (like Infographics). Tableau’s com-
parative advantage, though, is in using multiple large data sets to create charts in interactive dash-
boards from which the user can drill down to identify patterns in the data (Shaptunova 2017;
Solanki 2017; KnowledgeHut 2019). Tableau also has a user-friendly drag-and-drop interface. For
undergraduates looking toward the job market, data visualization is a skill in demand in the field
of data analytics.
 LEARNING TABLEAU 3

Figure 1. A story in Tableau.

To our knowledge, this is the first self-contained Tableau tutorial designed for an economics
course. Our tutorial combines the comparative advantages of the data librarian with the econo-
mist. A data librarian performed the processing of data (downloading and cleaning) and the ini-
tial data visualization in Tableau, and economists framed the data into the puzzle of gender
pay equity.

The tutorial: Components of Tableau visualization

Overview
In Tableau, the student begins by developing a single chart/graph in a worksheet. One or more
worksheets are combined into dashboards, and the dashboards are linked by tabs in a story. The
story is the final step of the data visualization and, as the name implies, tells the story of the pat-
terns identified in the data. The story developed in this tutorial is shown in figure 1.3
The gender wage gap is a well-researched topic in labor economics that is often measured as
the ratio of female to male earnings (Blau and Kahn 2017; Bolotnyy and Emanuel 2018).
Occupation and industry are important explanatory variables of the gender wage gap because
women are more likely to be in low-wage occupations that are less likely covered by union col-
lective bargaining agreements (Blau and Kahn 1997). Further, Blau and Kahn (2017) estimate
that, in 2010, 82.1 percent of the wage gap was explained by measurable human capital variables
(age, work experience, education) and 91.6 percent by including measurable variables for industry
and occupation. The residual was explained by discrimination and other unmeasurable variables.
Therefore, in our tutorial, we focus on the effect of the control for occupation,4 as combined
with weekly average earnings by gender and by occupation.
Tableau annotation tools can be used to add commentary on patterns observed in the data;
hence we recommend a few background reading assignments pitched at a level for undergradu-
ates with at least a course in principles of microeconomics, so they can draw conclusions from
the data in their presentation. As a primer on the gender wage gap, we recommend sections 1
4 S. BATT ET AL.

and 2 of Gould, Schieder, and Geier (2016). For a review of factors explaining the gender wage
gap among occupations, we suggest Hegewisch and Williams-Baron (2017). For a general discus-
sion of how students can include Tableau visualizations in research projects, we use Hennessey
(2014), which has an excellent review of the theory and principles of visualization using Tableau.
Our work complements it in that we provide detailed instruction in Tableau commands and illus-
trate these commands with a cleaned data set.
We use national data from the Current Population Survey (CPS) for weekly average earnings
by gender and by occupation. Our use of the CPS data is for reasons of relative ease of access
and wide use in the economic studies of the wage gap. The U.S. Bureau of Labor Statistics’ Web
site includes a “Labor Force Statistics from the Current Population Survey” page of tables on
employment and earnings data, with more than 20 years of data downloadable in Excel and
text format.
Table 39, “Median weekly earnings of full-time wage and salary workers by detailed occupation
and sex,”5 was chosen as the basis for this tutorial because it contains wage data for a relatively
high number of occupations for male and female workers. The original table was downloaded
and modified slightly for use in Tableau: the column headings were reformatted into a single
row, merged cells were deleted, and employment and wage estimates were converted to whole
values (the estimates in the original table are expressed in thousands of dollars and thousands of
workers). Only those rows with occupations that had estimated wage data for both men and
women were retained—125 detailed occupations out of the original 535. Although cleaning and
structuring data for use in Tableau is an important skill, we felt that preformatting the data in
the packaged Tableau workbook avoided confusion, saved time, and allowed the students to focus
primarily on understanding core Tableau functionality.
Tableau Public provides extensive support materials, including several how-to videos in the
Resources section of the Tableau Public Web site.6 Before the hands-on class sessions in which
students were introduced to Tableau, students were required to watch the first how-to video,
“Tableau Public Overview.” The other brief videos are helpful for understanding how to connect
to data, assemble and format dashboards, and other Tableau functionalities. In addition to exten-
sive, well-maintained software documentation, Tableau supports a robust online community
where users can post and search for questions and answers.
In figure 1, the title of the story “Wage Gap Project,” is labeled with the red letter “A” and
appears at the top of the main frame. Note that we include in the title directions from the tutorial
on how to build a story; however, the usual use of the title caption is to explain the purpose of
the data visualization. The pages of the story are the gray tabs opposite the red letter “B”; each
tab displays the related dashboard (red letter “D”), although the storyboard title (red letter “A”) is
constant. Opposite red letter “C” is the dashboard title caption; as shown in figure 1, we display
the instructions for constructing dashboard #1 of the tutorial. However, the student is instructed
to replace this with an appropriate description of the dashboard. Figure 1 also displays the dash-
board related to the first chart/graph (red letter “D”), which corresponds to the first worksheet of
our exercise.

