WEBLOG

Previous Month | RSS/XML | Current | Next Month

Check 'Em Out

(4/25/2006) An interesting article on psychological research on bias:
A Princeton University research team asked people to estimate how susceptible they and "the average person" were to a long list of judgmental biases; the majority of people claimed to be less biased than the majority of people. … Dozens of studies have shown that when people try to overcome their judgmental biases―for example, when they are given information and told not to let it influence their judgment―they simply can't comply, even when money is at stake. And yet, if decision-makers are more biased than they realize, they are less biased than the rest of us suspect. Research shows that while people underestimate the influence of self-interest on their own judgments and decisions, they overestimate its influence on others.

Source: Daniel Gilbert, "I'm O.K., You're Biased", New York Times, 4/16/2006
Ben Goldacre on appeals to expert opinion:
Authority is a shortcut to reliable information. You take stuff on faith because reading, critiquing and checking that academic references are valid and represent the material they refer to, and more, is very time consuming. I started out dubious about his claims. And after an afternoon in this hall of mirrors, I am in no sense motivated to spend my time reading a whole book on them.

Source: Ben Goldacre, "When in doubt, call yourself a doctor", Bad Science, 4/22/2006

Blurb Watch: "Scary Movie 4"

According to a newspaper ad for the new "Scary Movie" sequel, one critic pronounced it:

"HILARIOUS!"
-Chris Kaltenbach, THE BALTIMORE SUN

However, if you check Kaltenbach's review, the only place that the word "hilarious" occurs is here (emphasis added):

If you're Airplane!, of course, it doesn't matter that no one realizes you're parodying an old Dana Andrews flick, Zero Hour! The jokes are so consistently hilarious, and so relentless, that the source material doesn't really matter. Scary Movie 4 doesn't enjoy that luxury―something director David Zucker, who 26 years ago was one-third of the writer-director team on Airplane!, should have realized.

This is the worst contextomy that I've ever seen in a movie ad blurb, which is saying something. The word "hilarious" isn't describing "Scary Movie 4" at all, but the earlier movie "Airplane!", to which the current movie is being unfavorably compared.

However, while there is only the one occurrence of the word "hilarious", the adverb "hilariously" also occurs in the following passage (emphasis added):

Anna Faris, her deadpan comic timing still a joy to watch, returns as Cindy Campbell, one of two main holdovers from the first three movies…. Still a naive, ever-trusting island of cluelessness in a sea of utter insanity, she blithely, hilariously goes about her business while the jokes fly willy-nilly around her.

I suppose that the ad writer might claim to have been quoting this while dropping the suffix. If this is the source of the blurb, then it isn't quite as bad a contextomy as quoting a favorable word that describes a different movie. Instead, it quotes a favorable word that describes only part of the movie, which is one of the oldest tricks in the ad writer's bag.

Sources:

Ad for "Scary Movie 4", Go!, 4/21/2006, p. 4
Chris Kaltenbach, "Fun…But Done", Baltimore Sun, 4/14/2006

April 20th, 2006 (Permalink)

Close Enough for CNN

Here's the first paragraph of a recent CNN report:

Oil prices jumped above $72 a barrel to yet another record Wednesday after a government report said supplies of crude made a surprise decline and gasoline stocks fell far more than expected.

And here's the seventh paragraph (emphasis added):

Oil has been hitting record highs in recent sessions, unadjusted for inflation, on supply worries fed by fears of a confrontation with Iran, the world's fourth-biggest producer. But it's also within sight of inflation-adjusted highs of around $80 a barrel set in the late 1970s and early 1980s following the gas crisis and the Iranian revolution.

This is why you should always read a news story all the way to the end, because sometimes the later paragraphs take back what was said in the headline or first paragraph. But it gets better―or worse, depending on how you look at it: look at the Reuters chart included as a sidebar. There you will see that the inflation-adjusted price of oil in 1980 was actually closer to $90 a barrel than to $80, which the article claims. According to the U.S. Energy Information Agency, Reuters is correct, though the record appears to have been set in 1981. So, the current nominal "record" high price is in fact over $15 less per barrel than the real record. I guess that's "within sight", though, if you're a CNN reporter.

