Decision Science News

DOES EVIDENCE-BASED MEDICINE SUPPORT PARACHUTE USE?

In 2003, the characteristically less-than-hilarious BMJ published a satirical article entitled “Parachute use to prevent death and major trauma related to gravitational challenge: systematic review of randomised controlled trials.” It has its funny parts. The authors’ point is summarized in the abstract’s conclusion:

Conclusions: As with many interventions intended to prevent ill health, the effectiveness of parachutes has not been subjected to rigorous evaluation by using randomised controlled trials. Advocates of evidence based medicine have criticised the adoption of interventions evaluated by using only observational data. We think that everyone might benefit if the most radical protagonists of evidence based medicine organised and participated in a double blind, randomised, placebo controlled, crossover trial of the parachute.

Photo credit:www.flickr.com/photos/fongetz/1789740815/

JUNIOR PROFESSORSHIP AT CARNEGIE MELLON UNIVERSITY

The Decision Science News editor got his start at Carnegie Mellon University’s Children’s School and is happy to let readers know that the Department of Social and Decision Sciences at Carnegie Mellon University seeks candidates to fill a junior-level tenure track position in behavioral decision research.

Candidates must demonstrate a strong track record and have research interests related to the psychology of judgment and decision making (e.g., social psychology, cognitive psychology, social cognition, emotion, self-regulation, neuroscience, consumer behavior). We prefer candidates who can teach courses in the psychology of judgment and decision making, empirical research methods, or otherwise participate in the department’s undergraduate majors in Decision Science, Policy, and Management. Joint appointments are possible with other units on campus.

The department is interdisciplinary, including psychologists, economists, political scientists, and historians. It has particular research strengths in behavioral decision research, policy analysis, industrial organization, technological change, and computational social science. Current projects and faculty are described at the web site:

http://www.hss.cmu.edu/departments/sds/

Carnegie Mellon University is an Affirmative Action/ Equal Opportunity employer. We encourage minorities, women, and individuals with disabilities to apply.

Applicants should send a CV, reprints/preprints of peer-reviewed publications, 3 or 4 letters of recommendation, a statement of research interests, and a cover letter to:

Chair, Behavioral Decision Theory Search Committee (Psychology)
Carnegie Mellon University
Department of Social and Decision Sciences
208 Porter Hall
Pittsburgh, PA 15213-3890.

Review of applications will begin October 2nd. Applications received by October 1st are assured full consideration.

CMU-related posts on Decision Science News:

• Major choice strategies
• Looking for a good introduction to decsion science?
• The social dilemma of exaggerating the positive
• Robyn Dawes
• Honesty, No longer the best policy?
• Benjamin Kleinmuntz

GERD GIGERENZER ON IGNORING INFORMATION FOR BETTER DECISIONS

Who: Gerd Gigerenzer, Director, Max Planck Institute, Berlin
What: The Rationality of Heuristics: Ignoring Information for Better Decisions
Where: Westminster Business School, Hogg Lecture Theatre
When: 17h15-19h

The academic year in London will get off to a stimulating start as one of Psychology’s leading intellectuals, Gerd Gigerenzer, will take the stage on Tuesday September 23rd, 2008 to kick off the Economics of Behavior and Decision Making seminar in London.

Gigerenzer is Director of the Center for Adaptive Behavior and Cognition at the Max Planck Institute for Human Development in Berlin and former Professor of Psychology at the University of Chicago. He won the AAAS Prize for the best article in the behavioral sciences. His book Gut Feelings: The Intelligence of the Unconscious was one of six nominees for the 2008 Royal Society Prizes for Science Books. He is the author of Calculated Risks: How to Know When Numbers Deceive You, the German translation of which won the Scientific Book of the Year Prize in 2002. He has published several other academic books on heuristics including, Simple Heuristics That Make Us Smart (with Peter Todd & The ABC Research Group) and Bounded Rationality: The Adaptive Toolbox (with Reinhard Selten, a Nobel laureate in economics), Heuristics and the Law (Dahlem Workshop Reports) (with Christoph Engel), Rationality for Mortals: How People Cope with Uncertainty (Evolution and Cognition), and Adaptive Thinking: Rationality in the Real World (Evolution and Cognition Series).

