The Biggest Mistakes People Make When Getting Started With Data Analysis

I’ve spent a lot of time around people who are interested in learning how to do data analysis and visualization. I’ve found that there are some common misconceptions or mistakes, people make when getting started with data analysis and visualization.

I’d like to describe the most important ones here: not taking it seriously enough; thinking you can learn this on your own; not understanding the limitations of Excel (and other spreadsheet tools); expecting one easy trick to do it all; trying to be too clever, and forgetting about presentation.

Here are the Biggest Mistakes People Make When Getting Started with Data Analysis:

Not Taking It Seriously Enough

This is simply a lack of commitment to actually follow through with learning how to do data analysis/visualization using R. It’s missing the necessary work ethic needed in order to succeed. I think this attitude comes from an inadequate appreciation for what “learning” really means. There are some great online courses that give the appearance of being very easy. However, in order to really learn how to do something, you need to practice it. It’s not enough just to follow along with an instructor and “do the homework assignments”.

If you’re truly committed to learning data analysis/visualization (in R), you’ll keep pushing through when things get tough. You’ll find yourself frustrated because you don’t understand what’s going on, or make errors, and it will feel like progress is slow. But if you can recognize this experience for what it is (normal for beginners) then that feeling of frustration won’t discourage you. Instead, you’ll view this as a sign that learning truly is hard work!

Thinking You Can Learn This on Your Own

The first step to learning a skill is admitting you need some help. I must admit that the data visualization and R communities have been very kind, and generous with their knowledge. I’ve benefitted from many people’s willingness to publish books, blog posts, videos, etc., which are all freely available. In fact, these resources are so complete that it feels as though there isn’t any more information out there on R programming/data visualization! And yet, those who learn these skills will tell you that there is still much to learn…

In my experience, most people underestimate how difficult it can be to do data analysis using R. It requires a lot of technical knowledge in order for you to get started. For example, you need a basic understanding of statistics, mathematics, and software engineering in order to write your own functions from scratch. Finding the right packages for R is also very hard because there are so many options! This can make using R seem daunting at first.

Good news though: 

I think everyone should try to learn this stuff on their own. There is an incredible sense of satisfaction that comes from learning something new. A word of warning though: don’t become overconfident about your abilities (i.e. thinking you no longer “need” any help). No matter how much effort you’ve put into learning on your own, it’s still not as good as having someone explain concepts to you; or solve some of these problems for you.

Expecting One Easy Trick to Do It All

There are many people who are looking for the proverbial hack to do data visualization well. This is because it’s really hard! And I think if there was just one easy trick that would solve all of your problems, somebody would have already shared it online by now. As Stephen Few says-

“If only data visualization were as simple as choosing a chart type and setting some options, then anyone could create good-looking infographics.” -Stephen Few (2012)

Don’t let this discourage you though! Like with any worthwhile pursuit, learning how to do effective data visualization will take time and dedication. There are no shortcuts here…as much as we’d like to believe otherwise.

But if you can appreciate the fact that learning to do this stuff requires a lot of time and practice, then it will be less discouraging when things get tough. You’ll just view this as part of the normal process of becoming an expert and that all experts struggle at the beginning! Also, visualization is more than just choosing a chart type and changing some options (the latter can be done without knowing any theory). Showing your work through visualization means understanding how to use design principles such as composition, balance, etc. so that your work is aesthetically pleasing too.

Choose Data Visualization for its Problem Solving Abilities

Okay, I’m going to take a leap here and assume you’re already convinced about data visualization’s importance. In addition, I’m going to assume you’re open to learning what it takes to do this stuff well. If you can adopt a growth mindset and really want to learn how to visualize data, then this is for you…

The first thing you should know about doing good data visualization is that it requires a lot of problem-solving skills. This means being able-minded enough to understand a complex system (in the case of using software like R). It also means understanding which questions your data can answer and designing effective ways of showing those insights.

Conclusion: 

If you’re interested in data visualization simply for its “pretty pictures” (which don’t always give accurate insights), then I’d say your time is better spent elsewhere. This stuff takes a lot of time and dedication to improve at it!