On Thursday evening, ABC News director of polling Gary Langer appeared on behalf of the Institute for Social and Economic Research and Policy (ISERP) to deliver a lecture on the Whys and Wherefores of Newspolling. Lacey Harris-Coble, the Premiere Poobah of lecturehopping, reports.
Seen the real poll numbers on Obama’s ratings post-SOTU? Bullshit – unless, of course, they came from ABC News and were inspected by its celebrated data cop, the Premier Poobah of newspolling, Gary Langer. A two-time News & Doc Emmy winner and a leader in the newspolling industry, Langer, a pleasant fellow, arrived on campus and did not hesitate to begin bashing the omnipresent Internet click-in polls, which he calls manufactured data that easily “acquiesces to biases.”
That’s right – the data on the news is not real data, says Langer, who claims to be one of the few men left in news with standards and who can allegedly guarantee news reporters good, credible data while “shit-canning the bad.” To Langer, bad data is an epidemic spreading everywhere from CNN to The New York Times, the American Medical Association (AMA), and even occasionally ABC News itself (but only when the Good Morning, America hosts can do the fact-checking, of course). So where does this ubiquitous bad data come from? People like you! You’ve seen those pop-ups asking you to sign up to be an online survey-taker and get paid for it; well, some people actually do sign up, collectively completing thousands of surveys each day. The results of these convenience surveys are presented and sold to the news companies as accurate statistical fact, and the news companies in turn present the data as truth.
Take for example this gem from the AMA: 84% of college-age women increase their binge drinking and sexual activity over spring break. While it wouldn’t be particularly surprising if this were true, anyone who had bothered to check the data (someone like Gary Langer) might have found that 71% of the women taking the survey were not in college and/or had not taken a college spring break trip, and of the 29% remaining, only 4% had reported an increase in sexual activity. Granted, this realization does not negate the AMA’s statistic per se, but from Langer’s angle, the fact that 84% of these college-age women aren’t necessarily in college and don’t necessarily do spring break-y things over break suggests that they merely expect that alcohol- and sex-related behaviors would increase. Not exactly the scientific certitude one might expect from the AMA.
Still wondering about those latest Obama ratings? Mr. Langer revealed the truth, so far as he can reckon it: Obama shows a surprising similarity to Ronald Reagan in his first year – a strong 88% approval. As Langer said, “Reagan is misremembered as one of the most popular presidents.” So next time you hear people gloat over Obama’s apparent fall from popular grace, you can smugly tell them that Reagan’s were nearly the same while he was in office. Also, tell them to stop acquiescing to biases.