How to Sound Intelligent: P-Hacking

Every week new studies get passed around Facebook and on morning news programs. Astounding discoveries that defy logic suddenly have scientific proof behind them.

Losing weight has nothing to do with diet and exercise; it’s all about sleep.

Coffee cures every disease known to man.

Smiling is better for you than exercise.

On and on they go. Odds are, next week new studies will cancel out the findings of the old studies.

Why is this? How can coffee or eggs be so good for you this week and so bad for you next week?

Largely this is due to a statistical term known as “p-hacking.”

Statistics can be used to prove anything. In an effort to curb that, statisticians have a variable that stats must fit into in order for them to be considered statistically significant.

Without going into the details, p-value was introduced to ensure that a result is actually real, not the result of chance.

In order to hit the accepted p-value of credibility, researchers do p-hacking until their results meet the required p-value.

This can be done by reducing or expanding the test size. Adding or subtracting different variables such as age, gender, size, what day the study was done on, etc.

Therefore, if you play with the variables enough you can conclude that coffee cures cancer. Another group of researchers can take the exact same data, do some p-hacking by changing variables, and use the same data to prove that coffee causes cancer.

Scientists and researchers don’t get money, attention, or tenure unless they come up with startling test results. Therefore, to keep making money, they p-hack their way to riches and notoriety.

And since all the money goes to the researchers with the most fascinating results, there is no money left for fact-checking.

Another factor in all this is that money comes from people who want certain results. Researchers know this going in. In order to get that money they must come up with acceptable results.

The notion that science is this unyielding quest for truth, throwing out error left and right, is becoming increasingly ridiculous to swallow.

Science is based on human observation. But the weakness of human observation is that human part. We’re tainted with sin, selfishness, pride, and greed. Even scientists who have studies proving there is no sin, only genetic predispositions, are tainted with sin.

Although science is currently riding high on a wave of adulation and happy thoughts, it is shooting itself in the foot.

But that’s OK, they will soon have a study telling us how feet cause cancer anyway.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s