- New research finds that when people share advice, they get a shot of self-confidence that helps them on their own journey of self-improvement.
- Wharton's Lauren Eskreis-Winkler talks through the research.
Knowledge@Wharton: Can you talk about what past research says about some of the common things we do to help people achieve their goals, and why they don’t always work the way we think?
Lauren Eskreis-Winkler: I think that the status quo is to design interventions as scientists, to communicate our knowledge and teach people in the world of business, kids in schools — people who are struggling with a host of self-regulatory goals — about how they can more effectively achieve them.
Have you read?
I started off that way, designing interventions and running focus groups. In the process of interacting with the people I was trying to help, I realized that they had a ton of knowledge. There seemed to be a hole in this baseline assumption that people are not achieving to the level that they could because they lack knowledge. It seemed like that wasn’t really the full story. For example, I was running a study with a bunch of students who were trying to do better on the LSAT. In addition to communicating all of these strategies and ideas that we had come up with in our research lab, I started asking people, “What do you do to motivate yourself? What are the most effective strategies and techniques that work for you?” Over and over again, I was just amazed by how sophisticated people are. These are not people who are earning Ph.D.s in motivational psychology, the way I was. Yet they really seemed to be the experts in their own motivational topography and knew what they could do to push their own buttons.
One student stands out because I thought he was just brilliant. He told me that when he wanted to study for the LSAT, he had already decided on Vanderbilt University. He really wanted to be a student there. He lived 45 minutes away, so he could have studied in his house. But he said he realized that if he drove the 45 minutes every day to the cafe across the street from the Vanderbilt Law School, his studying improved. He was focused. He would look out the window and see the law school and refocus. I never would have suggested that to him, yet people seem to have these customized strategies that work really well. It led to the insight that maybe when people are struggling with goal achievement, what they lack is not really knowledge, it’s the motivation to put their knowledge into action. That led to this counter-intuitive idea that if people are struggling with goals, maybe they would benefit as much, or even more, from giving advice to other people than from receiving it.
Knowledge@Wharton: In the previous literature, is there evidence that advice-giving benefits the giver in general?
Eskreis-Winkler: I think there’s a lot of work on the benefits to the recipient. There are some lay programs, like Alcoholics Anonymous, which have baked into their program the notion that in addition to receiving help, you’re also a sponsor of a rookie in the program. But I’m not really aware of research that systematically looks at the benefits of giving advice and giving information, particularly when it’s something with which you struggle. I think it’s a pretty counter-intuitive idea that if you’re a smoker who can’t quit smoking, I would go up to you and say, “Hey, could you please advise somebody else?”
Knowledge@Wharton: From what you’ve said and the people you’ve talked to, it seems we may not be experts in nudging or how to design an incentive, but we may be experts about ourselves.
Eskreis-Winkler: Exactly. I think people do know in many ways what works best for them. It remains an open question as to why advice-giving works, but one of our primary hypotheses — and we’ve gathered some evidence for this — is that it really builds your confidence.
If you imagine you are constantly trying to lose weight and repeatedly failing to achieve that goal, it saps you of your confidence. The notion that suddenly you’re put in a position where someone asks if you have useful information and presumes that you do, that could raise your confidence. The act of giving advice forces you to focus on the things that you already know how to do versus things you don’t, the things that are in your control versus the things that aren’t. For all of these reasons, we think that advice-giving probably is a confidence booster that raises people’s motivation, giving them what they need to achieve their goals and to realize what’s holding them back.
“It remains an open question as to why advice-giving works, but one of our primary hypotheses … is that it really builds your confidence.”
Knowledge@Wharton: Maybe it’s something that’s often been a source of shame or embarrassment, whether it’s losing weight or stopping smoking. Then you’re taking that problem that is a source of bad feelings and making it into a source of good feelings because you’re being looked to as an expert. Is that correct?
