MKT 4561 Lecture 2: Data Visualization
2/14/19: Data Visualization
• Make a Qlik account (Qlik sense cloud)
• Rapidminer Studio
• Can you over-visualize your data? Yes. Pretty visuals are nice, but the message is
completely lost
• Key to good data visualization:
o Audience, story, data, goals, charts
o Why do all these things matter?
o If you don’t know your audience, then you tell a bad story. Your story you tell
varies.
o Stacking needs to be meaningful. How you structure your data makes a huge
difference
• What makes a good visualization?
o Information, story, goal, visual form, goal
▪ Use venn-diagrams to compare and correlate data
• Know your data
o Making websites. Look at what users mostly look at. How often do they come
back and look at it. Keywords. Use data to address user needs
o Create data
o Clean your data
▪ Outliers, missing data, formatting, condensing it, narrowing down the
specifics, duplicates
▪ Make sure to clean up doc for final project
o Document detailing the data (a key for what the data means) (like what is your
success?)
o What are you going to be doing long term with this information?
o Are you going to replicate it?
o Who does the information impact the most?
o How would knowing your audience?
• Know your story
o If you know your audience and you know what you want to tell them, you’re
halfway there
o Look at the data, take a survey
o Build a prototype that explains what is changing
o Storyboarding: UX, how do I get you where you want to go
o Once you know who you are speaking to, what you want to say, and how you
want to say it, then you can forward to the goal section
• Know your goals
o What are you trying to share?
• Know your charts
o Different kinds of data require different kinds of visualization
o Which chart is more effective in telling that story from another?
o Types of charts
▪ Comparisons: Compare one or more datasets over time. Stacked bar
graphs
▪ Relationships: Show a connection or correlation between two or more
variables. Bubbles and scatters
Document Summary
If you don"t know your audience, then you tell a bad story. Your story you tell varies: stacking needs to be meaningful. Information, story, goal, visual form, goal: use venn-diagrams to compare and correlate data, know your data, making websites. How often do they come back and look at it. Stacked bar graphs: relationships: show a connection or correlation between two or more variables. Bubbles and scatters: composition: display parts of a whole and change over time waterfall, pie chart, distribution: show how variables are distributed over time. bell curve. Make data clear and easy to read: line chart: reveals trends or change over time. 4 items maximum: pie chart: used to show parts of a whole, divide segments out of a 100. Donut, half donut, funky pie, full pie: bar and column charts: compare different items. Use if you have more than 10 items: dual axis charts: combine multiple charts by adding another y-axis.