ABOUT GIGERENZER:
Gerd Gigerenzer’s CV
Through Analysis, Gut Reaction Gains Credibility, The New York Times.
Smart Heuristics: Gerd Gigerenzer at Edge.org
Gigerenzer’s Books

NEW FACULTY OR GRAD STUDENT IN LONDON? JOIN THE EBDM SEMINAR EMAIL LIST:
To subscribe to the seminar series email list, please visit http://tinyurl.com/yvw2sr to opt in. You can easily unsubscribe anytime. Please pass this message on to those who may be interested in joining the email list.

The full schedule of talks at the Economics of Behaviour and Decision Making seminar series is maintained at http://www.decisionresearchlab.com/ebdm/

ASKING A LIBRARIAN

There is an interesting post and discussion at Statistical Modeling, Causal Inference, and Social Science about whether the Library of Congress stacks were ever open to the public. Andrew has detailed memories of roaming the stacks, including information on how they were configured (“like spokes on a wheel”), which if correct would suggest that he was in there. However, the commenters’ personal experience and Web research finds evidence of the stacks being closed. And when Decision Science News was a young Web log going to high school in Potomac, Maryland in the 1980s, it remembers being denied access to the stacks. Today, the grown-up blog has learned that questions about libraries are often best handled by librarians, and has made use of the “Ask a Librarian” features on the LOC web site. The response:

This is in response to your inquiry of August 14, 2008.

The book stacks of the Library of Congress General Collections are not open to the public. There are 75,000 titles of reference material in the Main Reading Room, some shelved on decks that are directly adjacent to this room, that self-serve for all researchers in the reading room.

Prior to 1989, temporary stack passes were granted to select researchers who provided information that such access was necessary to their research. The stacks were closed after this time due to the need to secure the stacks and protect the book collections. Additional information about the closed stack system is accessible at http://www.loc.gov/rr/security/stacks.html/

You and your friend are most welcome to come to the Library, register as researchers in LM140 (Madison Building), and use the reading rooms for research purposes. If you have any additional questions about the Library’s collections and services, please use the Ask-A-Librarian service again.

Thank you for using the Library’s Ask-A-Librarian service.

Sincerely,

Public Service & Collections Access Officer
Collections Access, Loan, & Management Division
Library of Congress
Washington, D.C. 20540-4630
202-707-7400

This blog hypothesizes that Andrew, probably a prodigious student, got a researchers’ pass. This hypothesis recently got some support as Andrew added, “They definitely let me in the stacks when I was in high school. But I don’t know the official policy. Maybe I had to fill out some sort of form to get into the stacks; I don’t remember.”

LEARN VI AND VIM BY WATCHING AND TYPING ALONG

For this week, Decision Science News has created a vi and vim video tutorial. (The tutorial is best viewed in your browser’s full-screen mode, try pressing F11 in Windows). Vim is a free and open source editor. DSN highly recommends vim.

This tutorial is for absolute beginners and probably should not even be looked at by people who already know vi or vim.

After finishing the tutorial, try this …

ADDITIONAL EXERCISE:
Find 4 ways to delete the last sentence typed (“End of the line”).

ANSWERS
(Try each of these, typing “u” after each to undo the deletion so that you can try the others.)

1) Move into the sentence with the movement keys and type “das” (delete a sentence).
2) Type dollar sign “$” to go to the end of the line and then “4db” to delete backwards 4 words, since there are 4 words in the last sentence.
3) Go to the end of the sentence as above, then hit “d(” to delete the sentence to the left.
3) Move to the beginning of the sentence with “(” then type “d)” to delete the sentence to the right.

