The National University of Singapore nudges

Marcus Tay Guan Hock, Sustainability Executive at the National University of Singapore, writes in to say that Nudge “gives me hope as an environmentalist,” and explains how the school used principles of choice architecture to redesign its recycling program.

Here at the National University of Singapore (NUS), we designed our recycling bins to tackle the issue of contamination, applying what you called “Expect Error” from users.

When users throw the wrong things in the recycling bins, it wastes the efforts of those who recycled properly. For example, paper bins are often contaminated with food waste, rendering all of the paper unrecyclable.

This situation is rather serious in Singapore. A Straits Times Article on June 15, “What rubbish,” indicates non-recyclable waste found in all 80 recycling bins surveyed.

At NUS, we did the following two things. They have worked wonders.

At the point of disposal, we help people decide if the item can be recycled using proper and clear labels. These labels are designed so that before users can throw trash into the bin, they will see the labels which instruct them what can and cannot be thrown.

We give people an option not to throw garbage into the recycling bin if the garbage cannot be recycled by pairing every set of recycling bins with a trash bin as well. Because some people are not yet environmentally conscious, they just want to get rid of the rubbish in their hands, whether it can be recycled or not.

What would it take to get you to take the stairs more? How about music and a view?

The stairs between the upper floors of the University of Chicago’s Booth School of Business are made to be used. They are placed smack in the middle of blocks of faculty offices. They zigzag back and forth up open columns, allowing people to see easily between floors. They have plenty of light, are carpeted, and blend into the design of the building.  By placing the stairs in a common area (the elevators are a longer walk away) and making them inviting, the architect created a nudge to encourage a smudge of healthier behavior in the work place.

Creating more accessible staircases through public policy and physical architecture is one way to promote an active lifestyle, say Dr. Ishak Mansi of Louisiana State University, and his wife, Nardine Mansi, an architect, in the Southern Medical Journal. A small 2.8 percent increase in stair use would cut 300 grams of weight from a typical person, they say.

So how does one design a building where people actually use the stairs? There are three key features.

1) Fewer turns between the stairs and the closest entrance.
2) Stairs with large surface areas (not too narrow and steep).
3) Create a view, either up, down, or across, from the stairwell. No one wants to walk up a tiny, white box.

The Booth School of Business staircases meet all of these requirements (perhaps it’s no surprise the building won a major design award last year). For those who can’t build new stairwells, there are a few other nudges to try. Displaying motivational signs in the lobby and throughout the building, and playing music in the stairwell can increase stair use. Together, these two nudges can increase usage by as much as 9 percent. Hanging artwork on the stairwell walls, closing elevators occasionally, and offering incentives like fruit are also known to work.

The journal article is gated, but a short news summary is here.

The dentist bib as choice architecture

Talya Miron-Shatz, a psychology post-doctorate at Princeton with Daniel Kahneman, recently received a crown and a root canal in the same sitting from two different dentists. Since the two procedures were separate from one another, neither dentist seemed to know the local anesthetic that the other was providing to Talya. With poor communication, Talya almost received a double dose of Novocaine (or something similar). There has to be a better way to prevent these kind of errors, she thought.

And here’s my two cents for human engineering. Dental patients wear a bib around their neck. How about if the (dental) office purchased a Sharpie, and had each doctor write down how many injections he/she gave the patient, and their exact location. Better still, how about if this was pre-marked on the bib?

Read the full post as Psychology Today.

The choice architecture of Time magazine’s online poll of influentials

Time Magazine’s list of 203 finalists for the 100 most influential people in the world is out for 2009 and thanks to Nudge, Richard Thaler and Cass Sunstein are on it. Almost making the final list is a great honor for them, but in Sunstein’s household, he’s just playing catch-up (his wife Samantha Power made the final list in 2004).

