With so many visualization options in Tableau, it can be easier to create nonsense than to create a meaningful viz. And with the amount of data fields and marks available to us, it’s easy for the canvas to get overwhelmed or become incoherent. So before jumping into Tableau, let’s take a step back and sketch out our viz on a piece of paper first.

When creating a visualization, keep in mind its purpose or research question, who the viewers are, and what the actions users may take from these insights. The point is not to predict what the final visualization will look like, although that may happen for researchers who are very familiar with their data already. Instead, we’re trying to get a sense of how our variables may work together by actually putting them on different axes.

Let’s walk through an example together:

Situation: Leslie is a university Program Specialist and wants to better understand exam performance across the Biology department. They have data on exam scores, instructor, student major, class size, and year of study.

Purpose: Determine what factors may impact student exam scores

Research Questions:

  1. Do exam scores differ across majors?
  2. Do exam scores differ across grade levels?
  3. Is there a relationship between class size and exam score?
  4. Which instructors have the highest average exam score?

Viewers: Department Administrators

Actionable Insights: What can Leslie propose to the department if differences in exam scores are found? She could suggest changing class capacity, limiting specific courses to upperclassmen, syllabi evaluations, or requiring prerequisites.

With these factors in mind, Leslie can decide which charts are most appropriate. A scatter plot might work best for #3 because the goal is to illustrate a relationship between 2 numeric measures. Whereas, a bar chart could work for showing differences across categories in #1 or #2. and #4 could be depicted with a simple table. (Feel free to make a sketch of each to see what we mean! We can do this with Leslie’s imaginary data, because remember that we’re thinking about how different fields interact with each other rather than trying to accurately predict how resulting graphs will look.) Because they will have multiple visualizations to address multiple questions, a dashboard may be the best option.

Starting out on paper is helpful in framing the question(s) your viz seeks to answer and narrow down the pieces necessary to craft your story.


  1. Pull up the Movie Ratings dataset we’ve provided.
  2. On a piece of paper, answer the following questions:
  • What is the objective and research question(s) of your viz?
  • Who is the intended audience?
  • What data will be used and what insights might be drawn?

With those answers in mind, make a sketch of your viz. Think about what goes on each axis. Will the data be represented with shapes, lines, points, or text? How many and what types of charts will you incorporate? In the next lesson we’ll start creating vizzes with this same dataset. Let’s get to it!

Take this course for free

Mini Info Outline Icon
By signing up for Codecademy, you agree to Codecademy's Terms of Service & Privacy Policy.

Or sign up using:

Already have an account?