BONUS: HOW TO EXIT AND SAVE FROM VI / VIM
From edit mode “:w filename.txt” will save your file. To quit vim, type :q. To quit without saving type :q!

We have received many questions about why one would want to learn vim in this day and age. A good answer is here.

For reference, the text of the tutorial follows:

Welcome to the Decision Science News vi and vim video tutorial.

To begin the first thing we will do is type “i”. This puts us in insert mode, which means the mode in VIM in which you can enter text. So type something, such as:

“I would like three eggs over easy.

Oh, and some coffee, too.”

Now that you have entered this text, you can move around a bit with the arrow keys and do some crude editing. For example, I can backspace over words and type them again, and the delete key also works.

However, this is generally *not* the way you edit text in VIM. Generally, you switch to “edit mode”. Do this by hitting ESCAPE or Control-OpenBracket Ctrl-[. I will do that now. Notice how the word “INSERT” has disappeared from the bottom of my screen. Now I’m in edit mode. If I hit “i” I’ll go back to insert mode, and ESCAPE takes me to edit mode: i ESC i ESC i ESC

The thing that really confuses people about edit mode is that you can’t just enter text in it. You can only enter special vim commands. We’ll go over some of the most common vim in this tutorial.

First, movement commands. To move down in edit mode, hit “j”. Notice how “j” looks like a down arrow. To move up, hit “k”, it makes sense that up would be right next to down. Since j and k are home keys, you can move up and down quickly.

To move left, hit “h”, which makes sense because it’s to the left of the up and down keys.

To move right, hit “l”. l, h, j, k. That is space by space movement.

Now, let’s give ourselves some more text to work with. Move up to your first line of text and hit “yy”. The “yy” command yanks a line of text, kind of like copying it. Now press “p”. “p” is like “paste” and you’ll see that it pastes the copied line, so we now have our first line twice.

To do a command many times in vi/vim, just type a number and then that command. For example, type 5p and you’ll see that it pastes 5 times.

Similarly, if we type “99p” it will paste 99 times. Try it “99p” I’ve just put 99 lines there. Yikes, how to undo it? Two ways. First type “u” and you’ll undo the last action. If you type “u” the 99 lines are gone. Another way would be to use the delete lines command, which is “dd”. I’ll type “99p” to paste back in the 99 lines, and now what I’ll do is type “99dd” and I just executed the “dd” command (delete line) 99 times. “dd” again is delete line.

How do you join lines? Simple, you use capital J. Up at the top, I can use captial J to put all these sentences into one line: J J J J J. Now you have this long line. I’ll do yy to yank this long line and I’ll hit “p” to paste it. Now we have some text to edit.

We’ve looked at moving left and right character by character, but there is a better way. If we type “w” we skip to the next word or punctuation mark (w w w w) and if we type “b” we move to the previous one: b b b w w w b b b. To jump to the next sentence we hit Close-paren ) Notice how close-paren ) looks like a right arrow. Similarly, Open-paren ( looks like a left arrow and jumps to the previous sentence.

To go to the beginning of the current line, type 0. To go to the end type dollar sign $: 0 $ 0 $ 0 $

Now lets delete some words. Hit “5b” to move back 5 words. Now to delete one word to the right type “dw”. To delete a word to the left you would type “db” (remember that b refers to the word to the left). To delete a sentence, move into the sentence and type “das”. Sentence is gone! Another way to delete a sentence, say the sentence to the left is to type d open-paren “d(” deleted the sentence to the left. I’m going to move back a sentence now to show that if you want to delete a sentence to the right use d close-paren: “d)”

To delete just one character, simply type x. You can use the delete key.

Ok, so we’ve done a bunch of stuff in edit mode, but you probably want to get back to actually typing something. How do you return to insert mode?

There are a few ways. We’ve already talked about i, this will let me insert before the cursor. So we see here the cursor is on the l in “like”, if I type “i” and start typing, I’m inserting before the cursor. I hit ESC to go back to command mode. If I hit “a”, I can start typing after the cursor. And I’ll hit ESC again to go back to command mode.