Worksheets
The ultimate result of the tutorial is to build a story that presents the findings from the CPS
data; this starts with a worksheet. The first worksheet of the exercise is displayed in figure 2. The
horizontal bar at the top (in the blue highlighted rectangle) has icons to activate commands that
apply operations to the chart in the worksheet. Mouse-over reveals the operations of the icons.
Extremely useful commands are those to: “undo and redo” (A), “sort data” (B), and the “Show
Me” pallet (C).
 LEARNING TABLEAU 5

Figure 2. Guide to the Tableau workspace.

The Data pane, (D), has the labels of the categorical variables (Dimensions) and quantitative
variables (Measures). Making a chart starts with dragging variable “pills” from Dimensions and
Measures (D) to the Columns and Rows shelves (E).
Clicking the Show Me icon (C) reveals the chart types appropriate for the data, highlighted
with a red border, and selecting one, a bar chart in our case (F), results in the chart appearing in
the rectangle “G.” The Marks card (H) has the commands to format the chart display (e.g., color
of bars, labels, numerical units, etc.).
In the exercise, we use the Caption field (I) for the worksheet to embed step-by-step instruc-
tions for the user to build the worksheet. In a completed story, the worksheet caption is used to
explain the visualization in the worksheet. In this first worksheet, we show the user a few basic
formatting commands in the Marks card. The horizontal bar “J” at the bottom has tabs to navi-
gate the worksheets, dashboards, and story.
The exercise has eight worksheets that gradually introduce the student to Tableau visualization
commands. We introduce several chart types: horizontal bars, treemap, scatter plot, and trend
line (with statistics). We introduce some of the options for formatting various attributes such as
labels, currency symbols, and axes labels. And, we introduce the command of Calculated Field to
make combinations and transformations of variables.
The worksheets build on the theme of exploring two dimensions of wage-gender difference—
(1) the differential concentration by gender in occupations and (2) the wage differential by gender
in occupations. The second set of worksheets (2a through 2d) builds charts that display occupa-
tions separately by concentration of males and females. This allows students to observe that some
occupations are more heavily concentrated with one gender than other occupations. The third
worksheet illustrates the 25 highest-paying occupations for men and the corresponding earnings
for women in the same occupations. Students should find it evident that in some occupations the
wage differential between males and females is larger than in other occupations. The fourth set of
worksheets (4a and 4 b) instructs the students to display a scatter plot and trend line of wage dif-
ferential mapped on the concentration of female workers in occupations. Students can discern a
correlation between the wage gap and type of occupation.
6 S. BATT ET AL.

Figure 3. Guide to the Tableau dashboard.

