Since it has been a while since we've had any articles in this series, you may want to refresh your memory by rereading the first and second installments in the 'Becoming a Critical Reader' series. I promise it won't be so long of a gap before the rest of them!

OK, having reviewed and identified your own personal biases, you are ready for the second read-through, where you can more critically read the article. We'll spend the next few posts in this series going through the various types of articles and the things you'll want to consider when doing this more critical reading. Some questions will be pretty universal, no matter what type of article you are reading. Others will be more specific to the various types of articles. We'll cover those specific questions over the next few posts.

The basic questions to ask as you read:

1. What did the authors set out to do? Hopefully you've already figured this out in your preliminary run through. If you're not clear on that, make that the first thing you look for. When you find it, write it down so you don't lose sight of that aim in the remainder of your reading.

2. Did the article really do what it set out to do? Look for the 'conclusions' or 'results' sections to see what the authors say about a study. Sometimes what is written here will have nothing to do with their original intent. Not that this makes the conclusions invalid, because sometimes studies do make important and interesting discoveries in tangential information. Ideally, the authors should at least address the original aim of the study, even if it was to say 'we did not find what we expected to find.'

NOTE: This question is NOT the same question as 'Did the study show what I think it should have shown' or 'Did the study look at what I wanted it to study?' Sometimes I hear people disparage a study by saying 'They looked at the wrong thing! Instead of studying 'when is the best time to do an induction?', they should have studied whether to do them at all!' This is unfair. The study is no less valid because it addresses a different issue than you would have chosen to research. The question is 'Did they do what they set out to do?'

3. Did the article use appropriate methodology? Some methods might not be a good way to study a particular question. Other methods might be a better approach. While some are fond of saying that only a double-blinded, randomized controlled trial can give you sufficient answers, it's not always realistic or ethical to do research in this way. We'll go more into detail on that in our next series on methodology, so don't worry if you don't know enough to make a good judgment on this aspect right now.

4. Did the author show undue bias or influence? Many studies will have a disclosure on the first or last page of a study that tells who paid for the study or if the researcher has any conflicts of interest. While I wouldn't necessarily dismiss a study because of a potential conflict of interest, I certainly would be using a VERY fine-toothed comb in my perusal of the study!

5. Do the conclusions match the data? Sometimes there is a pretty obvious mismatch between the two. I once read a study where the author concluded that a vaccine for GBS would save lives. However, the aim of the article was to find out if prenatal screening for GBS would reduce the incidence of serious GBS infections. Vaccines were not mentioned anywhere in the article, except in the conclusions area. It seemed an obvious mismatch to me.

Sometimes the mismatch might be more subtle. This is why you'll want to jot down that answer to the initial question, 'What did the authors set out to do?' At this point, go back and see if both their data and conclusions answer that.

These five questions are a good place to start as you review articles. I suggest that you take some time this week to find the full text of a study and read through it, answering these questions as you go.

Published: April 17, 2010

Tags