How to Read a Study When You Do Not Have a Science Degree | Strength from the Well

Home › Know Your Body

Watch instead

The short version

Do not read it front to back. Read the abstract, then who was in it, then what they compared it against, then what they actually measured. Four questions and you will know more about that study than the person who posted about it. You are not grading the science. You are deciding whether it has anything to do with you.

Key points

• Read out of order. The middle of a paper is not where the truth lives.
• Who was in the study decides whether it applies to you at all.
• Compared to what is the question that deflates most headlines.
• Watch for percentages with no real numbers attached. That is where people get fooled.
• One study is one study. Nothing is settled by one of anything.

Somebody sent you a study. Or a headline about a study, which is not the same thing and usually not even close.

And you opened it and hit a wall of words and thought: this is not for me.

It is for you. It is written badly on purpose, or at least written for other researchers, which amounts to the same thing from where you are sitting.

You are not trying to grade the science. You are trying to answer one question: does this have anything to do with my body?

Nobody reads these front to back. Not even the people who write them.

Read the abstract at the top. Then jump to the table that describes who was in it. Then the last paragraph of the discussion, where the authors admit what went wrong.

That limitations paragraph is the most honest writing in the whole paper, and almost nobody reads it. It is the first place I go.

This is the question that quietly decides everything.

Thirty college athletes is not you. Mice are not you. Men in their twenties are not a fifty-eight-year-old woman with a thyroid problem, and pretending otherwise is how half the bad advice on the internet gets made.

Look for the number of people, their ages, their sex, how sick they were to begin with, and how long the whole thing ran.

If the people in the study do not resemble you in any meaningful way, you have your answer and you can stop reading.

This one sentence deflates more headlines than anything else I know.

Compared to nothing at all? Then you are looking at people who got attention, got hopeful, and got better - which happens, and which is not proof of the treatment.

Compared to a placebo? Better. Now you can see what the thing did beyond hope and time.

Compared to the treatment people already use? Best. That is the question you actually care about, and it is the comparison studies most often avoid.

There is a difference between a study showing that people could walk farther and a study showing that a number on a lab sheet moved.

The lab number might matter. It might also be a stand-in for something that never actually changed in anyone's life.

The wins that count are the ones you would notice. Stairs. Sleeping through the night. Getting off the floor without planning the whole operation first.

When a study measures those, take it seriously.

Somebody tells you a treatment cuts your risk in half. Sounds enormous.

Then you find out the risk went from two people in a thousand to one in a thousand. Technically half. Practically, almost nothing.

So whenever you see a percentage, go looking for the actual headcount. How many people out of how many.

That habit alone will protect you from most of what gets posted.

Funding is worth knowing. It is not a verdict.

A company-funded study can be perfectly good. An independent study can be sloppy. Funding is a reason to read more carefully, not a reason to throw the paper away.

And here is the flip side nobody mentions. Plenty of promising things have almost no research behind them for the opposite reason - nobody funded them at all.

Follow the patent. A human trial costs hundreds of millions, and whoever pays needs to own something at the end of it. Cheap, old, or unpatentable means no line of funders. So the research stops early.

That leaves a gap, and the gap gets read as proof of failure. It is not. We do not know yet and it does not work are different sentences, and only one of them is usually true.

Nothing is settled by one of anything.

Findings get overturned constantly. That is not the system breaking, that is the system working out loud in public, which is uncomfortable to watch.

If you can find a review or a meta-analysis - somebody who gathered all the decent trials and looked at them together - start there instead. It is the closest thing to a straight answer you will get.

Six questions. Print them, or just remember them:

• Who was in it, and are they anything like me
• How many people, and for how long
• Compared to what
• What did they actually measure
• Real numbers behind the percentages
• Who paid, and does anybody else find the same thing

Reading one study will not make you a doctor and it is not supposed to.

What it does is let you walk into a room and ask a real question instead of a scared one. That changes the whole conversation. I have watched it happen.

You were given this body to look after. Learning enough to ask well is part of that. It is stewardship, not suspicion, and there is no shame anywhere in it.

Read one. Bring one question. Then come tell the rest of us what you found.

Trust God. Be committed. Stay informed.

Do I need to understand the statistics?

No. Who was in it, compared to what, and how many people is most of the value. If the statistics are doing all the heavy lifting in a paper, that is usually worth being suspicious about anyway.

What does peer review actually guarantee?

That a few other researchers read it and did not spot anything obviously wrong. It is a filter, not a blessing. Peer-reviewed studies get overturned all the time.

How do I know if a journal is any good?

If you have never heard of it and neither has anybody else, take it with salt - there are journals that print anything for a fee. A reasonable check is whether the study shows up indexed on PubMed.

What if two good studies disagree?

Then the honest answer is that we do not know yet. Look at who was in each one - often they studied different people and both are right about their own group.

Is a bigger study always better?

Usually more reliable, but not automatically more relevant. A small careful study of people exactly like you can be worth more to you than a huge one of people who are nothing like you.

This is the part that actually matters. Real people, honest questions, nobody selling you
anything.

Join The Well →

Chat with someone directly

Or get new posts by email. Three a week,
that's it. Unsubscribe any time.

• PubMed, National Library of Medicine
• Cochrane Library
• ClinicalTrials.gov, National Institutes of Health
• MedlinePlus, National Library of Medicine

• How to Research a Treatment Nobody Offered You

  You heard about something and got eleven seconds of it at your appointment. Here is how to go find out for yourself, in about twenty minutes, without losing your mind at midnight.

• How to Tell When a Health Claim Is Garbage

  Not everything that sounds too good is a lie, and not everything that sounds boring is true. Here is the pattern recognition, so you can stop getting played in both directions.

• Your Lab Results Came Back. Here Is How to Read Them.

  A flagged number is not a diagnosis and a normal one is not a clean bill of health. Here is what reference ranges actually mean and which questions get you a real answer.