Dashboards
A dashboard collects one or more worksheets. Figure 3 displays Dashboard 4, which displays two
worksheets. The instructions to the student for building a dashboard are in the Caption box for the
dashboard (A). The dashboard is constructed by dragging worksheets from the Dashboard pane
(B) to the center of the Dashboard box. The chart of women’s earnings as a percent of men’s is in
the box labeled “C.” The trend line regression of women’s earnings as a percent of men on the per-
cent of female workers to the total number of workers in an occupation is in the box labeled “D.”
This shows that the ratio of female wages to male wages increases as the occupation becomes more
concentrated with female workers, albeit with only an R-squared of 0.109 but a statistically signifi-
cant coefficient of 0.159 (t-value ¼ 3.89). To give a background for understanding patterns in the
data in “D,” the article by Hegewisch and Williams-Baron (2017) is assigned, which is an introduc-
tion to the research that examines the causal relation between the wage gap and gender segregation
of workers by occupation. The box labeled “E” gives the student the opportunity to make observa-
tions on the patterns in the data. Located in the horizontal bar at the bottom in the right corner
are the icons to navigate among the dashboards (F) and to create a dashboard page (G).

Story
The Story is a Tableau feature that binds the dashboards into storyboard presentation with a
shareable link that has tabs with captions that link to each of the data visualizations. It is a con-
venient and concise way to share the charts with annotations that the author can use to point out
highlights to the reader. The story for this exercise was introduced earlier in figure 1; this is the
sharable link.7

Suggestions for use

Tableau supports the use of Tableau Public and Tableau Desktop in an academic setting in a
number of ways. The “Resources” page on the Tableau Public Web site includes a Sample Data
 LEARNING TABLEAU 7

section with several dozen preformatted data sets in areas including sports, government and busi-
ness. The Tableau Community Forums provide an Instructor Resource that includes guides for
students and instructors, as well as information on free licenses for Tableau Desktop for students
and instructors. A Tableau for Teaching User Group provides sample assignments and curricu-
lum as well as assignment tutorial workbooks on Tableau Public that students can download and
complete (similar to our wage gap assignment).
For instructors considering using Tableau for the first time in the classroom, our advice is that
it can be a challenging and rewarding experience. Five pointers are offered: (1) if the type of chart
desired is not highlighted in the Show Me card, try changing the default setting in the Marks
card for the type of chart from Automatic to the type of chart desired; (2) to have several lines
(bars) appear in the same line plot (bar chart), from Dimensions, drag the pill with the different
categories onto Colors in the Marks card; (3) save your file to the Tableau Public server fre-
quently so your work is not lost; (4) practice with the four below examples; and, (5) if all else
fails, try Googling the question. There is a very large community of Tableau users who post help-
ful videos and comments.
Each of the next four examples uses cleaned data from Kaggle and has a few simple visualiza-
tions in Tableau accessible to a user with an introductory understanding of Tableau (see links).
The first exercise is the U.S. minimum wage by state from 1968 to 2017 in 2018 equivalent
dollars (2750 observations scraped from the U.S. Department of Labor). Possible courses for use
include public finance, contemporary economic problems, and labor.
For the Tableau exercise, use the “connect” tool in Tableau to import the workbook,
“Minimum Wage Data.xlsx.” From Dimensions, drag “Year” to the Columns shelf and from
Measures, drag “High 2018” to the Rows shelf. In the Marks card, switch “Automatic” to “Line”;
from Dimensions, drag “State” to the Filters card and select, for example, three states: CA, CT,
KY. Then, from Dimensions, drag “State” to the Marks card. The line plot compares the trend in
the real minimum wage for three states from 1965 to 2017. The visualization reveals some inter-
esting differences.
Next, to make a table, mouse-over “Sheet 1,” then depress the “Ctrl” key (right click on PC)
and select “Duplicate as Crosstab.” Select the “Show Me” card, select “text table” first row first
column, then select the “swap rows and columns” icon in the top toolbar. The table lists the real
value of the minimum wage for the three states for the time period.
Then, to make a treemap in a new sheet, from Dimensions, drag the Year pill on to the Filters
card and select year 2017. From Dimensions, drag the State pill to the Columns shelf, drag the
High 2018 pill to the Rows shelf, and change SUM to AVG. In the Show Me card, select treemap
(third row, first column). From Measures, drag the High 2018 pill onto Label in the Marks card.
Lastly, create a Story and drag Sheet 2 on to the first story point, add a new story point and
drag in Sheet 2, and do the same for the last sheet. Publish to Tableau Public and remember to
save while open on the first story point.
Links: Kaggle:
https://www.kaggle.com/lislejoem/us-minimum-wage-by-state-from-1968-to-2017/data
Example Tableau visualization:
https://public.tableau.com/profile/econ2327#!/vizhome/MinimumWageExample/Story1?publish = yes
The second exercise is the Boomers and Millennials, which contains a data set (2,750 observa-
tions) of Millennial and Baby Boomer Employment identified by state, MSA, and industry for the
years 2009–2013. There are separate worksheets for metropolitan areas, and for 2- and 5-digit
occupation codes. Possible courses for use include urban economics or capstone research course.
For the exercise, use the “connect” tool in Tableau to import the workbook,
“EMSI_MillenialsvsBabyBoomers.xlsx,” and drag the worksheet “States” into the data pane. From
Dimensions, drag the State pill to the Columns shelf and from Measures, drag the % Job Change
pill to the Rows shelf, and change SUM to AVG. In the Show Me card, select maps (second row,
8 S. BATT ET AL.

