Monday, March 19, 2018

Bad Science

This has me irritated.  Maybe more than it should... I am really irritated that people still find the need to share as if this was gospel truth.  This article claiming that childhood vaccinations are the cause of not just autism, but pretty much every other horrible thing that can happen to your child, is based on bad science.  Really bad science, in my opinion.

Let me tell you why.

But first, if you are a parent it is YOUR responsibility to make decisions in your child's best interest.  You need to base those decisions on good information from both your pediatrician and your own research.  Be an informed consumer.  If you conclude that vaccines do more good than harm, then get your child vaccinated without guilt and second-guessing yourself.  If you decide to forgo them, that is also fully your right.  I AM NOT OFFERING MEDICAL ADVICE OR EVEN AN OPINION ABOUT WHAT YOU SHOULD CHOOSE TO DO.  That's up to you!  I am only explaining why I believe this article shouldn't be used to inform that choice and, hopefully, giving you some tools to spot bad science when you see it.

First, you're going to have to read the article for yourself. While you are, notice some things that should be red flags any time you are reading an article based on a scientific study.

  • There's no link back to the original study just the landing page of the publishing journal.  You should be asking yourself why. We'll get to that in a moment.
  • The supporting studies that are, oddly, linked are old.  They are dated 2006 and 2007.  This is 2018.  Ten plus years will most likely have seen the findings in those studies obseleted many times over.  There are a few exceptions but generally, you should base your decisions on research that's not older than five years.
  • The author notes that the scientists conducting the study showed correlation.  And she hints that they recognize correlation is not necessarily causation.  Just hints at it with a single weakly worded sentence.  Correlation means they observed a relationship between the variable (vaccines) and a result (an illness).  Causation means they proved that the variable caused a specific result.  This article wants you to believe these scientists proved something they did not.

So... you're feeling some skepticism?  Good!  It's time to look a little deeper.  The lead scientist's name is given and you know the topic.  Google it.  Google knows everything!

Go ahead.  Google "Anthony Mawson vaccine study."  Dig through all those hits.  See where the Journal of Translational Sciences, the publisher of the study's findings, has retracted it?  When a peer-reviewed scientific journal retracts an article, it's a big deal.  That's not a decision they come to lightly.  It's an embarrassment to them so there has to be a very good reason.  Except...  You knew that was coming, didn't you?  The Journal of Translational Sciences isn't a respected scientific journal.  See the OAT at the top?  It stands for Open Access Text.  And if you dig deep enough, you see that oatext.com hosts a vast number of journals that you can pay to publish your study.  They give the reviewers 20% off the top, even.  You've heard the term "pay to play" on the evening news usually in regards to a government contract?  This is what it looks like in medical research.  And even they retracted this study!

If you do find one of the copies of the original study that still shows up now and then (because someone downloaded a PDF copy when it was still available and has posted it somewhere again) and start reading there are more problems.  John Q. Average probably doesn't know what a "cross-sectional sample" is in a scientific study so they may not pick up that what Dr. Mawson goes on to describe after those words is the survey of a "convenience sample."  Convenience samples do not have the same validity because all they can tell you is the opinion of a group of people who choose to answer your survey questions.  Think about who is going to choose to participate in any survey... only people interested in the topic (unless they are paid and there was no compensation involved in this particular survey) and ask yourself if that method provides a true representation of the entire population you are studying, in this case the American general public.  Sometimes it might.  This is called generalization.  Not being generalized is a big technical flaw in a study.  Big.  Huge.

And consider the survey itself.  He selected a bit more than 600 homeschooling mothers in 4 states and asked them if their children were vaccinated (not at all, partially, and fully were the choices) and what health problems those children experienced.  No records review to see if their memory was correct.  No independent assessment of the children's health conditions.  No control group.  Faulty data collection method, poorly constructed survey questions, no substantiation.  There are also several issues with bias ranging from Dr. Mawson's role as an outspoken advocate for the anti-vaccine movement, funding from anti-vaccine groups to do the study, and using the same parents he surveyed to promote and disseminate information from it. Bias is a bad thing in scientific research because it can lead you to swing results to a pre-determined conclusion. If you're cherrypicking for the answer you want, is it really research?

Vaccines are a subject where I have no expertise so I can't offer you an opinion about whether or not they are a good thing.  And that's not the point I'm trying to make anyway.  What I hope you've learned from my dissection of Dr. Mawson's study is how to spot bad science in the future whether it's descriptive journalism in the popular media or seeking out a scientific journal and reading the published study findings.  Good research and the truly awful exist side-by-side and are shared widely and quickly with today's technology.

In a world where everyone sells themselves as an expert, be a skeptic.  Ask questions and read critically.  Use your mind instead of your emotions even when it's an emotional topic.  Make sure you've got good information as the basis for your choice.  Google is your friend - look up words you don't know.  Consider both sides of the argument fairly.

And then...

Make your choice.

(And for goodness sake, when you find bad science DON'T SPREAD THE TRASH FARTHER by posting it on Facebook!)

1 comment:

Unknown said...

Love it Kathy!
The topic of vaccines and autism etc always makes me giggle! These people who believe its true, how do they account for all of us yes us who never been vaccinated but have autism and other illnesses?
My mom never vaccinated me, I know that because I have had measles, German measles, chicken pox and Scarlet fever! If I had been vaccinated, I would never had gotten those. Yet I have autism (as you very well are painfully aware of after all these years of friendship)!
You got right to the point on this and you can make remember my Civics Professor in college! He said: if I at the end of this class has made you question what you hear - lovely, if I make you distrust what you read, amazing and if you don't believe what you see: I have truly succeeded as a Professor!