Eskreis-Winkler: Exactly. We also studied this in kids in school, even the kids who aren’t necessarily feeling bad about themselves. I think it’s a totally novel experience in today’s educational world that you’d go up to a 15-year-old in school and say, “Hey, we think you have really useful knowledge. Could you share it with someone else?” The whole structure of education is that these kids are the recipients. They’re sitting in class taking notes, receiving knowledge. In general, it was just a very novel experience for a kid, regardless of the degree to which they’re failing or not failing, to be suddenly appointed as an adviser — to just make them feel that, “Wow, I have something to give to somebody else.”
Knowledge@Wharton: How did you design this experiment with high school students, and what were the results?
Eskreis-Winkler: This was a large-scale, randomized controlled trial that was run with close to 2,000 high school students. The scale of this was only made possible by partnering with the Behavior Change for Good Initiative and the Character Lab, both at the University of Pennsylvania. They really made this research frictionless and facilitated these partnerships. Typically, I think it’s very hard for researchers to test questions like this at scale.
We recruited about 2,000 students and randomized them to one of two conditions. Either they’re in the treatment, which is that they give advice to younger students, or they’re in the control condition, which is practice as usual and they didn’t receive anything in particular. The program was online, so teachers took their students to the computer lab and students signed in. There was this very aesthetically pleasing, graphically designed program that students walked themselves through. They were asked to be coaches. We said, “Help us help other students.” Then the treatment — they went through a series of exercises that tried to elicit their advice.
There were some multiple-choice questions that asked them to advise on optimal study locations. There were some open-ended questions where they were writing notes of advice to younger students. The whole experience was short, but it was meant to make them feel like bona fide advisers. They had information to give, and we were getting it, and we were actually going to give it to younger students.
The hypothesis was that this act of stepping into the adviser role would raise the students’ confidence, increase their motivation. It was a pretty high bar, but we were hoping and expecting that it would, in turn, raise the students’ achievement levels. We collected the students’ grades over the academic quarter to see whether this intervention, which was delivered to students at the beginning of the third quarter, would increase their grades. And it did. We specifically pre-registered and predicted in advance that it would raise their grades in a target class. This is a class in which students self-report that they’re most motivated to improve. Math is a subject that’s notoriously difficult to change student achievement and a subject in which many students lack confidence. We thought this advice-giving intervention would be effective in math, and we did find that the students’ target grades and their math grades improved relative to students in the control condition.
Knowledge@Wharton: How much did their grades improve?
Eskreis-Winkler: It was a couple of points. The grading scale was between 50 and 100, so students who are at 50 are failing out of their classes, and 100 is the best you can do. On average, students improved one or two grade points, so it’s not like a huge effect. But I think it’s really noteworthy given the cost of the intervention. It’s a marginal cost of zero. Any school district could implement this. Unlike many very intensive, costly programs like tutoring that often have very small, if any, effects, this program is costless not only in money but also in time.
“The act of giving advice forces you to focus on the things that you already know how to do versus things you don’t.”
Students had a very brief interaction with the program. I think one interesting direction moving forward would be to look at whether booster sessions help. If you’re an advice-giver not just once, but many times, or if you interact not just with a computer module, but with an actual younger student, could all these things increase the efficacy of the program and lead to much stronger effects?
Knowledge@Wharton: It’s kind of like a mentoring program in that the older students shared what they’ve learned with the younger students.
Eskreis-Winkler: Yes, and all of the advice really focused them on motivation: “What do you do to stop procrastinating? What would you tell a student who’s really not so motivated in school?” … It was really focusing students on what do you do when you get home and don’t want to do your homework? How do you get yourself to do it? Students have these ingenious strategies that, I think, sometimes they lack the motivation to put into action, but they have really great ideas about what they should be doing.
Knowledge@Wharton: Was there one from the student study that really stood out to you?