second column). From Dimensions, drag the State pill to the Filters card and deselect Alaska and
Hawaii. From Dimensions, drag the Generations pill to the Filters card and select Millennial.
From Dimensions, drag the % Job Change pill to Label in the Marks card and change Format
to percentage.
Then, to show a side-by-side comparison in a new sheet, from Dimensions, drag the State pill
to the Columns shelf and from Measures, drag the % Job Change pill to the Rows shelf and
change SUM to AVG. From Measures, drag the % Job Change pill to Color on the Marks card.
Click the Swap Rows and Columns icon. From Dimensions, drag the % Job Change pill to Label
in the Marks card and change Format to percentage with 1 decimal place.
Next, to show a line plot comparison in a new sheet, from Dimensions, drag the State pill to
the Columns shelf and from Measures drag the % Job Change pill to the Rows shelf and change
SUM to AVG. In the Marks card, switch “Automatic” to “Line,” and from Dimensions, drag the
Generation pill to the Marks card. The line plot compares the change in job share of Boomers to
Millennials from 2009 to 2013 across states. Generally, the peaks and valleys track each other, but
there are some exceptions: Georgia and North Dakota are interesting exceptions.
Lastly, create a Story, and drag Sheet 1 on to the first story point, add a new story point and
drag in Sheet 2, and the same for Sheet 3. Publish to Tableau Public and remember to save while
open on the first story point.
Links:
https://public.tableau.com/s/sites/default/files/media/EMSI_MillenialsvsBabyBoomers.xls
Example Tableau visualization:
https://public.tableau.com/profile/econ2327#!/vizhome/BoomersMillennials/Story1?= yes
The third exercise is the U.S. educational finances with annual state-level data on revenues
and expenditures for U.S. K–12 schools from the U.S. Census Bureau. There is a data file for
school districts with 374,162 observations (districts.csv) and another file for states with 1,276
observations (states.csv). Possible course applications include urban economics, public finance,
and contemporary economic problems.
In the Measures pane, click the Total Expenditure pill, in the dropdown carrot, select
“create,” then calculated field,” and create the variable Spending Per Pupil as the ratio of Total
Expenditure to Enroll. From Dimensions, drag “Year” to the Columns shelf and from
Measures, drag the Spending Per Pupil pill to the Rows shelf. In the Marks card, switch
Automatic to Line; from Dimensions, drag State to the Filters card and select, for example,
three states and the District of Columbia: CA, CT, NC, and DC. Then, from Dimensions, drag
the State pill to the Marks card. The line plot compares the trend in education spending for
three states and DC from 1992 to 2016 in thousands of dollars. The visualization reveals some
interesting differences.
To visualize a map, in a separate sheet, from Dimensions, drag the State pill to the Columns
shelf, from Measures, drag Spending Per Pupil to the Rows shelf and change SUM to AVG. In
the Show Me card, select maps (second row, middle column). From Dimensions, drag the Year,
State pills on to the Filter card, and filter by Year (2016), and State (uncheck Alaska and Hawaii).
From Measures, drag Spending Per Pupil onto Label in the Marks card, change from SUM to
AVG, select Format Currency Custom, and select 3 decimal places.
Create a Story and drag Sheet 1 on to the first story point, add a new story point, and drag in
Sheet 2. Publish to Tableau Public and remember to save while open on the first story point.
Links:
https://www.kaggle.com/noriuk/us-educational-finances
Example tableau visualization:
https://public.tableau.com/profile/econ2327#!/vizhome/StateEducationExpenditures/Story1
The fourth exercise is the U.S. unemployment rate by county, 1990–2016, data set scraped
from the United States Bureau of Labor Statistics. It has county-level unemployment data with
 LEARNING TABLEAU 9