Source: Steve Hargreaves, "Oil Jumps Above $72 to New High", CNN Money, 4/19/2006

Resource: "Real Petroleum Prices", Energy Information Agency, 4/2006

April 19th, 2006 (Permalink)

A Time Travel Puzzle

A traveller left Rotterdam on September 9th. After a two day voyage, he arrived in London on August 31st. After staying for a week, he left London on September 18th. After a three day voyage, he arrived in Stockholm on September 10th. Four days later, he left for home. After another two day voyage, he arrived in Rotterdam on September 27th. What year was it?

If you think that you can answer the puzzle, send your solution to Doctor Who? by April 1st, 2006. The Grand Prize is your very own TARDIS. All correct entries received after the deadline but before the end of the month will win an all-new, 100% natural, official Fallacy Files Boobyprize™, worth up to $195.99.

Solution

Acknowledgment: Thanks to John Congdon for the puzzle.

April 18th, 2006 (Permalink)

Check it Out

"The Numbers Guy"'s latest column is, as usual, worth reading as a whole, but here are a couple of specific points of interest:

His first item concerns an inflated number used by a special interest group to draw attention to its issue. Such groups often try to masquerade as disinterested experts, but they have a strong motivation to come up with dramatic numbers, and less interest in accuracy, so the numbers they produce should be treated with great caution. When the statistics aren't simply made up―as this one appears to have been―they are often inflated by using low redefinitions of the key terms.
His second item concerns another way in which interest groups try to influence issues with statistics, namely, by commissioning polls. With polls, the interest of these groups is in showing that large numbers of people are very worried about the group's issue. If the inflated numbers that they have released to the press haven't succeeded in alarming sufficient numbers of people, then they can try to inflate the poll results. For this reason, any poll paid for by such a group should be viewed with skepticism, though it may still be accurate if conducted by a professional pollster with a reputation to protect. The specific technique used in this case was to word questions to lead to the desired response; for example, here's one that tries to get the respondent to "jump on the bandwagon":
"More than 80% of Americans believe that gambling is a question of personal choice that should not be interfered with by the government. Do you agree or disagree that the federal government should stop adult Americans from gambling with licensed and regulated online sports books and casinos based in other countries?" You probably won't be surprised to learn that after being told that most Americans don't want the government to interfere, some 71% of the respondents to this question signaled they, too, were against a government ban.

Update (4/26/2006): The Mystery Pollster has an update on this poll, which turns out to have more problems than just leading questions:
[T]he survey release failed to disclose that it was conducted online and made the statistically indefensible claim to be a "scientific poll" with a "margin of error." The failure to disclose the use of sampling from an online panel is particularly deceptive given that an online activity was the focus of the survey. Add that all up and you get one remarkably misleading poll release.

Source: Mark Blumenthal, "An Online Poll on an Online Activity", Mystery Pollster, 4/26/2006

Source: Carl Bialik, "Measuring the Child-Porn Trade", The Numbers Guy, 4/18/2006

April 9th, 2006 (Permalink)

Headline

Colon a Little Hesitant

April 6th, 2006 (Permalink)

About Us

We specialize in strategic visionary innovative transformational solutions that leverage users, processes and information to enable change and attain best practice service and productivity. We easily comprehend, and advise on, vision and strategy. We possess the ability to quickly conceive strategic applications and process changes that enable attainment of strategy. We have delivered several solutions targeting user experience. We have extensive experience developing strategies, conceiving end-to-end solutions, and driving delivery to exceed intended results. Solutions are innovative, strategic, and typically improve delivery or internal productivity to best practice levels. We leverage our extensive experience to create, expand or revise strategies as needed to compete. Areas of expertise include the following:
User-centered solutions that maximize user experience, drive revenue and reduce cost.

