By Trevor Bothwell
web posted May 9, 2005
One of the most beneficial college classes I ever took was a summer graduate course at SUNY Geneseo called Educational Inquiry, Measurement and Statistics. It was taught, ironically enough, by one of the best professors I ever had: Dr. Phillip VanFossen, who has since moved on to Purdue University.
I actually think about (the original) Dr. Phil quite frequently, namely because nary a day goes by where I don't read about yet another biased college professor pushing his or her political views on unsuspecting students (Ward Churchill is simply an example of the extreme). So it's reassuring to recall those who don't make the headlines; the ones who show up to teach kids how to think not what to think. That's the Dr. Phil I remember.
But I thought of the good professor most recently upon reading the headline of a Sacred Heart University press release that was distributed on May 3rd via PRNewswire. It reads: "59.5% Nationwide Call for Tom Delay's (sic) Resignation According to Sacred Heart University Poll." Indeed, the first sentence of SHU's press release states, "According to a national poll by the Sacred Heart University Polling Institute, a majority of Americans surveyed, 59.5%, suggest that Congressman Tom DeLay should resign from his position as House Majority Leader."
Problem is, this headline is severely misleading; and it was Dr. Phil who encouraged me to be on the lookout for deceptive -- and sometimes downright fictitious -- surveys and studies. In fact, he began each lecture by leading our class through analysis of some new study he'd found reported in that morning's newspaper. You know, the kind claiming a linkage between fruit consumption and cancer because 75% of breast cancer patients participating in a research project happened to eat bananas every morning -- only you read on to discover that a grand total of four patients were surveyed for the study. Okay, I'm exaggerating, but you get the idea.
We examined plenty of perfectly sound studies with Dr. Phil that summer, too, but it was his insistence on inquiry and examination -- on never taking for granted what you're initially fed by pollsters, researchers, and the media -- that has always stuck with me. Because you just never know.
Alas, it looks like things haven't changed much since then. I've often been told that newspapers are written at a fourth grade reading level, but the public relations department at Sacred Heart must think only fourth graders read the news. Here's the excerpt of the university's press release detailing the statistics collected on Rep. Tom DeLay:
Nearly half of those surveyed, 46.5%, indicated they are following the U.S. House Majority Leader Tom DeLay's ethics issues. Another 49.1% suggested either not following the issue very closely or not at all.
Among those with an opinion and following the issue closely, 59.5%, suggested that Congressman DeLay resign his leadership position. Another 40.5% said he should not give up his post.
As for resigning from the U.S. Congress, 45.3% with an opinion and following the issue closely suggested he should while 54.7% said he should not resign.
Instead of clarifying that the 59.5% of Americans who favor Rep. DeLay's resignation as Majority Leader belong only to the group with an opinion and following the issue closely, Sacred Heart's PR department claims dishonestly that this percentage reflects the opinion of all Americans. This, my friends, is what you call poll bias.
In reality, 59.5% of only 46.5% of the people surveyed, or 27.6% in all (.595 x .465), actually support DeLay's resignation as Majority Leader. Moreover, when we calculate the 40.5% among the 46.5% who actually have an opinion, we find that 18.8% of all people surveyed do not support DeLay's resignation. And if there's any doubt as to whether Sacred Heart blithely neglected the whopping 49.1% of the population virtually unconcerned with this issue when blaring its 59.5% "majority of Americans" indictment, 27.6% and 18.8% curiously add up to 46.4% -- which looks a heckuva lot like the 46.5% who indicated they're actually following DeLay's ethics issues.
To be fair, I should make clear that the Sacred Heart University Polling Institute likely conducted an unbiased and responsible survey; it was the university's PR office that misrepresented the study, whether intentionally or not. When I spoke with polling institute director Jerry Lindsley about Sacred Heart's misleading headline, he admitted that he "might have suggested it be a bit clearer" had he seen it before the press release was distributed, and admitted that such deceptive reporting risks compromising the ethical standing of polling institutes and universities.
In short, the Sacred Heart University poll is rendered meaningless, as the headline of its press release distorts poll data and misleads readers into believing that almost a full 60% of the country supports Rep. DeLay's resignation from his post as Majority Leader. This is almost 32% more than those who actually do.
Everyday Americans don't necessarily need to take Dr. Phil's statistics class to be able to identify blatant statistical perversions in the media. I just wish I had the same confidence in the people running some of our universities' PR departments, who justify the existence of such courses in the first place. 
Trevor Bothwell is a freelance writer living in Maryland. He maintains a web log at The Right Report and can be contacted at
bothwelltj@yahoo.com . © 2005 Trevor Bothwell
 Printer friendly version |
|