885,548 observations. Course applications include urban economics, contemporary economic

problems, labor, and public finance.
For the Tableau exercise, use the “connect” tool in Tableau to import the text file,
“output.csv.” From Dimensions, drag Year pill to the Columns shelf and from Measure, drag the
Rate pill to the Rows shelf. Click the dropdown on the Rate pill and select Average. In the Marks
card, switch Automatic to Line. From Dimensions, drag the State pill to the Filters card, click the
dropdown, select Edit Filter and select, for example, CT. From Dimensions, drag the County pill
to the Filters card, select the dropdown to Edit Filter and click Select from List and select all the
counties in CT. Then, from Dimensions, drag the County pill to the Marks card, and drag it
again to the Marks Card on Color. Now we have a line plot comparing the unemployment rate
of CT counties since 1990. The visualization shows that 30 years later, unemployment is slightly
higher than at the start, and the variance in the unemployment peaks are differentially larger for
some counties.
To visualize a map, in a separate sheet, from Dimensions, drag the Year, State, and County
pills on to the Filter card, and filter by Year (2016), State (CT), and County (All). From
Measures, drag Rate to the Rows shelf and change SUM to AVG. From Dimensions, drag County
onto Color in the Marks Card. In the Show Me card, select maps (second row, middle column).
From Measures, drag Rate onto Label in the Marks card, change from SUM to AVG, select
Format Custom Numbers, and select 1 decimal place.
Create a Story and drag Sheet 2 on to the first story point, add a new story point and drag in
Sheet 1. Publish to Tableau Public and remember to save while open on the first story point.
Links:
https://www.kaggle.com/jayrav13/unemployment-by-county-us
Example tableau visualization:
https://public.tableau.com/profile/econ2327#!/vizhome/UnemploymentRateByCountyinCT1990-
2017/Story1

Summary
An emerging and complementary tool of data analysis is data visualization in order to envision
the relationships and then communicate those relationships convincingly to others. Tableau is a
widely-used data visualization tool, yet in the economics of education literature, there are only a
handful of exemplary research projects illustrating data visualization and none for Tableau. Our
tutorial seeks to fill this gap by identifying and describing a research question—gender wage dif-
ferences. As the student is creating visualizations, there are opportunities for the student to look
for patterns and to speculate about possible causal relationships. The visualizations generally sup-
port the theory that the gender wage gap decreases as the occupation becomes more concentrated
with female workers.
The contribution of this exercise is to make Tableau more accessible to undergraduate students
in economics, as well as to those instructors thinking about incorporating data literacy in their
course work. Our tutorial exercise is a self-contained module that provides a step-by-step tutorial
on how to develop data visualizations in Tableau with a cleaned data set. Along the way, the
student is encouraged to think about explanations for the gender wage gap and has the
opportunity to annotate their visualization with their explanations of data patterns observed in
the visualizations.
An extension for this exercise is to export the data to an Excel file to estimate a multivariate
OLS regression, which can further explore patterns observed in the Tableau univariate trendline
visualization. Another extension is to apply the framework of the exercise to another empirical
exercise with easily accessible data sources in Kaggle.
10 S. BATT ET AL.