Eskreis-Winkler: I’ve come across a host of different things students say. They’re very effective at rewarding themselves. The student who has two pages of math problems and really doesn’t want to do them tells me that he will put a candy at the end of the page. Each time he goes through the problems, he eats a candy. I think they’re very in tune to rewards and incentives. There are also very creative. One kid who I often quote, a seventh grader, told me very dramatically that he imagines his house is burning down, and if he doesn’t finish his homework in time, the fire is going to consume him. Just really imaginative.
I think psychologists have known for a long time that very young children engage in imaginative play, and they’re incredibly creative. Talking to students about how they motivate themselves just reinforced that that never goes away. I think maybe we channel that imaginative and creative play in specific ways, and this is definitely one of those ways.
Knowledge@Wharton: What are the implications here, not just for high school students or educators but also in different settings? If I’m a business manager, could I use this? If I’m an individual employee and my company is not going to do this, could I try it?
Eskreis-Winkler: Yes, absolutely. We expanded because we were interested in that exact question. We wanted to know not just with regard to academic achievement, but a host of self-regulatory goals — like people struggling to lose weight, to control their tempers, to save money, to motivate themselves in the job market. We explored all four of those domains.
We recruited people who self-identified that they really struggle to save money, that they really struggle to lose weight, they struggle to control their tempers, and unemployed individuals who are struggling to get a job. In all of these domains, we tested whether people were more motivated by giving advice or receiving advice from an expert source.
For example, the people who were trying to lose weight either gave advice or received advice from nutritionists at the Mayo Clinic…. [People] overwhelmingly said that they were more motivated by giving advice than by receiving advice.
“To the degree to which you’re lacking confidence, it seems like just being repositioned into the role of a giver versus a receiver can give you everything you need.”
What was especially interesting is that afterwards, we recruited people who didn’t go through the two activities but who were predictors. They were yoked to somebody who had given advice and received advice. They had to predict which would be more motivating. Overwhelmingly, the effect flips. People say, “Well, of course, somebody who can’t save money is going to be more motivated by getting advice from someone at America Saves than by giving advice.” So, people completely mis-predicted the phenomenon. We find that it’s really not confined to kids in school. Generally, when you’re struggling with motivation, when the problem is not knowledge, you really do seem to benefit more from giving advice than from receiving it.
Knowledge@Wharton: Instead of looking outward, you need to look inward.
Eskreis-Winkler: I think that’s absolutely right. People think that they’re not achieving because they’re lacking something, and often that something is information. It’s like, “I don’t have it within me. I have to go to a teacher or an expert or somebody else who can give me what I lack.” To the degree to which you’re lacking confidence, it seems like just being repositioned into the role of a giver versus a receiver can give you everything you need.
Knowledge@Wharton: What’s next for this research?
Eskreis-Winkler: I’m really excited by the real-world implications. This was the first large field study, so I’d be really excited to repeat it in other domains. I think that the light-touch nature of the intervention does leave this as sort of a proof of concept. I’m convinced by the concept but excited to see the degree to which making the intervention more heavy-handed, more involving could potentially have much larger effects.
We think that advice-giving raises your confidence because it makes you feel like an adviser. I could think of about a thousand different creative ways to not just have people sign into a computer program, but really pair them up with somebody, make it a more intensive, long-term advising program, give the adviser feedback on how effective their advice has been and how much they’re helping somebody else. I think all of those things could be done to make it more effective.
I think the other direction is less about ecological validity and less about exploring how this can be effective in the real world, and more about understanding the mechanisms behind it, which is more a researcher question. To the extent that you really understand the theory and what’s going on, in some sense, that’s the most practical thing in the world because then you can design an intervention and see, “Well, if it really works by building confidence, then how can we focus on redesigning the intervention to build confidence?” Alternatively, if it’s really about when people give advice, then how can we redesign the intervention to strengthen that piece of it?
I’m interested in these twin questions about how can we roll it out in the real world in ways that are effective, but at the same time understanding this seemingly very simple act of giving advice, which is actually probably very multi-determinant in terms of why it’s having an effect and how it impacts people.