Goffman and Veritas

June 9, 2014
Posted by Jay Livingston

Slate ran an article  by L.V. Anderson decrying the tendency of Ivy League graduates to be vague about their educational credentials.  Asked where they went to school, they say, “New Haven,” or “Boston,” or “New Jersey.”

If . . . you refuse to tell someone you went to Harvard, that reflects poorly on you – it implies that, on some level, you buy into the overblown mythos of Harvard and the presumption of Ivy League superiority.

Anderson’s course of study, wherever it was, must not have included even a paragraph of Goffman. One of the basic ideas of Presentation of Self is that people seek to control the impressions others make of them, and they do this by controlling the information others get. It’s not about what mythos Ivy Leaguers buy into. It’s about the mythos others have bought.

Ivy Leaguers have a very good notion, usually based on experience, of the impression that “Harvard” or “Yale” creates in others’ minds.  Alyssa Metzger in the Chronicle  sets the record straight.

When I would visit my former local bar in Philly . . . a reply of “In Boston” usually led to them returning to their beers with an “Oh cool … my friend’s sister goes to BU” . . .  If I said, “At Harvard,” it tended to lead to them turning on their stools to face me, wide-eyed, with an “Oh wow … you must be really smart.” I wasn’t Allyssa, I was SMART PERSON (TM)— more object than person.

Who wants to be seen as an exemplar of a stereotype? And stereotype we do, even those of us who should know better.  A few years ago (here) I reported a conversation from my playground days. I had gotten to know another playground dad (weekdays at the playground, the dad sample is a very small n). Brad was a Juilliard grad who was eking out a living as a conductor with a regional orchestra – five concerts a year. 

One day we were sitting on the bench, and Brad asked me where I’d gotten my Ph.D. I guess we’d never talked much about higher education. Harvard, I told him.

“I didn’t know that,” he said, surprised, “and I’ve known you all this time.”
“Don’t be impressed,” I said.
“But I am,” he said. From his voice and the look on his face, I could see that he meant it. I wanted to convince him not to be.
“Oh Brad,” I said, my voice rising in mock awe, “You went to Juilliard?! You must be this really great and talented musician. Juilliard – wow!” Or something like that.
He laughed.
“See what I mean?” I asked.
“Yes,” he said. Then a pause. “But I’m still impressed.”

Harvard grads don’t want to lie. So they tell the veritas, just not the whole veritas. Yalies may shade the veritas in order to present themselves in the lux that best fits the situation.  But, as Goffman pointed out, that’s what we all do all the time.

The Belmont – No Place Like Place?

June 7, 2014
Posted before post time by Jay Livingston

At Freakonomics Steven Levitt argues for making “a place bet on California Chrome” mostly because the odds to win will be so low.

When California Chrome won the Preakness, a $2 bet to win returned $3. A $2 bet to place also returned $3! . . . You can’t know with certainty what the place payout will be ahead of time because it depends on what other horse finishes in the top two, but if you watch the allocation of money in the place pool you can get a pretty good idea. Sometimes crazy things happen. When Big Brown won the Preakness, he paid $2.40 to win, $2.60 to place, and $2.40 to show!

Levitt is right when he says that the place payout depends on which other horse finishes in the top two. But he’s wrong when he says that you can get a pretty good idea by watching the place pool. The Tote board at the track does show how much money is bet on each horse to win, to place, and to show.  The place payout is determined by taking all the losing bets and dividing them up among people who bet on the winner and those who bet on the place horse.

The reason Big Brown paid more to place than to win was that horse who finished second, Macho Again, at 40-1 was the second longest shot in the race.  That meant more money in the place pool to be divided (all the money bet on the other ten horses) and fewer people that it would be divided among  So if you are betting a heavy favorite to place, you not only have to watch the place pool bets, but you also have to pray that the horses with big money bet on them finish no better than third.

Levitt’s best bet is Commanding Curve to win. The odds will be attractive. A dollar on California Chrome, if he wins, will get you fifty cents; if you bet him to place, you might win only a dime. The morning line on Commanding Curve is 15-1, but I expect it will by lower by post time. Commanding Curve closed six lengths on California Chrome in the final furlong of the Derby, an indication that he might have the stamina for the added quarter-mile of the Belmont.  Commanding Curve also skipped the Preakness, giving him an extra two weeks of rest.

My own long shot is Wicked Strong, another possible closer. The morning line is 6-1, but I predict it will be higher. He had some bad racing luck in the Derby and still got fourth.

Finally, I cannot do a post on horse racing without reiterating my pet peeve about the incorrect use of “track record” that has become so widespread (see my earlier post here).  In racing, where the term originates, it does not refer to a horse’s past performances. It refers to the record time at that track for a given distance. People don’t have track records, tracks do. The Belmont stakes is a mile and a half.  The fastest time for that distance at Belmont – the track record – is 2:24.  That’s way fast, and here’s what it looked like.

In a sport where the difference between win and place is usually a fraction of a second, Secretariat is four or five seconds ahead of the rest.

UPDATE: Both Levitt and I were wrong. The winner was Tonalist at 8-1, a horse who had raced only four times and only once against top horses, though he won that one (a grade-2 stakes). The place horse was an even longer shot, Commissioner at 20-1.

Game, Set, Match.com

June 5, 2014
Posted by Jay Livingston

Usually, you want to match up with someone at about your level, or a little higher.  The trouble is that many people overestimate their own level.  Maybe that’s especially true of men.

One summer many years ago at the tennis courts, a guy I didn’t know  came over and asked me if I’d like to play. I hadn’t arranged a game with anyone, but I didn’t want to wind up playing some patzer. 
“Are you any good?” I asked. He paused.
“Well, I’m not Jimmy Connors,” he said (I told this was many years ago), “but neither are you.” 