Notes
1. These department efforts are described in: Archibong et al. (2017), Lima and Tsiang (2017), Feyrer (2017),
Ferderer and Krueger (2017), Brunnermeier (2017), and Butcher and Weerapana (2017). Additionally,
Marshall and Underwood (2019) describe data literacy in the context of a capstone course.
2. The CV Compiler (Ste 2019) survey of 300 data science vacancies reports that data visualization is among
the top five key employer preferences and that employers prefer Tableau as the tool for data visualization.
The (Enlyft 2020) survey of 300 thousand firms that use data visualization software for business
intelligence reports that Tableau has the largest market share (15.5%) followed by Cognos (9.3%). The Hale
(2019) survey of several thousand job listings for data scientists in Indeed, SimplyHired, and Monster
reports that, among the top 15 technologies in demand, Tableau ranks #7, Excel #14, and Python, R and
SQL are the top three. For takeaways on technologies to learn: “Learn Tableau for data visualization. It’s
probably the technology on the list that is the most fun to learn and the quickest to pick up” (Hale 2019).
3. Links to the student and instruction version of the tutorial, to a YouTube video of a how-to step-by-step
tutorial for each Tableau worksheet, and to download the free public version of Tableau are located at
https://stamford.econ.uconn.edu/learning-tableau-a-data-visualization-tool/.
4. McDonald and Thornton (2011, 2016) describe a student exercise for simulating the wage gender gap
using the National Association of Colleges and Employers (NACE) database, which is based on the
technique in McDonald and Thornton (2007). A similarity with the tutorial herein is that they use college
major, whereas we use occupation as the explanatory variable to analyze the wage gender gap.
5. The data used herein are from https://www.bls.gov/cps/aa2018/cpsaat39.htm, the most recent data
available. Data from prior years are at this link: https://www.bls.gov/cps/tables.htm.
6. https://public.tableau.com/s/
7. https://public.tableau.com/profile/econ2327#!/vizhome/Wagegapworkbook02-24-20As/Story1

Acknowledgment
The authors are indebted to Bill Goffe and anonymous referees whose comments greatly improved this article.

Disclosure statement
The authors certify that they are not involved with Tableau.com in any way that would constitute a conflict
of interest.

References
Angrist, J. D., and J.-S. Pischke. 2017. Undergraduate econometrics instruction: Through our classes darkly.
Journal of Economic Perspectives 31 (2): 125–44. doi: 10.1257/jep.31.2.125.
Archibong, B., H. Dekker, N. D. Grawe, M. L. Olney, C. Rutz, and D. Weiman. 2017. Forging on-campus connec-
tions to enhance undergraduate student reasoning, writing, and research skills. Journal of Economic Education
48 (4): 317–26. doi: 10.1080/00220485.2017.1353466.
Blau, F. D., and L. M. Kahn. 1997. Swimming upstream: Trends in the gender wage differential in the 1980s.
Journal of Labor Economics 15 (1, Part 1): 1–42. doi: 10.1086/209845.
———. 2017. The gender wage gap: Extent, trends, and explanations. Journal of Economic Literature 55 (3):
789–865. doi: 10.1257/jel.20160995.
Bolotnyy, V., and N. Emanuel. 2018. Why do women earn less than men? Evidence from bus and train operators.
Working paper. Cambridge, MA: Harvard University, Department of Economics. https://scholar.harvard.edu/
files/bolotnyy/files/be_gendergap.pdf.
Brunnermeier, S. 2017. Learning by doing: The challenge of engaging undergraduates in economics research.
Journal of Economic Education 48 (4): 290–94. doi: 10.1080/00220485.2017.1353461.
Butcher, K. F., and A. Weerapana. 2017. Striving to involve undergraduates in economic research at Wellesley
College. Journal of Economic Education 48 (4): 295–300. doi: 10.1080/00220485.2017.1353462.
Enlyft. 2020. Companies using Tableau. https://enlyft.com/tech/products/tableau (accessed June 8, 2020).
ESRI. 2012. What is GIS? Redlands, CA: ESRI. https://www.esri.com//media/Files/Pdfs/library/bestpractices/what-
is-gis.pdf (accessed February 15, 2019).
Federal Reserve Bank of St. Louis (FRBSL). 2019. What is FRED? https://fredhelp.stlouisfed.org/fred/about/about-
fred/what-is-fred/ (accessed February 15, 2019).
 LEARNING TABLEAU 11