Now, let’s say I want to insert something on the line below the current line. If I just type “o” I can start inserting on the line below. Now I can type “Hi Dad” and hit ESC to go back to edit mode. To start typing on the line above, I type capital O and hit ESC to go back to command mode. If you have trouble remembering which is above and which is below, think of the Os as filled with helium. The big O will hold more helium and go above the lower case o.

To start inserting at the beginning of the line, type capital letter “I”. I’ll type “Start of line” and type ESC to go back to edit mode. To start inserting at the end of the line and type capital A (think A for “After” the line). I’ll type “End of the line” and then ESC to go to edit mode.

GET A FEELING FOR ACCIDENTAL SIGNIFICANCE

Click through to site to try

We were exploring Jerry Dallal’s site and came across this cute gizmo linked to as “a valuable lesson”.

Clicking the button simulates running 20 significance tests, each of which has a 5% chance of coming up significant when no effect is present. Underneath Jerry writes, “The chance that nothing is significant is only 0.3585, so don’t give up hope!”

This is probably a good thing to show to new grad students, whom I suspect get a bit too excited over significant results. We wonder how long it takes new scientists to realize that all that glitters is not meaningful.  Simply administering, say, the same 30 question survey to 4 different randomly-assigned groups should be enough to teach this lesson, so one would think that researchers ought to learn this quickly.

We notice that seasoned researchers, who are generally comfortable dismissing insignificant significance, fall into two camps. The first camp waves the results away as noise. The second camp believes that there was an underlying effect, but dismisses it as stemming from an ignore-worthy flaw in the design: “we’re seeing this because that group answered first thing in the morning”, “…right after lunch”, “…on Monday”, “…at the end of class”, etc. They never seem to say “we’re seeing this because a bunch of people who  answer that way got randomly put in that group”.

THE CENTER OF THE CIRCLE HEURISTIC

When a number of crimes, for instance burglaries, can be linked to the same offender, police often plot the locations on a map. The art of finding the location of the criminal’s home based on the crime sites is a key objective in what is known as geographical profiling.

Snook, Zito, Bennell and Taylor (2005) ran a competition between 11 techniques for locating offender residences. All techniques took as input the x-y coordinates of crimes (all committed by the same crook) on a map and made predictions of the criminal’s home location. There are many ways to turn these sets of coordinates into a point prediction. One stands out as exceptionally simple, so much so that it can be carried out with a pencil and ruler:

Center-of-the-circle heuristic: Predict that the offender lives at the mid-point of the line connecting the two farthest apart crime locations.

The figure up top shows the heuristic applied to a set of crime locations. Note that the midpoint is the center of the smallest circle encompassing all the crimes. Ten other methods for profiling were tested, including other “spatial distribution strategies” such as finding the centroid, harmonic mean, geometric mean, or point of minimum distance. Also investigated were computationally intensive “probability distance strategies” that involve fitting probability distributions such as the negative exponential, normal and lognormal and zooming in on maximally likely cells.

The 11 methods were applied to the crime locations of 16 UK residential burglars who had committed at least 10 crimes. Interestingly, all the strategies were run on a computer except for the center-of-the-circle heuristic, which was applied manually. We don’t know if that is because the crime-fighting software used is not clued into the power of the simple heuristic.

As shown below, in the aggregate analysis, the center-of-the-circle heuristic made the most accurate forecasts of where the criminals lived. In another study (Snook, Taylor and Bennell, 2004), laypeople who were trained with heuristic methods were as accurate at predicting home locations as a computerized geographic profiling system.

Decision Science News’ point is not that highfalutin statistics are bad. In fact, the more complex systems provide information that the “x-marks-the-spot” systems cannot provide, such as probable search areas. The blog merely wishes to observe that it is sometimes difficult to impossible to beat simple strategies at forecasting.

