Part 1 of Learning Tableau: Bronze Certified
Three weeks ago, I had never laid a finger on Tableau. Today, I uploaded my first dashboard to Tableau Public. How did I do it? Through the AMP (Analytics Management Program) developed and hosted by Tessellation.
What is AMP?
If you’re not already familiar, lets kick this post off by filling you in on what exactly the “AMP” is. The Analytics Management Program is Tessellation’s signature training offering which incorporates visualization training in Tableau with data preparation and agile delivery. What differentiates AMP from other self paced training programs is that Tessellation offers the experts’ time on demand and focuses on theory plus practice.
The Basics of the Craft
Step one in the quest towards Tableau mastery is learning the basics of the platform. The first set of lessons of the AMP training program are designed to be a comprehensive introduction to Tableau. These modules ramp up your skills through focused lessons, practice problems, and quizzes. The end goal of this section of training is to provide the skills necessary to create a dashboard from scratch that tells a compelling story and showcases your skills learned thus far. These skills include: learning different chart types, grouping and ordering data, color, filtering, and producing dashboards.
From the bronze level modules, I felt that not only was I learning the skills needed to create the desired visualization, but also the best practices for these skills. Tips such as an emphasis on formatting your tooltips, always keeping the end-user in mind, and proper use of color are given along the way to help you form good habits of creation within Tableau.
The Hunt for Data
Once I had completed all the bronze level modules, it was time to find some data. I spent a good chunk of time stumbling around online trying to find a data set that is big enough, with proper formatting, AND about a topic that interests me. This seemed like a tough task until I was introduced to the Google BigQuery archive of datasets. There were many interesting topics to choose from, including NOAA weather data and datasets from various cryptocurrency platforms. I settled on a dataset about crime in Chicago. I chose this because it is one of my favorite cities and the subject of crime sparked many interesting questions to chase in my exploration of the data story.
To access the data within BigQuery, I copied the table into my personal project, connected to it live with Tableau Desktop, and ran an extract which filtered out 2019 because it was incomplete. I verified that the data set included the necessary elements of date, measure and location (in latitude and longitude), and then I proceeded to connect to the dataset within Tableau.
I’ve learned the basics and I’ve found the data, so now what? Here comes the fun part: showing off your Tableau skills with a dashboard. Through the modules, I learned that any good story starts with a solid hypothesis that you can stick to, and eventually drill deeper through to reveal the whole story of your data. In the case of the Chicago Crime data, I hypothesized that crime rates are higher in the summer months. I let this hypothesis drive my initial dashboard design by focusing the layout of the board on the 4 seasons of the year. Personally, I like to draw out a design on paper before attempting it within the program. Once I have the layout planned, I can then move forward with manufacturing the worksheets for my dashboard.
First, I created the text chart that showed the total number of cases for each season. That number was later subjected to filtering by year and action buttons that further broke it down by type of crime. This action was kicked off by clicking on a box within a heat map of all of the types of crime. I could tell that the Summer value was consistently higher, in support of my hypothesis, as I viewed it separately over a few years. I wanted my end-user to be able to see this trend immediately without having to click through many years to understand the trend. Based off of this, I created a line graph of total cases over the last 10 years. This created a strong supporting message upon just glancing at my dashboard that crime was indeed higher in the Summer months.
As suggested by the modules, I spent a lot of time and energy on properly formatting my dashboard. I sought to hide all redundancies and unnecessary labels. I was also careful to match my color scheme across all elements of the board for the ease of understanding of my user. Once I was satisfied with the design and formatting of my creation, I uploaded it to Tableau Public and it was off to be graded by my experienced Tableau trainer.
I’m only through the first third of the AMP program and I already have confidence in the tool and self sufficiency with the basic skills of Tableau. If you’d like to check out my dashboard yourself you can find it here.