Ferderer, J. P., and G. Krueger. 2017. Building research skills in the Macalester economics major. Journal of
Economic Education 48 (4): 301–5. doi: 10.1080/00220485.2017.1353464.
Feyrer, J. 2017. Undergraduate research in the Dartmouth Economics Department. Journal of Economic Education
48 (4): 306–9. doi: 10.1080/00220485.2017.1353463.
Gould, E., J. Schieder, and K. Geier. 2016. What is the gender pay gap and is it real? Report. Washington, DC:
Economic Policy Institute. https://www.epi.org/publication/what-is-the-gender-pay-gap-and-is-it-real/ (accessed
August 17, 2019).
Griffith, A., and T. A. McFall. 2013. Using PGA tour results to illustrate the effects of selection bias. Journal of
Economic Education 44 (3): 238–48. doi: 10.1080/00220485.2013.795456.
Hale, J. 2019. The-most-in-demand-tech-skills-for-data-scientists. https://towardsdatascience.com/the-most-in-
demand-tech-skills-for-data-scientists-d716d10c191d (accessed June 5, 2020).
Halliday, S. D. 2019. Data literacy in economic development. Journal of Economic Education 50 (3): 284–98. doi:
10.1080/00220485.2019.1618762.
Hansen, W. L. 2001. Expected proficiencies for undergraduate economics majors. Journal of Economic Education
32 (3): 231–42. doi: 10.1080/00220480109596105.
Hegewisch, A., and E. Williams-Baron. 2017. The gender wage gap by occupation 2016; and by race and ethnicity.
Fact Sheet. IWPR #C456. Washington, DC: Institute for Women’s Policy Research.
Hennessey, J. 2014. A picture can be worth more than a thousand words: Integrating data visualization into an
undergraduate econometrics course. Paper presented at the Conference on Teaching and Research in Economic
Education, Washington, DC, May 28–30.
Hoyt, G. M., and K. McGoldrick. 2017a. Models of undergraduate research in economics: Advice from eight exem-
plary programs. Journal of Economic Education 48 (4): 288–89. doi: 10.1080/00220485.2017.1353465.
———. 2017b. Promoting undergraduate research in economics. American Economic Review 107 (5): 655–59. doi:
10.1257/aer.p20171069.
———. 2019. 50 years of economic instruction in the Journal of Economic Education. Journal of Economic
Education 50 (2): 168–95. doi: 10.1080/00220485.2019.1582388.
Kassens, A. L. 2019. Taking a path less traveled: Mastering metrics without a textbook. In Teaching economics:
Perspectives on innovative economics education, eds. J. Hall and K. Lawson, 137–53. Switzerland: Springer
Nature.
KnowledgeHut. 2019. Why Tableau is the most important tool for data visualization. Blog. Last updated October
17, 2019. https://www.knowledgehut.com/blog/business-intelligence-and-visualization/why-tableau-is-so-popular
(accessed February 20, 2019).
Lima, V., and G. Tsiang. 2017. Training young researchers: Successful strategies from University of Chicago college
economics. Journal of Economic Education 48 (4): 310–16. doi: 10.1080/00220485.2017.1353467.
Marshall, E. C., and A. Underwood. 2019. Writing in the discipline and reproducible methods: A process-oriented
approach to teaching empirical undergraduate economics research. Journal of Economic Education 50 (1):
17–32. doi: 10.1080/00220485.2018.1551100.
———. 2020. Is economics STEM?: Trends in the discipline from 1997 to 2018. Journal of Economic Education 51
(2): 167–74. doi: 10.1080/00220485.2020.1731387.
McDonald, J., and R. Thornton. 2007. Do new male and female college graduates receive unequal pay? Journal of
Human Resources XLII (1): 32–48. doi: 10.3368/jhr.XLII.1.32.
———. 2011. Estimating gender wage gaps. Journal of Economic Education 42 (4): 405–13. doi: 10.1080/00220485.
2011.606094.
———. 2016. Estimating gender wage gaps: A data update. Journal of Economic Education 47 (2): 140–41. doi: 10.
1080/00220485.2016.1146100.
Mendez-Carbajo, D. 2015. Visualizing data and the online FRED database. Journal of Economic Education 46 (4):
420–29. doi: 10.1080/00220485.2015.1071222.
Mendez-Carbajo, D., K. G. Taylor, and M. A. Bayles. 2017. Building a Taylor rule using FRED. Journal of
Economics Teaching 2 (1): 14–29.
Peterson, K. D., Jr. 2000. Using a geographic information system to teach economics. Journal of Economic
Education 31 (2): 169–78. doi: 10.1080/00220480009596773.
Rouse, M. 2012. Infographics. WhatIs.com. Last updated February 2012. https://whatis.techtarget.com/definition/
infographics (accessed February 15, 2019).
Shaptunova, Y. 2017. Tableau software review: Pros and cons of a BI solution for data visualization. SamSolutions.
Blog. https://www.sam-solutions.com/blog/tableau-software-review-pros-and-cons-of-a-bi-solution-for-data-visu-
alization/ (accessed May 8, 2019).
Siegfried, J. J., R. L. Bartlett, W. L. Hansen, A. C. Kelley, D. N. McCloskey, and T. H. Tietenberg. 1991. The status
and prospects of the economics major. Journal of Economic Education 22 (3): 197–224. doi: 10.1080/00220485.
1991.10844710.
12 S. BATT ET AL.