References:
Snook, Brent, Michele Zito, Craig, Bennell, Paul J. Taylor. (2005). On the complexity and accuracy of geographic profiling strategies. Journal of Quantitative Criminology, 21(1), 1-26.

Snook, Brent, Paul J. Taylor & Craig Bennell (2004). Geographic profiling: The fast, frugal, and accurate way. Applied Cognitive Psychology, 18, 105-121.

Notes:
This is extracted from a paper the DSN editor is writing.
Decision Science News does not promote vigilanteism.
Yes, that is a rotated map of Rhode Island.

POSITIVE / NEGATIVE / ZERO SUM ADVICE AND THE JOB MARKET

Every year, Decision Science News republishes this wonderful, caring advice on the AMA (Marketing) job market. Recently, our friend and kindred blogger Andrew Gelman asked how much of the advice is positive sum.  Your humble editor, not knowing much about what positive sum advice is, looked into it.

A cocktail-party-worthy example of things positive / negative / zero sum comes from the entertaining Why Not?: How to Use Everyday Ingenuity to Solve Problems Big And Small by Barry Nalebuff and Ian Ayres, who argue that The Club (a steering wheel lock) is zero sum while LoJack (a chip in your car that reveals its location) is positive sum. Their logic is that The Club simply causes car thieves to chose another car to steal instead of yours. So if you get +1 points for not having your car stolen, some other schlub gets -1 points for losing a car and -1 +1 = 0, the sum of the game. They then argue that LoJack leads the police to the location of the car theft rings, which they can subsequently bust. With the car thieves in jail, car theft overall goes down and the net benefit to society is argued to be positive.

Why Not? is a book we wholeheartedly enjoy and endorse, despite our upcoming diss of one of its points. The authors state that the net value of The Club to society is zero. Not probably zero, but zero. But dear authors, are you sure that people always steal another car upon seeing The Club? Is there a a fixed quantity of crime that each criminal must commit? A law of conservation of crime? In the Club / LoJack example, every criminal either ends up shifting crime to where its easier, or in jail. This blog’s inutition is that some crime just disappears when there are fewer temptations in the environment, much like some overeating disappears when there are fewer oversized portions in the environment. This blog has moved between the US and Europe many times and experienced first-hand the effect of smaller portions on body weight.

(Note also that the conservation of crime idea would suggest that in a fully LoJacked future, the new generation of criminals (or those just spring from the joint) would steal something else. How’s that positive sum?)

Back to the academic job market. Andrew’s comments are indented. The quotes within the indents are Andrew quoting Decision Science News.

If I tell people a secret way to put their proposals at the top of the pile for a granting agency, that’s zero-sum

We will buy that. The way that a stack works, something going to the top position moves the things formerly above it down one position such that the sum is zero. This point will come in later.

Just for analogy, if I give people advice about how to make cleaner powerpoint presentations, that’s positive-sum (better communication for all)

Or is it? Let’s say we were an obnoxious blog. We could then argue that the person who receives said PowerPoint advice benefits and the person who doesn’t does not. To this you might say, yes, but Decision Science News, presenters are not in a fixed sum competition in which one person’s gain is always another person’s loss. But is this true? If I’m in your field and my PowerPoints rock, I may hurt your chances of getting the scarce endowed chair, winning the teaching award, landing the best book contract on our obscure topic, etc.

“Get yourself a room in the conference hotel, preferably on the floor where the express elevator meets the local elevator for the upper floors.”: Zero-sum. If you get a room at the conference hotel, somebody else will have to find a room elsewhere.

Job candidates generally benefit from being in the conference hotel and on the elevator-transfer floor. Regular conference attendees do not benefit from being on the noisy, congested transfer floor, so both win if the job-seekers are on that floor. About conference hotel capacity, every single seeker could stay in the conference hotel easily (and probably most could stay on the transfer floor). Some non-seekers might like to stay at the conference hotel (though many don’t like to), but seekers are most benefited by staying in the conference hotel, and most harmed by not staying in it because of the running around they must do. Sounds positive sum to us.