On Time’s web site, readers are able to help editors pick the influentials. Already, it is clear that the online rankings are quirky, and that’s not just because readers have ranked Thaler and Sunstein dead last as of Sunday March 29. A number of readers on Freakonomics have pointed out how awkward some of the choice architecture of the site is. Say you want to vote on who belongs on the list. Time provides a slider on a 1-100 scale that visitors can use to score influence. But you might be easily confused. Should influential people get higher or lower numbers. Seems like you’d want to give higher number to influential people; but the winner of any best/worst list always comes in at No. 1.

Reader Tom notes that Federal Reserve Chair Ben Bernanke has an average score of 5. “People are interpreting the top 100 scale to mean 1 is the highest and 100 is the lowest,” he writes. E agrees that the slider is poorly conceived. “I started by sliding toward 1 to say the person has less influence (like 1%) and toward 100 for more influence…Now I think it’s ranking that the slider indicates, not ‘influence’ as the text indicates…. So I just gave up, not understanding the paradigm.”

Initially, Time gave no guidance about how to interpret the scale. Since the site went online, Time has added a description to the scale that makes clear 100 is “most influential.”

Meanwhile, a lot of people are voting for moot, the founder of the image bulletin board 4chan.org, and the leader with a score of 75. The Korean pop star Rain (he’s Korea’s answer to Justin Timberlake) finished second on the online poll, and is currently ranked No. 3. Online polls are one good way to open up the selection process to people, but they can be vulnerable to manipulation, especially when mobilized groups participate in small, relatively unknown ones. Basically, they are a highly flawed form of choice architecture.

Comedian Stephen Colbert recently outwitted NASA by asking his viewers to write him in as part of an online poll to name a room on the international space station. One U.S. congressional member is backing Colbert’s win, telling the Chicago Tribune, “Funding for space exploration is something where getting the public’s interest is challenging, and having Colbert would bring interest to NASA’s program. Over a quarter of a million people or so came online to chime in on the naming question…It just shows what happens when you reach outside the normal circles.”

Colbert is doing pretty well again in Time’s poll, holding steady at No. 4.

Slang for “choice architect”

Courtesy of Philip Frankenfeld. Add your own…

1. Nudgeletarians
2. Noodges
3. Noodgeniks
4. Vectordictorians
5. Chess Theorists
6. Suasarians (Moral suasion)
7. Carrot and Stickleteers
8. Homo Vectoris
9. Boxitects
10. Funnelists (akin to finalists)
11. Carrotodomists
12. LibPats (libertarian paternalists)
13. SofPats
14. ComPats
15. Gridologists
16. Matrix d’s

What choice architects can learn from a new medical checklist

A new study on another medical checklist is out, this one conducted on noncardiac surgery in eight hospitals in eight cities around the world to ensure a diverse range of clients. As with the now-famous previous checklist – call it Checklist I – this one produced decreases in death and infection rates. What is most interesting about this checklist – call it Checklist II – is that is undercuts one of the supposed insights of Checklist I, namely that shorter was better.

Checklist II is long. Very long. Totaling 19 items. (Recall that Checklist I uses only six.) Does this mean checklist designers need not worry about keeping the number of items to remember to a minimum? Though it would tempting to draw this conclusion, it would be as foolhardy as pronouncing the absolute virtues of brevity based solely on the findings from Checklist I. What is more likely is that checklist length matters less than the overall environment in which each step is taken; the mechanisms in place for enforcing each step are what matter most. This enforcement is all the more necessary since human memory is not reliable enough to count on recalling every step. Indeed, the authors of the Checklist II study point out that, in spite of the medical improvements, “omission of individual steps was still frequent.”

Designers of Checklist II broke the 19 steps into three subcategories, administering the steps at different points in the surgical process. The number of steps and points of administration make the identification of key steps difficult. But the authors explain the overall choice architecture this way:

Use of the checklist involved both changes in systems and changes in the behavior of individual surgical teams. To implement the checklist, all sites had to introduce a formal pause in care during surgery for preoperative team introductions and briefings and postoperative debriefings, team practices that have previously been shown to be associated with improved safety processes and attitudes and with a rate of complications and death reduced by as much as 80%. The philosophy of ensuring the correct identity of the patient and site through preoperative site marking, oral confirmation in the operating room, and other measures proved to be new to most of the study hospitals.

