Last week we attended a workshop run by the consulting firm Ideas42 on behalf of the Robin Hood Foundation. The goal of the day-long event was to provide teams with an overview of behavioral economics, the study of the psychological, social, cognitive and emotional factors that go into decision-making.
Students lose their way in school for a broad range of reasons, many of which are non-academic, and the hope behind Robin Hood’s prize is that technological solutions can help them scale already proven strategies for improving matriculation rates.
The day began with an overview of the day to come by ideas42’s Josh Wright:
- Psychological “Scarcity” – how even unrelated stress and anxiety levy a cognitive tax.
- Hassle Factors – how seemingly insignificant logistical challenges can discourage and therefore effectively prevent task completion.
- Limited Attention – Information overload.
- Self-control – how to shore up self-control through social bonds, incentives and tricks you play on your own psyche.
- Prospective Memory – not only remembering to do something, but following through to actually do it.
- Social Norms – the human tendency to choose “normal” over “right”.
On the whole the sessions were lively, peppered with informal experiments, anecdotes and studies that illustrated key points through examples rather than jargon and formal definitions. Every session provoked incisive questions from the audience. For our part, we walked away with much more specific ideas for the design and implementation of our solutions as well as a host of questions for the folks at Robin Hood and CUNY, mostly around how the program will be introduced to the students.
Sendhil Mullainathan’s well-argued presentation on the Psychology of Scarcity made abundantly clear how poverty in one area of life (financial) creates poverty in another (academic performance). Study after study showed how even subtle reminders of financial stresses degraded cognitive performance. (You can get a synopsis from his New York Times piece on the same topic.)
- The obvious next question to ask then was: Can you prime in the opposite direction? Can you help people forget their stresses and perform better than they would otherwise? The answer? Nothing conclusive so far, although an interesting study found that activating Asian female students’ positive (ethnic) vs negative (female) stereotypes affected their quantitative performance which suggests it may be possible here too.
- Given that the nature of our relationship with the students will be long-term, another question we had was: Does the effect of priming wear off over time with repeated exposure, positive or negative?
- Another issue this brought up for us is whether the mere fact that students are participating in this program remind them of their “remediation” status thereby undermining our efforts to bolster their performance? As we understand to date, only remedial students will be using our technologies. Are we missing an opportunity to build technologies that help remedial students feel a part of the CUNY community as a whole?
Filling out Gigantic Forms
William Congdon talked about hassle factors. We could all relate to the hassles of coordinating calendars that span different aspects of life (work, school, childcare, family, commuting). Ideas42 in particular is working on improving the onerous process of applying for financial aid. Two approaches that came up repeatedly was the idea of 1) defaults to reduce the cognitive load of making decisions; and 2) pre-filled out forms to remove the hassle factor of having to “look up” information the university already has. So we’re wondering:
- Will participation be mandatory or will students be asked to decide? If the latter, will their participation be assumed with the option to opt out or will they be asked to opt-in?
- Will students be pre-registered, or will they need to go sign up? Can we piggy-back on their CUNY accounts?
- Will we have API access to student schedules and class syllabi, or will we need to ask the student to provide that information to us separately?
William also covered Limited Attention, a familiar topic in modern day life. One interesting tidbit from this session: It’s generally assumed that students’ preferred mode of communication is SMS. However, like Twitter, email before it and perhaps Yo! to come, what happens when every system and organization shifts to using text messages? More specifically, how will our communications with students interact with / collide with CUNY’s existing student support program START?
Hey, remind me to…
Matt Darling presented on Prospective Memory, the art of following through on future commitments. Memory, or the lack of it, is clearly the first problem to overcome. But assuming you are able to implement some kind of reminder system, how do you actually make those reminders count? Hassle factors and self-control (see below) come into play for sure. But Matt pointed to the power of “being specific” as one simple technique that doesn’t rely on the student to be more disciplined.
It made us reconsider how we’re thinking about designing our reminders. When you send them, how often you send them and the language you use in them of course remain important factors. But what exactly are you reminding the student to do, and how do you want them to respond to the reminder is where the real design problem lies.
Specifically for us, we’re working on ways to make tasks more concrete and bite-sized (aka, doable), tasks students can easily imagine completing successfully in a limited amount of time.
Allie Rosenbloom reminded us of the now famous marshmallow test. Self-control or willpower is a tricky issue in light of Sendhil’s ideas about scarcity. In an environment of scarcity, there simply isn’t a lot of self-control to go around. Social supports and personal incentives (e.g. betting against yourself) were 2 approaches discussed. The challenge we see ahead is how to create social supports through our technology when the students who will be using our service may or may not be in the same classes or even campus due to the structure of the Randomized Controlled Trial (RCT). We are encouraged though that the student population will be big enough that we can build community around shared interests and career aspirations, if not coursework. Allie’s talk also supported the idea that “getting specific” with tasks would be a boost to performance because as we all know, focusing on “exercising today” is a lot easier than thinking about the 20 days of exercise you committed to for the month.
What is everyone else doing?
Finally, social norms come into play – the emphasis of the studies noted here had to do with public service announcements intended to discourage problematic behaviors that end up reinforcing them. Examples included posters designed to discourage binge drinking that make the reader who doesn’t drink feel like they are abnormal, since everyone else must clearly be drinking, or provoke petrified tree theft rather than discourage it.
- Most relevant here is that commuting community college students (which is the majority of them) often feel isolated, and don’t have a good sense of how other students are handling the challenge of college. Social norms seem most relevant to us in terms of Ponder providing an atmosphere where students feel connected with their classmates, are aware they are working hard, and engage with one another through their college and career interests. We wondered if we could coordinate with existing CUNY support services to reinforce the somewhat disparate nature of the randomized participating students, and provide an in-person, face-to-face experience.
We’re thinking about how we can use the data we have about student progress to reshape students’ sense of “what’s normal” when it comes to school. Our goal would be to not only show students how others like them are succeeding in school, but to also paint a realistic picture of how much time and effort it takes to succeed at school. At the very least we can prevent students from feeling discouraged because it takes them ‘too long’ to study; or because they feel uniquely selfish in spending so much time on school in light of their other obligations.
At the beginning of the day David Crook from CUNY voiced his enthusiasm for the teams and the prize; having had an opportunity to digest all of the above, we can’t help thinking it would be great to have a second webinar to drill into the data with this new perspective.
All in all I think I’ve demonstrated here that the workshop provided much food for thought and advanced our thinking greatly towards our prototype for January. We also finally got an opportunity to meet and learn from the other teams in the Challenge. Thank you Robin Hood and ideas42 for organizing!