Everything a faculty member usually wants to know before bringing Resonance to a large class.
Resonance is a platform I built, out of about twenty-five years of research on the neuroscience of social connection, to help students in a shared community find a few people they are highly likely to become real friends with. Students take a short survey, get a report about themselves, and are matched with the two or three classmates they are most compatible with. The platform then helps them get past the awkward first conversation. Below are the questions instructors most often ask.
Four things, in order. They go to resonance.info, sign in with a university-specific address, and join the official group for your class. They take the Lens survey, which takes 10-15 minutes. They receive a personal report, which most students find surprisingly interesting on its own. And once enough classmates have finished, they get their matches: the two or three people in the class they are most compatible with, along with some information about why they are a good match. From there the platform gives them structured ways to start talking, described below.
The survey shows students statements about how they tend to notice and interpret situations — what comes into focus for them and what stays in the background — and asks whether that describes them more than other people, less, or about the same. We call these lenses. Two people who share hobbies but not lenses are just two strangers on a ski lift; two people who share lenses tend to click even when their interests differ. Nothing in the survey is sensitive.
A report about themselves. It describes their strongest lenses, how their profile compares to other students (in the aggregate, not by name), and the perceptual type that best fits them. In our experience this report is part of why students finish the survey and part of why they talk about it with each other afterwards.
It is a web platform now, at resonance.info. There is nothing to download and it works on a phone or a laptop.
Typically, two or three. We compute a compatibility score for every pair of students in your class and only make a match when that pair scores in the top five percent of all pairs — the 95th percentile or above. We have found that when there are 100 or more people who sign up for a class, there is always 1-2 above threshold matches for each student.
Because we have found that 100 is a key threshold for the algorithm to successfully find everyone a match, I am only contacting instructors of large classes. What we have found in past terms is that if the instructor gives out .50 (or even .25) pts of extra credit for completing the survey, setting up a profile, and joining the class group, 60-80% will happily join.
Your class comes first. When students join your class group, their matches are drawn from classmates. Students can also choose to join other communities and groups — their residence hall, another course, the university as a whole — and receive matches from those spaces too. That is entirely their choice, and it does not affect what happens inside your class.
Suppose a student in your class had a clone who gave exactly the same answers on every question. You might expect that pair to be the best possible match. It isn’t. In our data, a person and their clone score at about the 78th percentile of all pairs — better than most, but nowhere near the top. Resonance only matches pairs at the 95th percentile or above. So a student would essentially never be matched with someone identical to them; the pairs that clear our bar are always people who differ from each other in specific ways that turn out to matter for connection. Seeing things in compatible ways is part of the story of friendship. Seeing things identically is not.
Two other points. Our algorithm ignores race, gender, sexual orientation, nationality, and major entirely, which means it routinely pairs students who might never have found each other across the campus’s usual demographic lines. And the alternative for many students is not a more diverse friend group; it is no friend group at all. We take the echo-chamber concern seriously, but it is a small worry next to the larger worry of loneliness and isolation.
Matched students see each other’s profiles and a short note on why they were matched, and can message each other on the platform. Beyond that, two tools are built in. The first is a version of Art Aron’s Fast Friends procedure — the closeness-generating questions that have been used in social psychology for nearly thirty years — which students can run together on the platform. The second is a set of short games designed to get two strangers talking without either one having to make a move that feels like they are risking something. Students who want to go deeper use Fast Friends; students who want something lighter use the games. There is no feed, nothing to scroll, and nothing designed to keep them on the platform. If they leave because they made a friend, it worked.
None. You do not send me a roster, student emails, or anything else. Students sign up on their own with their university address and join your class group themselves. At the end of the term I can give you a list of who completed the process, for extra credit, but nothing flows from you to me.
Because I am not using a class email list, this part depends on you. The pattern that works is a brief mention in lecture with the link showing “resonance.info” on a slide, followed by an email in the next 24 hours. One reminder a few days later really helps. The reminders matter more than you would think: most students who intend to sign up and don’t are the ones who never got the nudge. I will send you a short announcement you can read or adapt, and a slide.
Extra credit is up to you. The default that has worked in other classes is about half a point toward the final grade for completing the survey, setting up a profile, and joining the class group. Students are free not to participate at all, and are also free to start the survey and not finish it; in either case they are not matched and their data is deleted. The platform requires an explicit opt-in before a student’s name is shared with anyone on the app (and then it is just their first name until they accept a connection from somoene. If you want an alternative route to the same credit for students who would rather not participate, I can give you an article on our mistaken beliefs about striking up conversations with strangers and they could submit a one page essay to me and then I will include them on the extra credit list for your class.
A second way to do extra credit is the ‘kickstarter’ approach where everyone in the class gets a certain amount of extra credit once at least 100 students have signed up, published their profile, and joined the class group. Its totally your call.
Flexible, but a sensible default is: students sign up and take the survey in one week of your class. Students will start getting class specific matches as soon as 100 students have signed up in your class. Some will get them much sooner if there happen to be matches that we know are well above ‘the bar’ for matches.
I do, personally. Give students my address — matt@resonance.info — and tell them to write to me directly if anything goes wrong. There are very few issues, but getting students successfully onboarded and benefiting from the app is my number one priority.
No. The survey never asks students what they believe, who they voted for, their religion, their sexuality, or anything about their health history. More importantly, their answers are never shown to anyone else. The app simply uses these responses to get the best matches for each person. You are welcome to take the survey yourself at resonance.info and see every question.
No. I created this app because I study social connection and want to help people be more connected. We do not sell ad space or give data to any third parties at all.
No. FERPA governs a university’s disclosure of education records — rosters, grades, and the like — to third parties without consent. Resonance receives no records from you or from your university. Students create their own accounts and consent directly to what they share. Because an alternative assignment is available for the extra credit, no student is compelled to use the platform to earn a grade. Participation is a personal choice, which keeps the arrangement outside the scope of FERPA.
Nothing, for students and for you. Individual use of Resonance is free, and a contract between Resonance and the UCLA Office of Student Affairs covers all student usage at UCLA.
Three things, with their limits stated.
The classroom rollout. Last winter and spring we ran Resonance in large UCLA classes with roughly 1,600 students in total. That is where the pattern I describe above — the announcement, the reminders, the extra credit — comes from.
The transfer-student pilot. An earlier pilot with about 2000 UCLA transfer students, a population that struggles with belonging on every campus, found that among students who completed the survey and then actually contacted their match, self-reported loneliness fell by roughly 41 percent relative to their own baseline.
The Fast Friends study. In a separate study, 200 UCLA students were paired at random and ran the Fast Friends procedure over video. Four in five enjoyed it more than they expected; two in three said they could see becoming real friends with their partner; and the pairs with higher Resonance compatibility scores were the ones more likely to say they planned to stay in touch. This is why Fast Friends is now incorporated into the app itself.
I am happy to share more detail on any of these, and to send you the aggregated feedback from your own class at the end of the quarter.
Still have a question? Write to me directly at matt@resonance.info — I answer these myself.