Solutions that seamlessly integrate enterprises.

Our own next-generation approach to systems delivery that delivers order-of-magnitude improvements in time to market and ability to exceed expectations.

Speaking of doublespeak―or double speaking of it―the above blather is a Reader's Digest version of an egregious example of inflated business language that I recently came across. It's so full of hot air that it doesn't really mean much, but it might translate into English as: "We can help you to make much more money than you thought you could." So, if you need your enterprises seamlessly integrated, your user experiences maximized, and your strategies leveraged, we're the ones for you!

April 5th, 2006 (Permalink)

Doublespeak Watch

Slate has an interesting article on a popular mistake about military doublespeak:

Countless news articles and blog entries over the past two and a half years have claimed that "transfer tube" is the new Pentagon-speak for "body bag." … But the U.S. military does not refer to body bags as either "transfer tubes" or "transport tubes." Mortuary suppliers have been using the designation "pouch, human remains" since at least 1965, and the Pentagon has recognized "human remains pouch" (or HRP for short) as the official term since the first Gulf War.

"HRP" is actually better military doublespeak than "transfer tube". For one thing, the military loves initials, as in "MRE", which stands for "Meal, Ready-to-Eat". Moreover, both "body bag" and "human remains pouch" are uncomfortably explicit about what they're for, but "HRP" calls up no images of dead bodies, unless you happen to know and think about what the letters mean.

Source: Benjamin Zimmer, "How Does the Pentagon Say 'Body Bag'?", Slate, 4/4/2006

April 4th, 2006 (Permalink)

Sentence First―Verdict Afterwards

Rebecca Goldin of STATS has an excellent analysis of a recent Associated Press article on anti-pornography efforts. Goldin's analysis is worth reading as a whole, but I have a few passing comments on the AP article:

…[I]f you're a consumer [of pornography], John Harmer thinks you're damaging your brain. Harmer is part of a cadre of anti-porn activists seeking new tactics to fight an unprecedented deluge of porn which they see as wrecking countless marriages and warping human sexuality. They are…raising funds for high-tech brain research that they hope will fuel lawsuits against porn magnates. "…I'm convinced we'll demonstrate in the not-too-distant future the actual physical harm that pornography causes and hold them financially accountable. That could be the straw that breaks their back."

This is what's called "advocacy research" or "junk science", since it's funded by advocates who've already decided what results they want.
"The Internet is the perfect delivery system for anti-social behavior…," said Mary Anne Layden, a psychologist and addiction expert at the University of Pennsylvania. … She says many of her patients, rather than improving their sex lives with porn, suffer sexual dysfunction.

As Goldin indicates, the patients seen by an addiction expert are not a representative sample of people who use pornography. "The clinician's error" is the mistake of judging that those who are seen by a clinician are representative of the population of people with a condition, when doctors and psychologists see those who have the worst problems.
Interest in porn is age-old and normal, says psychologist David Greenfield of West Hartford, Conn., an expert on Internet behaviors, but it can become a destructive obsession for a minority who indulge in it at the expense of healthy relationships. … He estimates that for up to 10 percent of porn users, relationships suffer with many husbands spending so much time online that they cease to have sex with their wives.

"Up to" are weasel words that you often see in advertisements: "lose up to 30 pounds!", "save up to 50%!", "win up to a million dollars!" However, zero pounds is up to 30 pounds, 0% is up to 50%, and $0 is up to a million dollars, so the advertiser is not really committed to anything. The phrase "up to" simply puts an upper limit on an amount. In the same way, "up to 10 percent" merely puts an upper limit on the number of users who are abusers, but the actual percentage may be much lower.
…Morality in Media president Robert Peters, a Dartmouth-educated attorney[,]…struggled in his 20s to kick a porn habit that started in grade school. "It was hell," said Peters, recalling a six-year stretch where he regularly visited porn outlets on New York's 42nd Street. "It's a very hard habit to break."

