MTID Thesis draft


This thesis aims to explore ways of using tangible design elements to encourage people in public places to interact positively with each other and to learn something meaningful about themselves or the world. The foundational idea is that given a carefully designed opportunity, people can be brought to have enriching experiences, even in places (such as waiting spaces in the public transit system) that normally inspire more boredom than joy. The purpose of these interventions is to increase social cohesion in the broader public, particularly across lines of race, class, urbanity, and politics.


This background chapter consists of two sections:

  1. A theory and literature review section that situates the core ideas of the present work relative to previously published ideas. A logic model of the practical value of the current work is proposed.
  2. A critical review of related prior art, consisting of a study of the relative strengths and weaknesses of selected pieces for the purpose of learning from them.

Social theory

The underlying claim of this thesis depends on social connection having value in that it helps to maintain modern civil society. But what is the value in maintaining or improving the state of civil society? Ultimately there’s something axiomatic about the belief that it is good for people to share positive interactions with each other. It is a leap that we will make.

So: beginning from the belief that there is value in bringing people together positively, how might a designer of tangible interactions in the public sphere go about achieving that end? Thankfully there is a body of empirical research alongside well-developed theory to provide guidance on this question.

People are happier when they’re connected, though they don’t know it

Social rules generally prohibit unnecessary meaningful interactions with strangers

Why do many people have a revulsion to speaking with strangers? There is a complicated history at play and it's not a question that's easily answered. However, some foundational social science writing by Goffman1 addresses the question nicely, and even though the work is now more than fifty years old it still feels in many ways like it describes contemporary norms. In the eighth chapter of this now-canonical work, "Engagements among the Unacquainted," Goffman mentions some categories of people who are generally allowed to be engaged by strangers without any need for social justification. These include police officers, priests wearing their vestments, and also the very old and the young. The general rule is that unless a stranger fits into a special category like these select few, one may not initiate meaningful communication with them.

But in addition to the members of these special categories, Goffman points out that otherwise regular people who are temporarily "out of role" can be engaged by strangers; examples of this sort of person are drunk people or people dressed in clown costumes. Why? “Presumably on the the assumption that the self projected through these activities is one from which the individual can easily dissociate himself, and hence need not be jealous of or careful with."2 Goffman’s read is that alternative selves can be commented on socially, engaged with safely, etc., because any commentary on this alternative does no injury to the “real” core of the person in question. The emotional protection of the ego, of the inner self, emerges as the primary rule in all social interaction.

Bigger cities produce less friendly connections between strangers

Goffman coins a lovely term to describe one difference between small-town and big-city expectations of comportment:

“In Anglo-American society there exists a kind of ‘nod line’ that can be drawn at a particular point through a rank order of communities according to size. Any community below the line, and hence below a certain size, will subject its adults, whether acquainted or not, to mutual greetings; any community above the line will free all pairs of unacquainted persons from this obligation.” (Goffman p. 132–3.)

This is a quaint formulation of an idea that is supported by folk wisdom: people in small towns are friendlier. Decades after Goffman's book, a wide-ranging sociological investigation by Levine et al.3 of kindness and courtesy in 24 small, mid-size, and large American cities experimentally substantiated this message.

The researchers did a few tests where they artificially generated situations calculated to encourage strangers to help a person in need. They refer to them as "helping measures," and in increasing order of urgency, they were:

  1. "Change for a Quarter": a researcher holding a quarter in their hand asks if an oncoming walking person can give them change for it.
  2. "Dropped pen": a researcher walking down the street drops a pen out of their pocket before crossing paths with the test subject to find whether they will say or do anything to help.
  3. "Hurt leg": “walking with a heavy limp and wearing a large and clearly visible leg brace,” the researcher drops a big pile of magazines on the ground and reaches for them but can’t pick them up to find if anybody will help.4

The findings were fairly damning for large cities (which in this case means a population greater than 4 million). Of the 24 cities studied, 8 of them were large, and in the ranked list of helpingness, large cities scored 5th (nicely done, Dallas), 17th, 18th, 20th, 21st, 22nd, 23rd, and 24th. New York City was dead last and in America's largest metropolis, 13% of people helped a visibly disabled person who had dropped some magazines.5 The statistical result was clear: "Population size and density showed significant negative correlations all of the measures of helping.”6

Strangers like interacting—they just don't know it

An interesting recent study by Epley and Schroeder7 used empirical research on subjects traveling on public transit in Chicago to investigate the extent to which people enjoy social interaction. Many people are understandably afraid of social interaction with strangers—just on social grounds, it is inherently risky to talk to somebody who is not an acquaintance in a public space. The stranger may lash out, or, worse, even take offense to the friendly overture. And so it’s understandable that this study found that most respondents said speaking with a stranger would worsen their daily commute on a train or bus.

The subjects in the study were given the opportunity to put that prediction to the test: a group of study members were asked to start a conversation with somebody on the train or bus they were riding, whereas another group was asked to keep to themselves. (A third control group was given no special instructions regarding social interaction.) If social interaction with strangers is risky and off-putting, as participants tended to assert prior to the study, then one would expect those tasked with engaging socially to have bad experiences.

In fact, the opposite was true. In statistically robust findings, the authors explained that “people liked their conversation partners, had pleasant conversations, and had more positive commutes the longer their conversations lasted.”8 It’s an impressive repudiation of the notions most people have about strangers, and it suggests that perhaps getting people to engage may be functionally introducing a minor discomfort in the service of a larger payoff.