Solanki, A. 2017. Tableau is more than a data visualization tool. Softweb Solutions. Blog. https://www.softwebsolu-
tions.com/resources/tableau-data-visualization-consulting.html (accessed February 15, 2019).
Ste, A. 2019. How to become more marketable as a data scientist. CV Compiler. Blog. https://cvcompiler.com/blog/
how-to-become-more-marketable-as-a-data-scientist/ (accessed May 7, 2019).
Tableau.com. 2019. Data visualization beginner’s guide: A definition, examples, and learning resources. https://
www.tableau.com/learn/articles/data-visualization (accessed February 15 2019).
VanderMolen, J., and C. Spivey. 2017. Creating infographics to enhance student engagement and communication
in health economics. Journal of Economic Education 48 (3): 198–205. doi: 10.1080/00220485.2017.1320605.

Appendix
Three other clean data sets that can be used.
#1—US Veteran Suicides
2005–2011 veteran deaths outside of combat by state
https://www.kaggle.com/residentmario/us-veteran-suicides
The data consist of six files, one for each year between 2005 and 2011. Each year’s worth of data includes the
general population of each U.S. state, a count of suicides, a count of state veterans, and a count of vet-
eran suicides.
#2—Fatal Police Shootings, 2015–Present
Civilians shot and killed by on-duty police officers in United States
https://www.kaggle.com/washingtonpost/police-shootings
A database of every fatal shooting in the United States by a police officer in the line of duty since January
1, 2015.
#3—Home Price Index
Housing indexed prices from January 1991 to August 2016
https://www.kaggle.com/PythonforSASUsers/hpindex
For the Federal Housing Finance Agency House Price Index (HPI), monthly and quarterly time series from
January 1991 to August 2016 for the U.S., state, and MSA categories. Analysis variables are the aggregate
nonseasonally adjusted value and seasonally adjusted index values. The index value is 100 beginning in
January 1991.