My Digital Analytics Quest
Chapter (6 of 12):
Well here we are in week 6, halfway through my allotted time to complete this CXL Digital Analytics Mini-degree program and I am continuing to be humbled all while learning so, SO much. So, instead of digressing ad nauseam like a recipe blogger, I will cut to the chase and get into the “meat and potatoes” of what I have learned this week. This week was mostly dedicated to the Data Visualization and Presentation module of my program. I’ll be honest, I didn’t expect to get too much out of this module since data visualization and presentation are a significant portion of my job.
Once again my expectations of the module have been proven tragically and epically wrong. Now, why is that, do you think? Here, let me give you a hint…by asking you a question…how much time do you think I had devoted to understanding the brain science behind human-to-human communication? On a scale of one to ten, where do you think I would land? If I’m being completely honest, a solid 3.5 maybe a 4.0 if I’m being generous to myself. (Enter heaping slice of CXL’s finest humble pie)
This particular module is taught by none other than my latest man-crush Monday candidate: Tim Wilson, Senior Director of Analytics at Search Discovery. Tim did a phenomenal job of breaking down each video into an insightful, but easily digestible lesson on the principles of effective data visualization and presentation. One of the things that I loved the most about this module was that it was data visual platform agnostic. The principles can be effectively implemented across whatever your platform of choice is. Pick your poison!
Tim starts the course out by prefacing with an intriguing method he uses to approach his data visualization strategy. The overall strategy is centered around the practice of “reducing the cognitive load” of the data you’re presenting. But, wait! What is a cognitive load? Well contrary to popular belief (a belief I admittedly aligned with until recently), Tim says, “data visualization is not about making the data pretty, it’s about making it understandable”. Simple, yet sound strategy in my opinion.
Still don’t understand the “reducing cognitive load” principle? Let me explain it this way, in the Gestalt school of psychology it is believed that humans are more adept at perceiving entire patterns or configurations, as opposed to individual components. For example, say we have a table of data packed with valuable information and right next to it, we have a time series line chart depicting that table’s data trends over time. Which do you think is easier to perceive and understand? Well, if you’re a member of the species, homo sapien, and not a bionic time-traveling cyborg that processes data as easily as we drink water, chances are you’re going to choose the time series chart every day of the week, and twice on Sundays.
If I can distill this any further, I would explain it like this…through the miraculous process of evolution over the last few billion years, human existence is pretty recent, and our ability to read even more so. Before writing was invented, we spent a significant portion of our existence drawing pictures on cave walls. As a result, our brain is best at reading, interpreting, and understanding information in picture form. In general, it takes a considerable amount of concentration and focus to read a data table and if there is one thing our body does better than anything, it’s overwhelming instinct to conserve the body’s calorie consumption. Where reading and interpreting information in a simple, picture-type visual format consumes fewer brain calories, this is how we reduce “cognitive load”. Ultimately, the idea is to make understanding the data as easy as possible.
A few of Tim’s methods for doing just that, that I like are::
- Horizontal Bar charts > Pie Charts
- Maximize the Data / Pixel ratio
- Use color strategically and SPARINGLY
Why are horizontal bar charts better than pie charts? Let’s unpack this real quick here, shall we? The reason Tim gives for using horizontal bar graphs instead of pie charts, especially when conducting a categorical comparison is that humans are not all that great at guesstimating areas…especially areas of a circle. Anyone who has eaten more pie than their eyes perceived understands this reasoning intimately. I guess it kind of gives credence to the old saying after overeating, “your eyes were bigger than your stomach”. The same holds true with data consumption where pie charts and donut charts are concerned. And! whoever came up with the idea that putting a hole in a pie chart would all of a sudden make it easier to discern the correct area, was seriously disturbed. I mean, it’s a donut, how many waistlines have been compromised as a result of faulty donut area perception? Oh, just one more won’t hurt…
The horizontal bar chart, on the other hand, lends itself to how we naturally process data. As it turns out, if you do a quick Google search you will find that our friends at Speaking Powerpoint, who have so kindly allowed me to borrow a visual from their blog post arguing why Tufte is wrong about pie charts, would agree:
I think you and I would both agree that in an apples-to-apples cross-categorical comparison, using a pie chart against a horizontal bar chart — the horizontal bar chart would win by an epic margin.
Maximizing the Data / Pixel ratio, now this probably sounds super complicated and overly technical, but aside from the name, it’s pretty straightforward. You take away any extraneous details that distract from the point you are trying to make with the data visualization. Basically, it’s a strategy for effectively helping your audience to focus on what you want them to focus on and see what you want them to see. There are myriad ways of doing this, but this segues beautifully into Tim’s third method that I like.
Using color sparingly AND strategically is a highly effective way of getting your audience or even yourself to focus on the details that matter. Think highlighter in college…you used your highlighter as a tool to signal to your brain what information among the overwhelming, steaming pile of academic hot air that composes the majority of over-priced college textbooks and @$$-numbingly dull college lectures, was important for you to learn in order to pass your classes. Do you see what I did there…? Pretty clever, right? When I drafted this, I had the italicized, bolded text highlighted…but it didn’t transfer over. Oh, well…you get the point, right?
