Why extract data?

At this point in your systematic review, you should have a pile of studies that appear to answer your research question. All of them are slightly different: different populations, different outcomes, different study designs, and even different results.

Now what?

Before you can write a comprehensive report about this existing research, you need to transform those piles of studies into digestible and organized pieces of information.

You do that by going through every study and pulling out the details that will help you answer your question.

In other words, you extract your data. You might also see the phrase “data abstraction” in reference to data extraction. These terms are used interchangeably.

In an ideal world, all of your studies would use consistent and unambiguous approaches to reporting their data and each paper’s abstract would clearly state the key information about the research and the results.

Unfortunately, this is rarely the case. Data extraction is not always as simple as copying and pasting numbers from a report into a spreadsheet. Sometimes you will have to do some calculations to standardize your numbers. Other times you’ll have to do an interpretation of the results or the author’s intentions.

When you see conflicting data or poorly reported methods, this is often a red flag that the research might be unreliable. That, or the authors made a typo. Get ready to think carefully about how to put all of this information into a clear, useful format.

Next section: