Teaching with Tableau
Successfully integrating Tableau into the classroom
go.ncsu.edu/crash2
Alison Blaine
ablaine@ncsu.edu
Goals
Make an informed decision about using Tableau for classroom useKnow the strengths and limitations of Tableau for class assignments
Know how to prepare data for Tableau
Use the Tableau interface
Create a well-designed visualization
What is Tableau?
Tableau for Educators
Free 1-year licenses for:
academic instructors
students
classes
Tableau is Big
$199.9 million company
3,193 employees around the world
Leader in visual analytics and business intelligence
The Case for Using Tableau in the Classroom
It's free (for educators)
Easier to do visual exploration of data
Integrates well with other platforms
Marketable skill
Is it easy?
Open Tableau Public on your computer
Visit go.ncsu.edu/crash2
Download the data set
Create a well-designed visualization in Tableau with that data set
Discussion
What were your emotions?Were you successful?
What preparation, guidance or resources would you need to be successful?
Did you want to figure it out yourself or seek help?
Considerations
Tableau is not collaboration software
Tableau requires clean data
Interface and menus are not intuitive
Procrastinators will be sorry :(
What Tableau is Good For
Encouraging students to think visually during data analysisVisual experimentation without altering original data
Creating calculations on data during the vis process
Filtering data
No need for pivot tables
Interactivity
Integrating Tableau: Questions to Consider
What are the students ultimately trying to accomplish?What skills do they need to get there?
What level of support will they need?
Determining level of support
How crucial is Tableau to the overall project? (the more crucial, the more support)What does the original data look like? (the messier, the more support)
Will statistical analysis be required? (the more complex, the more support)
Is streaming or big data involved? (if yes, much more support)
Minimal Instruction Support
making a basic bar or line chart with two variables using clean data
More Instruction Support
filtering
designing a chart for communication with annotations
using dashboard actions
selecting a non-traditonal chart style successfully
pivoting data, splitting data, or making data unions
Most Instruction Support
working with multiple data sources
extracting data from PDFs
cleaning messy data
creating custom calculated fields
integrating with R or other software
professional-looking design
Data Pre-Processing
Characteristics of clean data:
single row of data per line
no nested headers or sub headers
no sub tables or extraneous descriptive text above or below table
each variable measured should be in own column
only one thing measured per table (ex: life expectancy)
So clean
Messy Data
NHIS Data
Clean Data, but not Tidy
Gapminder Data on C02 Emissions Per Capita
Demo
