Monday, July 9, 2012

Bad even for Drudge

Two bits of data which are important for you to know before getting to the main post:

1.  As some of you may be aware, I have recently moved back to Akron, Ohio.  We've been furniture shopping, and I've found myself frustrated by the lack of good reviews for couches.  Every review I've found on any non-seller website is negative, without exception.  Does every couch suck?

Of course, it wouldn't be reasonable for a person whose couch is doing just fine to go out of their way to go online and tell the world how well their couch keeps their ass off the floor.  It's only the people whose couches break in half six months after they buy them that type out a review on Amazon.

2.  I'm a physician, and I can personally attest that filling out paperwork is the bane of most doctors' existence.

With that in mind, on to the main post.

This evening, the Drudge Report's top headline was a link to this article, which discusses a survey performed by an organization called the Doctor Patient Medical Association.  The survey, per Drudge's headline, showed that "83% of doctors have considered quitting over Obamacare".  The article describes the DPMA as "a non-partisan association of doctors and patients".  Which is a strange claim to make, given that the DPMA has an "Affiliate Relationship" with the National Tea Party Federation.  As one might expect based on the study's origin, there are major flaws with the article.

The DPMA's description of the survey states that it was faxed to over 16,000 offices, with 699 respondents, meaning that 4.3% of doctors actually returned the survey, which is kind of a shitty response rate for a survey, but which might be expected given the method of contact, faxing.  Considering the two facts above (re: couch reviews and doctors filling out forms), it is reasonable to conclude that there would be a significant selection bias in the survey's sample.  Few would take time out of their busy day to fill out a survey to tell a faceless, generically-named organization that they think the medical system is, is just doing great, goshdarnit!  You're going to wind up with people looking to complain, so you'd expect a negative bias.

More importantly, I would totally answer "Yes!" if I were asked "Have you considered quitting medicine because of Obamacare?", and I'm more pro-Obama than most.  Obamacare represents a significant change in the system that allows me to provide food, clothing and shelter to my family.  It would be foolish to not consider if such changes would make my current set-up untenable.

So Drudge (and the Daily Caller, the referenced source) lie about the source of a survey which uses flawed methodology to rack up as many tallies as possible for one particular response which doesn't actually indicate anything other than a recognition that Obamacare's a big fucking deal to doctors, which is obvious to anyone at all paying attention.


Anonymous said...
This comment has been removed by a blog administrator.
Anonymous said...

Just to help you out,
I've worked with a lot of furniture lines...
I really like Rowe, and they just revamped their entire quality control system.

Good luck!

Anonymous said...

You make an interesting point re. selection bias, but how far do you propose to extend that logic? Is a negative survey on anything automatically biased if it has a low response rate? What if it has a high response rate? Does that refute your theory? What if a survey with a low response rate is positive; does that make it reliable?

Merely asking.

Anonymous said...

For what it's worth the question the doctors were asked also made no explicit reference to the ACA, but rather asked "How do current changes in the medical system affect your desire to practice medicine?"

So thaks Drudge and friends. Right-wing pseudo-journalism ftw.

Anonymous said...

Reminds me of questions on a dating site I belong to that ask things like "Would you consider dating a neo-Nazi/someone with an arse for a face/someone 3x your age?" etc. As soon as you read the question, the considering is happening, so the answer is yes, I would consider it...but then I would say no.

Anonymous said...

>Is a negative survey on anything automatically biased if it has a low response rate? What if it has a high response rate? Does that refute your theory? What if a survey with a low response rate is positive; does that make it reliable?

The low response rate is not considered biased, but the methodology is. By faxing surveys to doctors' offices, it is considered a "voluntary response" survey, which the doctors can choose to response, or not at all. For anyone who's been to college and studied elementary statistics, "voluntary response" is a big no no, and represents a big discrepancy in data collection. Doctors who felt more strongly to the survey will likely to answer, and those who doesn't will be less likely to answer. If the organization called a selection of doctor offices across the US, their results may be more believable.

Panama City Doctors said...

Thanks for sharing such interesting post,it really very valuable for us. Panama City Doctors