More Stories
Data can prove anything ... no matter how absurd

This is my latest column in the Blue Chip magazine at the Eugene Register Guard...

It’s hard sometimes not to be a skeptic. The Internet today is riddled with endless surveys and data along with headlining results. Who can we trust? What do we believe?

One of these days I’m going to start a new consulting company based on the sad truth that in today’s world a good search turns up nice-looking data to prove just about anything

A while back I read the following on a post on

“A UK study found that over 50 percent of social media users evaluated their participation in social networking as having an overall negative effect on their lives. Specifically, they singled out the blow to their self-esteem that comes from comparing themselves to peers on Facebook and Twitter as the biggest downfall.”

In defense of the author of that post, she’s obviously making fun, going on to say that there’s bad behavior in social media that will make your blood pressure rise.

She was joking, but the study wasn’t. And seriously: When something starts with “A study found that …” do you pay attention?

I used to, back before the Internet, when information was hard to find. These days I don’t have the same respect for “a study found” because there are studies to find anything you want to say. No matter how preposterous.

Take the post I found on the blog by David Mielach of BusinessNewsDaily called “Can Facebook Make You Fat and Poor?”

The post says researchers found that social media users were more likely to binge eat and have a higher body-mass index. Frequent Facebook users also were more likely to have certain financial problems, including a lower credit score and higher levels of debt.

But wait. It says the research was based on the responses of 541 Facebook users in the United States. So what does it really mean?

First, who makes up the sample? Is it Facebook users or Facebook users who answer surveys? Those are different sets of people. Is it balanced for age, demographics, technology, geography?

Maybe the people who answer surveys on Facebook have less self control — that’s why they took their time to answer the survey. And maybe people who answer surveys have less money because they waste their time answering surveys. Maybe they are just younger, on average, and that causes the money difference. And maybe they lie to surveys.

So maybe what it really shows isn’t about Facebook users but rather about people who answer surveys.

Then there is that whole issue of causation and correlation: Could we just as easily say living in a large house makes you rich, or attending college makes you young? That’s as logical as saying Facebook users have less self control and less money. Right?

Here’s a direct quote from the research: 

“These results are concerning given the increased time people spend using social networks, as well as the worldwide proliferation of access to social networks anywhere, anytime via smartphones and other gadgets. Given that self-control is important for maintaining social order and personal well-being, this subtle effect could have widespread impact.”

So now it’s “widespread impact.” Wow.

I’m not saying that information is bad. Misinformation is. I’m not saying that research is bad. Believing it is. Letting “research” end the discussion and trump common sense is bad.

Question the research, question the assumptions, look through it, and then take what’s valuable in it. Never just believe it.

More posts are loading...