In chess and other games, serious players have ratings. Give a roomful of possible partners, they can sort through the ratings and find a match with someone at roughly the same level.  It’s called assortative mating, though that term usually refers to the other kind of mating, not chess.  It’s the basis of the conflict in this poignant scene from “Louie.” (The scene was also played on a recent  “Fresh Air” interview with Louis C.K.)

Vanessa is not a ten, neither is Louie.  According to principles of assortative mating, the tens will wind up with other tens, the nines with nines, and so on down the attractiveness scale. One problem in the “Louie” scene is that Louie seems to have an inflated view of his own attractiveness.  He’s aiming higher than Vanessa.  That’s typical.  So is the importance that Louie, the man, places on physical attractiveness. This excerpt begins with Louie telling Vanessa that she’s a really beautiful . . . . He can’t bring himself to say “girl”; he was probably going to say “person”  or worse, “human being.” In any case, he’s obviously not saying what he thinks.

Or as Dan Ariely and colleagues concluded from their study of HotOrNot members (here)*

[Men] were significantly more influenced by the consensus physical attractiveness of their potential dates than females were. [Men also] were less affected by how attractive they themselves were . . .  In making date choices, males are less influenced by their own rated attractiveness than females are.

Another dating site, OK Cupid, found a similar pattern when they looked at data about who gets messages (here).**  They asked their customers to rate profile photos of the opposite sex on a scale of 0 to 5. They then tracked the number of messages for people at each level of attractiveness.  The graph below shows what women thought and what they did – that is, how attractive they found men, and who they sent messages to.

Men who were rated 0 or 1 got fewer messages than their proportion in the population.  That figures. But even men who were only moderately attractive got more than their share. Generally, the fewer men at a level of attractiveness, the fewer total messages women sent. The male 4s, for example, constituted only 2% of the population, and they got only 4% of all the messages. The Vanessas on OK Cupid are not sending a lot of inquiries to guys who look like George Clooney.

But look at the men.

Men are more generous in their estimates of beauty than are women. But they also ignore the Vanessas of the world (or at least the world of OK Cupid) and flock after the more attractive women, even though there are fewer of those women.  Only 15% of the women were rated as a 4, but they received about 26% of the messages.  Women rated 5 received messages triple their proportion in the population. 

What about those with so-so looks?  Women rated as 2s received only about 10% of the messages sent by men. But men at that same level received 25% of the messages women sent.  The women seem more realistic.

Vanessa too has no illusions about her own attractiveness. She refers to herself as “a fat girl,” and when Louie, trying to be kind, says, “You’re not fat,” she says: “You know what the meanest thing is you can say to a fat girl? [pause] ‘You’re not fat.’”*** But it’s only when she challenges Louie’s view of his own attractiveness that their relationship starts to change.

Y’know if you were standing over there looking at us, you know what you’d see?
That we totally match. We’re actually a great couple together.

She doesn’t explain what she means by “totally match.” It could be their interests or ideas or personalities, but the imaginary stranger looking at them from over there couldn’t know about any of that. What that generalized other could see is that they are at roughly the same place on the assortative mating attractiveness scale.

* Ariely discusses this research in his book The Upside of Irrationality.

** OK Cupid was founded by Harvard math graduates. On the Website’s blog, they would post graphs like these – big data that could be very useful for the site’s members. A few years ago, they sold out to Match.com, and the data analyses ended.

*** This occurs early in the clip, at about 0:25. The entire 7and a half minutes is worth watching.

Tide and Time

June 4, 2014
Posted by Jay Livingston

Survey questions, even those that seem simple and straightforward, can be tricky and yield incorrect answers.  Social desirability can skew the answers to questions about what you would do – “Would you vote for a woman for president. . . .?” and even factual questions about what you did do.  “Don’t ask, ‘How many books did you read last year?’‘ said the professor in my undergraduate methods course. “Ask ‘Did you read a book last week?’” There’s no shame in having been too busy to read a book in a seven-day period. Besides, people’s recall will be more accurate.  Or will it? Is even a week’s time enough to distort memory?

Leif Nelson (Berkeley, Business School) asked shoppers, “Did you buy laundry detergent the last time you went to the store?” Forty-two percent said yes.

Nelson doesn’t question the 42% figure. He’s interested in something else:  the “false consensus effect” – the tendency to think that others are more like us than they really are.

So he asks, “What percentage of shoppers do you think will buy laundry detergent?” and he also asks “Did you buy laundry detergent.” Sure enough, those who said they bought detergent give higher estimates of detergent buying by others. (Nelson’s blog post, with other interesting findings, is here.)

But did 42% of those shoppers really buy detergent last time they were in the store? Andrew Gelman is “stunned” and skeptical. So am I.

The average family does 7-8 washes a week. Let’s round that up to 10.  They typically do serious shopping once a week with a few other quick express-lane trips during the week.  This 50 oz. jug of Tide will do 32 loads – three week’s of washing.

That means only 33% of customers should have said yes.  And that 33% is a very high estimate since most families buy in bulk, especially with items like detergent. Tide also comes in 100-oz. and 150-oz. jugs.

If you prefer powder, how about this 10-lb. box of Cheer? It’s good for 120 loads. 

A family should need to buy this one in only one out of 12 trips. Even at double the average washing, that’s six weeks of detergent. The true proportion of shoppers buying detergent should be well below 20%.

Why then do people think they buy detergent so much more frequently?  I’m puzzled.  Maybe if washing clothes is part of the daily routine, something you’re always doing, buying detergent seems like part of the weekly shopping trip. Still, if we can’t rely on people’s answers about whether they bought detergent, what does that mean for other seemingly innocuous survey questions?