The article gives anecdotes such as this, which is standard journalistic practice, and you would be hard-pressed to find any news articles on social problems without at least one such poignant example. However, anecdotes are very poor evidence, but a powerful psychological boobytrap which can lead the reader to commit "the anecdotal fallacy"―also known as "the Volvo fallacy". This is the tendency to remember such anecdotes more vividly than dry statistics, and to overestimate the prevalence of a problem as a result of the vividness of such stories.

Sources:

Lewis Carroll, "Alice in Wonderland", Chapter 12: "Alice's Evidence"
David Crary, "Activists Lament Porn's Move to Mainstream", Associated Press, 4/2/2006
Rebecca Goldin, "Porn Causes Brain Damage", Statistical Assessment Service, 4/4/2006

Resources:

Book Review: Voodoo Science, 11/30/2004
The Clinician's Error, 3/3/2005

April 1st, 2006 (Permalink)

Reader Response

"I've tried to keep quiet, but I just can't hold it in any longer. I have to tell everyone that the Fallacy Files has graduated from occasionally exempting itself from the few principles it has to betraying them altogether. Here's a quick review: To make sure you understand, I'll spell it out for you. For starters, the Fallacy Files is not interested in what is true and what is false or in what is good and what is evil. In fact, those distinctions have no meaning to it whatsoever. Why? We must obviously ask ourselves questions like that before it's too late, before the Fallacy Files gets the opportunity to mold your mind and have you see the world not as it is, but as it wants you to see it. I am making a pretty serious accusation here. I am accusing the Fallacy Files of planning to abandon the idea of universal principles and focus illegitimately on the particular. And I don't want anyone to think that I am basing my accusation only on the fact that its arguments would be a lot more effective if they were at least accurate or intelligent, not just a load of bull for the sake of being controversial. The next time the Fallacy Files decides to divert us from proclaiming what in our innermost conviction is absolutely necessary, it should think to itself, cui bono?―who benefits? As a parenthetical note, I want to give people more information about the Fallacy Files, help them digest and assimilate and understand that information, and help them draw responsible conclusions from it. Here's one conclusion I certainly hope people draw: The Fallacy Files can't attack my ideas, so it attacks me. It could be worse, I suppose. It could blow the whole situation way out of proportion. Finally, this has been a good deal of reading, and undeniably difficult reading at that. Still, I hope you walk away from it with the new knowledge that the Fallacy Files is consistently inconsistent."

Source: "My Complaint about the Fallacy Files", 4/1/2006

Solution to Time Travel Puzzle (5/2/2006): 1752. Here is John Congdon's explanation of the answer:

The key to the puzzle is the change from the Julian calendar to the Gregorian calendar. All the Netherlands had adopted the Gregorian calendar by the end of 1700, so our traveler left Rotterdam on September 9, 1752 (Gregorian). Great Britain was still on the Julian calendar when he arrived on September 11 (Gregorian), so the official date was still August 31 (Julian). However, Great Britain changed from Julian to Gregorian three days later, so September 2 (Julian) was followed by September 14 (Gregorian). Therefore, when our traveler left London, the date was September 18 (Gregorian). When our traveller arrived in Stockholm, the date in London was September 21 (Gregorian). However, since Sweden did not change to the Gregorian calendar until the following spring, September 21 (Gregorian) was still September 10 (Julian) in Stockholm. When he left Stockholm and returned to Rotterdam, he also returned to the Gregorian calendar, arriving on September 27 (Gregorian), having been away 18 days.
In other words, our traveler did not travel forwards and backwards in time. He only travelled forward in time, at the same rate as the rest of us, but into and out of places that did not reckon the date the same way as each other.

Thanks again to John, and congratulations to the following winners who determined the correct year: Paul Farrington, Steven Haddock, Kranthi Marupaka, Przemys�aw P�askowicki, and Vamshi. An honorable mention goes to John Lavett. The boobyprizes have been shipped and should have arrived sometime last year.