A conservative organization called the Intercollegiate Studies Institute has released a new report which shows (yet again) that most Americans don't have extensive factual knowledge of politics. (Whether this matters very much is a question that has been widely debated among political scientists.)
Based on this finding, the Washington Post's Kathleen Parker embarassed herself and her newspaper by suggesting that "passive activities... diminish civic literacy" while "[a]ctively pursuing information... and participating in high-level conversations... makes one smarter":
What's behind the dumbing down of America?The ISI found that passive activities, such as watching television (including TV news) and talking on the phone, diminish civic literacy.
Actively pursuing information through print media and participating in high-level conversations -- even, potentially, blogging -- makes one smarter.
But as anyone who has taken statistics 101 knows, correlation does not equal causation. The fact that people who score lower on the ISI quiz tend to watch TV and talk on the phone does not mean that those activities reduce their civic literacy. A more plausible explanation is that people who have low civic literacy scores also tend to watch more TV and talk on the phone more. (At a minimum, Parker and ISI can't distinguish between these two possibilities with cross-sectional survey data.)
How did this passage make it into the newspaper? Does anyone at the Post know anything about statistics?
I think you misunderstand the Post's perceptual frame.
Parker is merely issuing a timely reminder of what 'everybody knows'. It's no more than 'common sense' and she's only pointed out the latest study to support the received wisdom.
Kind of a case study in the difference between inductive and deductive reasoning. Statistical literacy isn't relevant - the epistemological problem lies much deeper than that.
Posted by: Ken Lovell | November 29, 2008 at 11:51 PM
Parker made no mistake. She reported accurately that ISI had found a cause and effect relationship. Indeed, ISI wrote:
"...discovered that the civic knowledge gained from the combination of engaging in frequent conversations about public affairs, reading about current events and history and participating in advanced civic activities is greater than the gain from an expensive bachelor’s degree alone.
Conversely, talking on the phone, watching owned or rented movies and monitoring TV news broadcasts and documentaries diminish a respondent’s civic literacy."
Did ISI mistakenly assume that correlation = causation? Possibly so, but their report doesn't specify what analysis they did, so they may have had a basis for their causal conclusion.
Following up Ken Lovell's comment, although correlation doesn't necessarily imply causation, such an implication can be derived if various other conditions obtain.
Posted by: David | December 01, 2008 at 01:44 PM
David, it seems that you're defending Parker's column by claiming that she merely parroted the ISI's report, which would make her a pretty lousy columnist.
And I think you're going too easy on the ISI. They make the extraordinary claim that people actually get stupider if they talk on the phone, watch TV news or watch documentaries (those things "diminish" civic literacy). Since they don't back up the claim, it's hard to believe they did any analysis at all that would justify that conclusion.
And what is civic literacy? Apparently, it's knowing the answers to a collection of multiple choice questions (mostly definitional) that the ISI considers important.
Posted by: Jinchi | December 01, 2008 at 06:41 PM
Jinchi, I don't know if people get stupider if they watch TV news or documentaries, but I'm convinced they get stupider if they watch American Idol and Dancing with the Stars. Or listen to rap music. Or read US Weekly. Or see Rocky VII. And the list goes on and on.
Posted by: Rob | December 01, 2008 at 08:33 PM