Define Your Data Jargon

One of the most important parts of a presentation on data analysis happens right at the beginning - setting the scope and expectations. Defining your visualization tools and data jargon with graphics keeps everyone focused on the important stuff during the meeting - your data story.

Using simple graphics to explain jargon can be very useful. Finding the right way to explain your tools is critical in the beginning, especially when speaking with people who aren't touching data every day. Executives, managers, and other stakeholders need to understand your analysis - they shouldn't get caught up in the jargon. These are your customers when presenting data.

I once watched a 60 minute presentation with incredibly interesting analysis and take-aways. The problem was these lessons and next steps weren’t given adequate attention, because 15-20 minutes was spent trying to explain why the analysis was using median, instead of average. The executives knew their business and were used to seeing their customer metrics as averages. The analysts used median consistently through the presentation, and this caused the executives to questions the numbers.

The executives weren’t wrong to question the data. They weren’t convinced the numbers truly represented their customer, because they weren’t used to looking at their data this way. They weren’t data scientists, and had to ask for a definition of median, and an explanation for why the presentation was using it.

The analysts also weren’t wrong. They had good reason to use median - they reviewed the raw data and saw major outliers that were skewing the average. Because they wanted to represent the bulk of the customers, they chose to use median, to represent a bigger portion of the client base.

Your data should be a pleasant experience. The more your audience has to work to understand you, the less time they have to focus on the actual analysis. Photo from Unsplash. Wall clock with mathematic equations for numbers.

Your data should be a pleasant experience. The more your audience has to work to understand you, the less time they have to focus on the actual analysis. Photo from Unsplash. Wall clock with mathematic equations for numbers.

From my perspective, the biggest issue was the analysts not taking their customers (the executives) into consideration when making the presentation. Once their customers started quesitioning the data, even though they eventually accepted the definitions and explanations of median, they were unconsiously less convinced of the rest of the analysis. They listened to the take-aways with more skepticism, and they didn’t have time to really dive deep into the important areas, because 1/3 of the time was spent on explanations.

Imagine an alternate presentation where after the introductions are made, there is a 60 second explanation of assumptions and definitions clearly laid out, with simple graphics that simultaneously convey the data definitions and business reasoning. Words, combined with visuals, can demonstrate with instantaneously. This is helpful for many reasons:

  1. You get ahead of any confusion with your jargon or analysis choices - there is a clear expectation from the beginning about what tools you’re going to be using and showing.

  2. You present yourself as an expert from the beginning. Instead of losing credibility by having your customers question why you’re doing something, you gain their trust that you’ve made informed decisions based on their data.

  3. You educate your customers for future presentations. Especially if you’re in an internal role, and are often meeting with the same group, you can begin to increase the level of understanding about your analysis and process with every talk and presentation you do, by adding in simple and easy to understand definitions.

Now, in the case above, the analysts were not an internal team, and they didn’t know what comfort level the executives had with data. They might not have been able to anticipate that they needed to define median (though defining or avoiding all jargon is always my suggestion). When the question came up, they would have been better served to get ahead of it. Find a whiteboard, use a pad of paper and make a quick sketch of the difference between average and median to answer the question. Instead of going into the formulas for with words, show the concept visually: where the average is in one color (skewed because of the outliers), and where the median is in another, right in the bulk of the customers. In an ideal world, you can create a hand out, use a whiteboard to the side, or have a second monitor that always has these definitions up during your presentation so they can be referred to throughout.

The sketch below shows one way to do this. The data isn’t real, the scale isn’t defined, and the points aren’t in the spots they truly would fall, but none of that matters to your audience! They need to understand the “why” - why you chose one method over another without getting caught up in complications. If they can see this and understand it with one graphic, you gain their trust and can move forward with the real analysis with fewer unproductive questions.

An example of a quick sketch that can clear up the reasoning behind choosing to use the data median over the data average. By Andrea K Haley. Graph with data points showing median in the middle and average nearer the outliers.

An example of a quick sketch that can clear up the reasoning behind choosing to use the data median over the data average. By Andrea K Haley. Graph with data points showing median in the middle and average nearer the outliers.

This quick sketch exercise is something you can practice on your own, to explain your concepts better. If you can’t sketch and explain the term it in less than 2 minutes, it’s probably too complicated to define this way, and you should avoid that term. That’s not to say you can’t use it for your data analysis, but be creative in making your presentation user friendly. Use storytelling and comparisons for data concepts such as negative regression or standard deviation.

If your customer/audience/user can’t understand your analysis, they can’t see the value in it, and they won’t be able to make data driven decisions from your analysis. Let them see the value by using simple graphics that define your data toolbox from the beginning.