“One of the biggest risks facing you is that you will be forgotten. Make sure the interviewers know something unusual about you.”: Zero-sum. Or maybe positive-sum, I don’t know.

That would be positive sum. If you think, “but Decision Science News, one person being remembered means that another person will be forgotten”, we delight in declaring that memory is not like a stack of paper.  As one candidate’s probability of being remembered increases, the other candidates’ probabilities do not decrease in proportion. Memory, like much of the world (we are starting to realize as we write this), is not zero sum. And if more applicants are remembered, the better the chances that the market will arrive at good matches between schools and candidates, and that’s a good thing.

“It’s a good idea to hit a school with 2 packets, 3 if you suspect they’re a little disorganized.”: Negative-sum. I’m not saying this wouldn’t work–a couple of years ago, our department missed out on a top candidate because we literally lost his file. But it can’t be good to have duplicate letters flying around.

Or can it? When the job market fails, it can also be quite bad: couples can get stuck in different cities, people can take jobs then relocate a year later, etc. These events have environmental impact, too.

“Get your advisor / sponsor to write a cover letter encouraging people to meet with you at AMA.”: Zero-sum, I think.

This isn’t even really advice, more of a how-to. Candidates must have introductory letters if they want to be in the game at all. It’s like saying “if you want to get from Columbia to NYU, head downtown”.

“Don’t gossip.”: Negative-sum. I say this because Dan illustrates with a story where the gossiper provided him with useful information! So the gossip was probably helpful.

Kidding aside, I would guess that there are laws concerning rumors in financial markets because they find them to be bad for the larger system.

Given that in most real problems, one can futz with the reference classes, populations, probabilities and payoffs to arrive at whatever total one wishes, this blog wonders if zero sum analysis actually adds up.

THE MPG ILLUSION

In their carpooling conversations, friends of Decision Science News Rick Larrick and Jack Soll have a come up with poignant example of how information that is mathematically equivalent is psychologically different (to paraphrase Feynman). Their “miles per gallon illusion” has been passed by our editorial staff to policy makers and may someday, hopefully, change policy. It is definitely worth reading. Check out:

• The article in Science. (Psst: If you access the article from the link on this page, you don’t need a pricey subscription to Science to read the article.)
• The blog
• An interactive quiz
• And, of course, the summer blockbuster movie.

useR! CONFERENCE, AUGUST 12-14 2008, DORTMUND GERMANY

Impressive statistical computing types like Andrew Gelman, Gary King, and others will be presenting at this year’s useR! conference. Decision Science News might just have to hop over and check it out. The program looks great. Those interested in learning R might be interested in our Decision Science News R tutorials one and two.

About the Conference

useR! 2008, the R user conference, takes place at the Fakultät Statistik, Technische Universität Dortmund, Germany from 2008-08-12 to 2008-08-14. Pre-conference tutorials will take place on August 11.
The conference is organized by the Fakultät Statistik, Technische Universität Dortmund and the Austrian Association for Statistical Computing (AASC). It is funded by the R Foundation for Statistical Computing.

Following the successful useR! 2004, useR! 2006, and useR! 2007 conferences, the conference is focused on

1. R as the “lingua franca” of data analysis and statistical computing,
2. providing a platform for R users to discuss and exchange ideas how R can be used to do statistical computations, data analysis, visualization and exciting applications in various fields,
3. giving an overview of the new features of the rapidly evolving R project.

As for the predecessor conference, the program consists of two parts:

1. invited lectures discussing new R developments and exciting applications of R,
2. user-contributed presentations reflecting the wide range of fields in which R is used to analyze data.

A major goal of the useR! conference is to bring users from various fields together and provide a platform for discussion and exchange of ideas: both in the formal framework of presentations as well as in the informal part of the conference in Dortmund’s famous beer pubs and restaurants.

Prior to the conference, on 2008-08-11, there are tutorials offered at the conference site. Each tutorial has a length of 3 hours and takes place either in the morning or afternoon.