After devising the steps, determining the appropriate points in the surgical process to implement them was the key challenge. For example, the designers encouraged administering antibiotics in operating rooms instead of preoperative wards, “where delays are frequent.”

The paper’s authors are unsettled about the potential of a Hawthorne effect – in which an improvement in performance is the result of doctors’ knowledge of being observed – in the overall results. At the risk of carelessness, for now, this concern is perhaps best left to the academics. Building in an observant whose presence leads to changes in behavior is a perfectly acceptable option for someone who is not so worried about the exact contribution of each casual mechanism. As social scientists know, even trying to isolate the effect of observation can be futile, for in some cases observation is one way to enforce particular social norms, making the combination of the two what influences behavioral change. So while the Hawthorne effect has confounded social scientists for years, everyday choice architects may need not worry so much. After all, the Hawthorne effect itself is just another nudge.

A reader spots a convenience store nudge

Paul Sweeney writes in about a retail store sales manager:

A friend of mine stopped selling lighters and only stocked matches. When asked why he said, “It stops the riff raff from coming in,” which of course means less need to stock low, low cost beers in the front of shop, opening up new stock, and increasing the safety in the shop.

People are too smart for choice architecture sometimes II

Via Dan Lockton’s Architectures of Control

The strange power of the free test score report

Amanda Pallais, a Ph.D. Candidate in Economics at the Massachusetts Institute of Technology, is currently working on research about how small changes in how test score reports are sent to colleges can affect student behavior. Pallais kindly agreed to discuss her findings thus far in a guest post.

By Amanda Pallais

Picking the right colleges to apply to is a daunting challenge. Students must choose one of over 2^2,400 combinations of colleges (and that just includes four-year schools in the U.S.), while facing great uncertainty about the costs and benefits of attending each college. Even deciding how many schools to apply to is not easy. Students looking for a rule of thumb can find one in an unexpected place—the number of colleges to which they can send their standardized test scores for free.

I’ve been looking at one college entrance exam in particular, the ACT. The ACT is especially popular in the U.S. Midwest and more students actually take the ACT than the SAT each year. In the fall of 1997, the ACT Corporation increased the number of free score reports it provided from three to four; additional score reports still cost $6. While the difference in the cost of sending four score reports dropped by only $6, the student response was enormous.

As you can see from the two charts below, I find that before the change, over 70% of students sent their ACT scores to exactly three colleges, while less than 5% sent their scores to four. Afterwards, less than 10% of students sent three score reports and approximately 60% sent four. As a result, 23% of students sent an additional application. Over the same period, there was no similar change in the number of score reports sent by students taking the SAT.

Number of ACT Score Reports Sent

Number of SAT Score Reports Sent

What caused the large response to the fourth free score report? It was not the $6. If it were, students should be equally responsive to a $6 decrease in application fees. But I find there is no relationship between changes in a college’s application fees and in the number of applications it receives. Instead, students are likely reacting to an implicit rule of thumb about how many colleges to apply to. They may interpret the number of free score reports the ACT provides as an indication of the number of score reports it recommends sending. This is consistent with the results of other economists who have looked at choices of 401(k) and prescription drug plans.

The change in student behavior is not just an interesting artifact. It has real world consequences that can permanently affect these students. Sending an additional score report could have large benefits for low-income students in particular. I very conservatively estimate that by increasing the probability that a low-income student attends college and attends a selective college, sending an additional score report could increase her lifetime earnings by over $6,000. She might also be more likely to end up at a school that’s a better fit for her personally. All of this, because of a simple $6 change.

The paper that this post is based on is titled, “Why Not Apply? The Effect of Application Costs on College Applications for Low-Income Students.” The full version can be found here.