6 Commonly Made Errors in Data Interpretation

Data interpretation

There are many ways that data interpretation can go wrong. Sometimes, things outside of your control get in the way and you have to deal with those challenges as they come. But much of the time, errors are made because those who work with data fall into common traps and make mistakes. There are many mistakes you can make when it comes to data interpretation, but there are a handful of mistakes that seem to occur more frequently than others. Here are six commonly made data interpretation errors to be aware of so you can avoid them.

Not Identifying a Clear Research Question

Many times, researchers get so excited about collecting and analyzing data that they rush through a critical first step: identifying a clear research question. Developing a specific question that’s relevant to your end goals helps you set a foundation for the rest of your project. It also helps you figure out which data is relevant and which isn’t. When you’re dealing with the overwhelming amount of healthcare data, or big data, you need to have laser focus so you can find the information you need and sort out the irrelevant information quickly. A clear research question can help you do that. Without a specific, focused question in mind, you can find yourself following lots of rabbit trails that don’t ultimately help you achieve your goals, and that can mean lots of wasted time and money.

Seeing Patterns That Aren’t There

Humans are hard-wired to notice patterns. In fact, we’re so hard-wired to see patterns that we’ll try to find them even when we’re told that what we’re looking at is random. Several experiments have been done that prove how quickly human beings are to jump to the conclusion that a pattern exists, even when there isn’t one. That tendency can be helpful when there really is a pattern in the data you’re looking at, but it can be disastrous if you’re merely trying to make connections that aren’t there. It’s important to double check yourself as you work with data to ensure that the patterns you’re finding are truly patterns and not the result of your mind trying to find one.

Confusing Correlation and Causality

Lots of factors are correlated, but they don’t necessarily indicate causality. Confusing the two can be dangerous. If there’s merely a correlation between a drug and a positive outcome and you assume it’s causality, it’s not hard to imagine how detrimental the results of that mistake could be. Always make sure that you’re not rushing to judgment too quickly when a correlation presents itself. Take the extra measures to ensure that causality is truly in place instead of merely assuming it is because it appears to be. After all, shark attacks and ice cream sales spike at the same time every year. But that certainly doesn’t mean that eating ice cream increases your chance of being bitten by a shark. It’s just that sharks are nearer to shore in the summer and people eat more ice cream in the summer.

Ignoring Context During Data Interpretation

As we’ve discussed previously, context is incredibly important. What appears to be true in a vacuum, or without surrounding data, can prove to be a much different scenario once you take into consideration all the contextual data surrounding it. This is similar to the mistake of confusing causality with correlation, but it’s different in that you’re not linking two events. Instead, ignoring context means you’re focusing only on one piece of data. Always make sure to look at every side of a given issue, look for surrounding data, and view any information you’re looking at in the context of existing data. Nothing exists in a vacuum, and nothing is totally unique. If you’re coming up with results that nobody in history has ever come up with, you’re probably doing something wrong.

Not Looking into the Numbers

Don’t take numbers at face value without digging deeper. If you’re looking at apparent facts and not asking questions like “Why is this so?” and “Were there mitigating factors?”, you need to slow down. It’s true that numbers don’t lie. But they can be misinterpreted. Accurate data interpretation relies on a full understanding of the data at hand. It’s important to understand how something happened and why it’s true in order to truly understand the data in question. For instance, knowing that a certain drug increases brain function in a given population is well and good. But looking deeper to find out that it’s a particular ingredient in that drug that causes this effect could lead you to develop a drug or treatment that’s even more effective. Obviously, this is a hypothetical and simplified example, but it makes the point.

Rushing Through Data Review

Having a specific research question is important, as mentioned above. However, rushing through the data review portion of your research — the part where you look at the data you’ve collected without trying to find anything in particular — is a mistake. It’s easy to want to get on with the research and find out if the data you’ve carefully gathered answers your question or not. But it’s important to take some time to view the data for what it is without any goals, aims, or objectives. You can miss some interesting and beneficial patterns and insights by rushing through this part of the data analysis process, so slow down and let the data “speak” to you before you start looking for specific answers.

Conclusion

There are many mistakes you can make in data interpretation, but by avoiding these common pitfalls you’ll be well on your way to a successful project. It’s exciting to delve into the research and answer questions, and when you accomplish your goals it’s even more exciting to know that your project, product, service, or development can help people live healthier and more enjoyable lives. But amidst the excitement, it’s important to stay focused, to follow the data instead of forcing it into preconceived expectations, and to avoid putting data in a vacuum or finding causality where none exists. By avoiding these mistakes you’ll set yourself up for an efficient, successful research project and you’ll sharpen your analytics and research skills in the process.