Wednesday, May 23, 2012

Serial Killers And Bad Science - Braintenance Reports: Question Everything!

Share this ARTICLE with your colleagues on LinkedIn .

The fact of the matter is that correlation dies not necessarily imply causality. Just because two things tend to be correlated positively (one usually accompanies, or is associated with the other), doesn't necessarily mean that a 'cause and effect' relationship exists.

In the extreme case, take this correlative study, which I just conducted myself in the The Braintenance Blog Laboratories (located in a concrete and lead bunker located deep below the desert in Area 52 -- as Area 51 was already spoken for when we were working with the realtors in our search for cheap laboratory space.):

In my experiment, I found out that in excess of 95% of all heroin addicts were given milk or formula when they were babies. Therefore, milk and/or formula, given to someone as an infant, will cause eventual that infant to graduate to heroin addiction.

Stupid, yes. Bad science always is. Numbers don't lie, but the inferences drawn from them can be manipulated or interpreted in many, many ways. The results of some studies could never, ever pass the common sense test.

The article excerpt and the video which are embedded below are made available though SmartPlanet, a wonderfully informative publication:


Scientists have found a curious similarity between a serial killer's murders and the firing patterns of brain cells.
Read the article which follows; After doing so, please hit the "BACK" button on your computer's browser to come back and read some interesting observations about this type of "science" and its drawing of sadly useless conclusions!
Serial killer's behavior explained with mathematics
As interesting (yawn) and informative (yawn) as this correlative (as opposed to causative) "study" seems, the refreshingly informed readers of this bit of "oh so what?" had quite a bit to say in criticism of this inactionable and ridiculous "study" of the obvious. Their comments, replete with misspellings, tangential campaigning, and off-topic wisecracks, follow for your pleasure:
Not very interesting
So the scientists have found that serial killing is somehow associated with becoming agitated. Well that's quite a revelation. I guess we can rule out that they did it in their sleep.
You put the cart before the horse
Neurons do not "fire" and then we have thoughts. We have thoughts and then neurons "fire". This is not the first time scientists have been able to model some behavior mathmatically. This same law of distributions could apply to the race car driver, one night stander, or any of a myriad things that result in a rush of adrenaline and other neurotransmitters. This guy some how latched onto murder as his high of choice. The problem will be that if this particular link is made then anytime scientists can show this particular neural activity pattern then the association that the person is a serial killer.
bumps on the head
Yeah, and Phrenology was once considered real too!
It's only a matter of time...
Ok... So now it's only a matter of time until some lawyer uses this as a defense in a murder case. "My client was the innocent victim of misfiring neurons."

Wait for it... wait for it...
Fitting the curve
I wonder how long it took to find a mathematical relationship that the data would fit?

1) Define the data points
2) Run the software to find the curve
3) Publicise the curve

It really sounds like a "Publish or lose tenure" situation. You get into the soft sciences and they can play some real games with data. I remember seeing a Biology (borderline hard/soft science) paper where an R factor of 0.3 was considered good. I'm a chemist. If I got results like that I'd service my measuring device.
Pareto principle
Your offhand throwing in of the 99% slogan of the occupy movement is off base.

If 20% of people controlled 80% of wealth, that would be a stable, or normal situation. The problem now and either a symptom of or cause of instability is the fact that 80% of the overal wealth is controlled by a much smaller group than 20%.

Every time such a disparity has occured historically, a correction (or war or revolution) has occured.
The pareto law applies to optimum functioning of the economy
The problem being that if the distribution of wealth to owners reaches or exceeds a parato distribution, the economy fails to function optimally. It also doesn't function optimally at a totally level(1 to 1) distribution of wealth to owners. The optimum being a small dynamic range lying between those two extremes.

Unfortunately, the majority of economists and financial experts are too stupid (i.e. unable to be trained out of it) to understand this concept and usually adhere to the fallacy of unlimited markets.

Now, back to Douglas E. Castle (who I happen to be)...

A statistical relationship is only a means, ultimately, of lending credence and credibility to a common-sense 'cause and effect' hypothesis, and even then, only if the proper experimental controls and conditions required by the Scientific Method are rigorously adhered to.

When it is doctoral thesis time, or when a batch of university professors get a publish-or-perish ultimatum, expect a rash of bad science to make its way into the news. I'll bet you that my reasoning here is right.

I could go a step further and make some emotionally-charged comments regarding the large pharmaceuticals companies and the FDA, but I wouldn't want to get myself put on a pseudo-ephedrine restriction list in the event that I have nasal congestion from whatever it is that I have been breathing...or perhaps reading.

Grow that brain. Ask questions. Stray off of the main roads. The world is your laboratory. Your life is a learning opportunity - have some fun with it. 


View DOUGLAS E. CASTLE's profile on LinkedIn

Douglas E Castle
All Blogs & RSS Feeds

Share this page
Contact Douglas Castle

No comments:

Post a Comment

Blog Archive

Bookmark and Share