This book was filled with a variety of research already done and what Behavioral Science in the Wild does helps apply all of those long pages of research, into brief summaries on applying what has been learned and studied in an easier way to implement and see transform the business you are in. I would recommend it for more technical personnel and as a reference book.

In LinkedIn:

Need to up your reading? There’s a wealth of new books out in May covering #business #science #health #mindandbody #history and #essays.

📚Behavioural Science in The Wild by Nina Mažar (Mazar) & Dilip Soman addresses the "why" and "how" behind BI’s origins, and how best to translate and scale behavioral science from lab-based research findings.


(I also did a podcast with Nina and Dilip for 42Courses)

It's more technical and oriented to research on behavioral change than what I expected. There are references to studies that worked and also to others that didn't provide the predicted results. Therefore, we can get a broad perspective, but at the same time it requires a surgical patience to extract solid conclusions.

Thank you to the publisher and @NetGalley for the E-ARC copy of this book. The rating of this book is entirely of my own opinion. #NetGalley

This was a really good primer on behavioral science and the ways that it can be applied! As someone who studies psychology, I found a lot of it interesting but a lot of it I also already knew so I was trying to look at it from the perspective of someone with less knowledge, and I think it's a little technical for general audiences but there is still plenty of useful information. It doesn't come out for awhile yet so I will do a more in-depth review on other sites closer to release date.

Wow! This book was very in depth and really had me retracing my steps to see if I fall into any behavioral science normalities. This book felt well researched and well written, definitely something i would recommend to anyone interested in this particular area of study.

Really interesting overview of difficulties translating behavioral science/nudges into working policy. Consistently, policies have much smaller effects than they do in the lab—a “voltage drop.” That doesn’t mean we should stop—the average effect of a government nudge is significant and positive, and even a 8% change in behavior can mean a lot of people helped, but it is worth thinking about how we go from 30% in the lab to 8% in practice. A lot of this seems to be explained by publication bias: studies that don’t show a (big) result don’t get published.

Sometimes the nudge was the wrong lever. For example, whether organ donation is the default or opt-in varies across countries; when Wales flipped the default, the organ supply didn’t seem to increase, perhaps because a default opt-in can be overridden by family members. It’s possible that, pre-intervention, a country’s default simply matched its population’s overall willingness to donate organs.

Sometimes the studied nudges systematically vary from implemented ones. Academic nudges are more likely to be delivered in person and to be about asking people to choose between 2 options, both of which are more effective than other types. The government may face barriers changing a service from opt in to opt out despite the effectiveness of that kind of nudge. Still, nudges are generally quite cheap for government, with often no marginal cost to changing how they communicate (though if we wanted them more effective, we’d need more people to do things in person).

But backfires are possible. A South Korean test of messages to deter excess credit card spending worked for 12% who were the heaviest spenders but backfired for 88%; maybe better targeting would have helped. A Mexican pension experiment showing people how much they’d have for retirement with the goal of encouraging savings showed a similar backfire effect (though that may have shown that they were targeting the wrong problem for that group, which was invested in a low performing pension fund).

The contributors constantly emphasize that context is everything. “Every detail of how the intervention is implemented matters. The fundamental behavioral insight, say peer comparison, may transfer as is, but how it is packaged and delivered often does not.” So, for example, different groups may need different tones, different timing/frequencies of messages, and different graphics. It was fascinating to learn that, at ASU, “a four-year school with a traditional student body,” an intervention to help students get more financial aid needed to email both students and their parents. “But at CUNY, students tend to be older, commonly with children of their own, and are often the first in their families” to go to college; they didn’t need emails to their parents. ASU emails written in a friendly, casual tone were “approachable and unintimidating,” but they seemed “unprofessional and untrustworthy” to CUNY students, “who see college as providing a service more than an experience.” For recorded messages, having the research assistants record them leads to hangups, but voice actors and clear scripts get listened to more.

There’s an interesting chapter on debiasing interventions and how we shouldn’t try to change people’s minds, but rather “redesign their systems and environments, so that biases have no place to hide.” You may have heard of blind auditions and their effects on gender bias, but there’s also guidance on job listings, replacing words like “entrepreneurial” and “strong” with neutral synonyms like “creative” and “dedicated”—the latter attracted more women, who represented 4% more of the resulting pool, but also more applicants in general, especially “men who were more weakly identified with their gender.” In evaluating job candidates, it helps to compare individual responses “horizontally,” by “looking at one question or criterion for all applicants and then moving on to the next question or criterion,” instead of trying to assess each candidate separately. The chapter didn’t talk about this, but I expect it may fight back against the known problem that position requirements get shaped to the strengths of the most attractive male applicant. Likewise, it may help to change numerical evaluations, as dumb as that sounds. It turns out that “[e]valuators are actually less likely to give 10/10 – an indicator of perfection and brilliant performance – to high-performing women,” but a 1-6 scale instead “closed the gender gap on perfect scores.” And within organizations, an opt-out scheme for consideration for promotion helps women by eliminating the gender gap in who puts themselves forward.

What about environmental acts? Can we avoid moral licensing/backlash? How do we get people to think of themselves as the kind of people who are environmentally responsible? Turns out that people who take “small, ‘token’ public actions such as a social media post supporting environmentalism” can behave less responsibly afterwards, whereas if you can get them to take “larger, effortful actions” or private actions like a home energy audit, that can have positive spillover effects on their other actions.

But again, everything has to be contextualized. One great chapter discussed localization in Kenya: The researchers started with “a well-cited intervention from psychology to induce mild stress …, in which those assigned to the treatment group were given a mock job interview in front of a panel of ‘experts’ while those assigned to the control group were simultaneously to give a speech about a friend.” It didn’t work in Kenya, because public speaking was not culturally understood as stressful and also, as one respondent asked, “Why are we being interviewed by butchers?” The experiment had followed the Western convention of “white coats = scientists.”

A slow read, but interesting, and especially relevant for the current time. I felt like I got a masterclass in new management styles from reading this book!