Correction: Facebook Does Not Make You Stupid

We blogged a while back about some research suggesting Facebook use was correlated with low grades. Well, one Facebook-using professor named Eszter Hargittai thought the data looked fishy. So did Josh Pasek, a graduate student who got in touch with Hargittai — through Facebook, of course — and asked if she’d like to work on a paper with him challenging the earlier findings. Their paper, written with Eian More, has just been published. Analyzing a larger sample size than the original study did, the authors found no correlation at all between Facebook use and G.P.A. In defense of her work, the original study’s author, Aryn C. Karpinski, told the Chronicle of Higher Education: “I completely acknowledge the limitations of my research. What I found is so exploratory — people need to chill out.” [%comments]


Julie333w

I love it when researchers use technical jargon like "chill out" to defend their work.

Andrew

OK so technically Facebook may not make you stupid. But after a while of reading that your "friends" just made a salad or what rock star or 80's icon they are most like, or how many points they have in Mafia Wars, you can feel brain cells dying from the creeping banality.

Liz D

I'm just really tired of baby-boomers claiming that new technology and trends are somehow ruining our youth or society.

Everyone just needs to stop worrying about facebook and cell phones and TV and video games and all that. Kids are both as smart and as stupid as they ever were.

It's the future, get used to it.

JonA

The original study author is right. The results weren't her fault, the poor reporting is thanks to the media that is unable to report accurately on scientific studies.

But science moves on, replication and refinement of experiments win! But I doubt these results will grab any media headlines, so the public's state of being misinformed will be sustained. Media = poor at reporting science.

Brent

#2 said everything that needs to be said. No one else should comment

sunny

I agree with JonA, just the title of this article is misleading. Correlation is not ncessarily causality.

EP

Next time I get a bad review back from a journal, I'm going to tell the reviewers that they need to 'chill out.'

Seriously though, hypotheses are made to be disproven. That's science.

Eric H

Correct me if I'm wrong, but isn't it true that either study would only show correlation, and not causation? I know it seems like just a matter of how you frame the issue, but it has big implications.

Angie

I concur. I'm with #2 on this one.

Zach

I've been following the discussion of this study over on crooked timber for a while, and the original study never made claims of causation but was reported as so by the media because they love that kind of story. Hargittai was disputing the idea that there was a causation, but in expanding the sample size found that not only was there no reason to suspect causation, there is no likely correlation either.

Eszter Hargittai

Zach, glad to see that you got the point. The issue is that it was problematic to claim either correlation or causation given the major limitations of the original study.

Sarah

For a good article related to this, read the following...

http://www.cjr.org/the_observatory/facebook_and_procrastination.php?page=1

The latest study by Pasek and others was opportunistic and participating in the same media sensationalism that they are shunning. Who gets pleasure from beating up on young researchers before they even start their careers? That's unprofessional and definitely not collegial.

Eszter Hargittai

Sarah, I invite you and others to read our original article. Our goal was to make sure people didn't walk away with the wrong conclusions about the relationship of Facebook and grades. There are no personal attacks in the piece.

Sebastian

Oh my. I just read Karpinski's response.
He really, really, did not do himself a favor.
http://www.uic.edu/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/2503/2183
Some of the stuff he writes is just awful - e.g. Pasek et al. recode their dependent variable to a 0-1 scale, which he clearly doesn't understand, suggesting logit regression?

He also doesn't seem to appear the relationship between sampling and statistical inference (if you random sample - like in Eszter's data, you can make inferences about the population from which you sample - UIC first years in that case. If you sample by convenience you cannot draw any inference whatsoever. )
and so on.

He really should have had a faculty advisor stop him. I hope he isn't looking for a career in academia - the first study - flawed as it was - might still have helped him. This response is going to hurt. a lot.