How To Choose The Right Graph Types of Graphs and When To Use Them

So, these are the world's five most popular charts, and in this video, I'll show you why these are the only charts you'll ever need, except for one, and how you can start using them. In first place, the world's most popular chart is the bar chart. There are few versions of them. So, let's go through the main types. Vertical bar charts are the best for category comparison. So, if you're comparing countries or teams, and in this example, we're looking at different apps and how many downloads there were in 2020. With bar charts, we can quickly and easily compare the differences in data with the story in this one being that despite what most people think, Tik Tok, not Zoom, was the most downloaded app of 2020. But vertical bar charts lose that quick comparability when there's lots of different categories. So try and only use them if there's less than seven categories. However, if you absolutely must have between seven and 15 different data points, flip your vertical bar chart on its side and you have horizontal bar chart. Really though, these are best when you want to communicate ranking. So things like election results, like this chart about the London May rule elections that shows the candidates ranked by votes. They're also great for showing things like performance stats like sales performance or this example here which shows Oscar nomination performance. You should use stacked bar charts if you want to show part to whole. So how much different elements contribute to that overall total. For example, we already showed this chart where the main message was that Tik Tok, not Zoom, was the most downloaded app of the year. But if you were then to use stacked bar chart and break out the downloads for the US, you still get the overall picture, but you also see that in the US, the difference between the two is actually lot smaller, which is probably why people thought that Zoom was more popular. warning though for stat bar charts, keep them simple. As soon as stat bar chart becomes far too complicated, like this one, comparing the different components becomes almost impossible and they become more or less useless. So that's bar charts. Now for the second most popular chart in the world, which is the histogram, which is basically bar chart. But if bar chart breaks down data by different categories, histograms should show the distribution over continuous variable. So things like time, age, weight. These charts are great to understand things like this one showing the income distribution in the US and probably most famously population distribution. like this one showing data for the US where age is on the y- axis and population on the x- axis. It's split by male and female. And histograms are generally less about the individual data points and more about the broader pattern that it shows. So if we quickly change this and look at data for Nigeria, you can immediately see from the pattern that Nigeria has much younger population. And taking bronze, the world's third most popular chart is the pie chart. Okay, let's talk about pie charts. Pie charts are hated by data visualization experts the world over. Mainly because they're really bad at showing multiple data points like this one. Fail. But also, our brains aren't really very good at comparing things like areas and angles. And here's challenge for you. So, look at this chart here and quickly work out which was the third most downloaded app. Three, two, one, times up. Okay. Now look at this one which is showing exactly the same data and tell me now which is the third most downloaded app. 3 2 1 bar charts are just much better than pie charts for communicating data. But if you absolutely must use one, follow these tips. Firstly, never use more than five data points. Then rank the values in order where the biggest segment begins in the top right hand corner. Then make sure you put all labels next to the segments which means you can also remove the key. and then use colors to help you communicate what you want to point out. In fourth place, we have scatter plots. Now, if you're looking to show correlation or how two things relate to each other, then scatter plots are your best mate. They can be used to either show that there is correlation or that there isn't correlation to show clustering trends or to spot outliers. They're probably one of the most persuasive charts to use if you want to convince somebody of something. The world's fifth most popular chart is the line chart. So if you want to show how something changes over time, then go for the line chart. It's great for showing things like stock market price, visitors to your website, or if you want to go really meta. This is line chart showing the search volumes of the term line chart since 2004. So we can quickly see that it's popularity has increased while also dipping every December and every July. So these are the world's five most popular charts. It's not to say there aren't other charts that you can use, but unless you're data viz master, stick with these and spend that extra time focusing on how to communicate these charts better. The best way to do this is to use the data to inc ratio, which you can learn in less than 5 minutes by clicking on the link on your screen. The data to inc ratio is one of the simplest and most effective things to keep in mind when you want to create more persuasive charts, graphs, and data visualization. So, click on the link on your screen to find out how you can start using it right now.