Give Me All the Charts!

Why care about data visualizations?

If you know me,

1) I CAN-NOT get a saying correct to save my life

2) For better or worse, I try very hard not to use cliches

3) I'm actually not a very good storyteller irl (see 1 and 2)

I could use a 🧀 cheesy 🧀 quote like this "Storytelling is the most powerful way to put ideas into the world.” or I could just try to explain in my terms why data visualizations are helpful.

But annoyingly this is true, visualizations are a form of storytelling with data. You're objective is to bring a narrative and guide the viewer through the story of your data.

Luckily for you and def for me, you don't have to be a naturally AMAZING storyteller to be good at creating data visualizations. You just need to think strategically. 🤓

First off, let's talk about why we make visualizations. Even if you understand numbers well, it can be hard to process them in a quick glance. Basically you are creating pictures that represent the same stats, so they can be more easily digested. Not to mention 65% of people are visual learners. 🖼️ People will be more likely to recall the numbers associated with the visualizations if presented to them in this form. Lastly, through visualizations you can narrate a cute little data story (ugh I know).

This story will help the viewers understand what is happening within the marketing, sales, workforce, risk management, or whatever other data you are analyzing.

The goal of a dashboard, or a group of charts and tables, is to show meaningful insights where an action could be taken.

Example

Here's a good example of how charts can help someone understand the data in a more efficient way. ⬇️

We have a market research response where you are showing two questions that have logic built into it. I.E. do you have a dog? If no, do you have a cat?

From this image, it shows how these two questions interact. You can easily see that 30% of those who don’t have a dog, also don't have a cat.

So where to start irl?

Cleaning data

For me even at the cleaning stage I am making note ✏️ of what visualizations would register this information best. Even though charting will typically be the last step in the process. If you need a refresh on best practices for cleaning data, read up here.

Get An Overview

After cleaning the data, you can run some summary statistics... frequencies, means, outliers, then potentially some cross-tabulation tables. You'll start seeing a good overview of the data. Are there any patterns that stick out? Are you curious if this one variable would vary by different segments? Is there a correlation between age and if they have a pet?

Continue performing whichever analysis' are necessary and then it's time to formulate a plan! 📋

START CHARTING

Go back to the objective of your report

• What were you trying to figure out?

• What do the stakeholders/boss/client of want out of the report?

• Are you trying to compare two different time periods?

You'll want to make sure you highlight any insights related to those aspects.

Generally, you'll also want to group charts together that are related or in the same category. How you organize, format, and annotate your dashboard will make a big difference in telling your data story (last time, I swear!).

This was just a small snippet of what I think through when creating any reports or dashboards. Stay tuned because next time we'll go over the different chart types.

Get excited for scatter plots, scorecards, maps, pie charts, and more!