Further confirmation of this same finding came from a study by Sandstrom and Dunn, which asked study subjects to engage in a small-talk conversation with their coffeeshop barista (and asked control subjects not to speak unnecessarily with the barista).9 The study aimed to compare the "inefficiency" that some may think would arise from small talk with the social benefits to those engaged in that small talk.

Just as in the Epley and Schroeder study, findings were that people reported greater “positive affect” and lesser “negative affect” after they had a small engagement with their barista versus when they were more “efficient” in that they just grabbed their coffee and ran.10 It's worth noting that the coffeeshop in question was a busy one, at an urban location on a weekday afternoon, which is the type of environment in which a customer may think that transactional efficiency is more important than chatting in the moment. The authors suggest that it's an advantageous habit to intentionally turn a stranger into a "weak tie" (see discussion of Granovetter's term, below) for the purpose of creating greater all-around positive affect among both interactants.

People exhibit live physiological mirror responses to strangers with whom they have even a weak connection

There is a wide berth between a complete stranger and one with whom somebody has some sense of camaraderie, however distant. This has been studied in a variety of ways—perhaps most famously as described by Granovetter's landmark paper, "The Strength of Weak Ties."11 Granovetter explores theoretically how a multiplicity of loose ("weak") connections between people, as opposed to the "strong" ties such as those between close friends or family, can actually be equally or more important to defining the strength of an individual's overall social network.

Consider the case of two people who have a car accident on the highway. Initially speaking angrily to each other, they discover that they hail from the same town. This simple connection, meaningless as it may be given the circumstance, may significantly diffuse tensions between the drivers; they are no longer complete strangers since they've located a meaningful, albeit weak, tie. An interesting study by Cwir et al. explores the power of this sort of seemingly unimportant connection between people who are otherwise total strangers.12

The study divided subjects into two groups. Subjects in the experimental group were introduced to a stranger (actually a research confederate) and, based on a small opinion survey the subject had filled out about a week prior, the subject would discover to her surprise that the confederate shared three out of her five opinions. Subjects in the control group matched zero out of five opinions.

Then the confederate was made to run in place for three minutes as the subject sat nearby in a chair. At the end of the stranger's exercise session, researchers measured the blood pressure and heart rate of the subjects and found that the greater sense of social connection between subject and confederate correlated with a physiological connection as well: socially-connected subjects' heart rates went up by an average of 5 bpm while socially-unconnected subjects' heart rates went up only 0.5 bpm.

People distrust or dislike people from outside groups

Just as test subjects demonstrated a shared physiological response to people with whom they shared some opinions, a study by Zebrowitz et al.13 examined people's responses to images of faces of strangers from different racial groups. In short, perhaps unsurprisingly the research found that people had more negative responses to images of members of different racial groups than their own. The authors propose the "familar face overgeneralization hypothesis" (FFO), which asserts an evolutionary basis in humans' preference for members of their own familiar group. The reasoning is that historically, humans were right to be wary of outsiders who may pose a threat to their community; outsiders can be defined as such in part because they look different than the ingroup members.

This would appear to be a difficult to surmount obstacle to efforts to create social cohesion, if indeed there are literally biological factors that are encouraging people to distrust or dislike members of other racial groups. However, the Zebrowitz et al. study did not merely examine people's aversion to outsiders: it also found a surprisingly simple mechanism of reducing that aversion.

How might one go about making a visually distinct outsider more appealing? According to the findings, simply by showing images of members of that person's racial group to a subject. Simply seeing a variety of images of East Asian or Black people significantly increased the odds that a white test subject would respond more positively to an image of another person of that race. This is referred to in the psychological literature as the "mere exposure" effect—namely, that simply being exposed to a stimulus like a person's face (i.e. merely being exposed) can increase one's positive feelings towards that stimulus. The familiar, in other words, is preferred over the unknown.

This seems like a fairly obvious conclusion and not especially noteworthy, but an interesting wrinkle about the mode of exposure people had to outgroup faces illustrates the depth at which this type of judgment is taking place. The researchers exposed white subjects to images of East Asian and Black faces supraliminally, meaning that the subjects were able to see and be aware of seeing the faces. In these cases, the authors found that people responded as described above. However, the researchers also exposed different subjects to subliminal images of East Asian and Black faces, meaning that the exposure times to the images were so brief that they did not consciously register that they'd seen those faces. Even in the subliminal cases, they found tound that the subjects' attitudes changed in the same way, even though they were not conscious of the exposure they'd been subjected to.

The practical implications of this research are significant: it demonstrates that mere exposure, even if conducted fleetingly, can be a meaningful way to increase positive associations across racial divides.

Various measures of social connection show declines on small and large scales

Casual civic engagement has been declining in America for decades

Robert Putnam's 1995 article "Bowling Alone" launched a wave of research aimed at understanding decades of informal civic disengagement in the US.14 Putnam's claim was that in many ways, casual ties between members of civic society had been eroding since the postwar heyday of the fifties.

He cites a variety of declining trends indicating the same pattern: turnout at elections, public affairs meetings, houses of worship, the Lions, the Elks, and the Masons. Union membership has declined, as has the number of people volunteering for Boy Scouts and the Red Cross.15 The title of the article is drawn from a shift in the way people recreationally bowl: whereas there were for a time a great many thriving bowling leagues, bringing together friends and neighbors in regular friendly competition, Putman cites statistics showing that while there are more people bowling in America today than in the past, they are now bowling alone instead of in leagues.

Putnam suggested some explanantions for these social trends, and writing in 1995, suggested that one of the contributing factors was the "technical transformation of leisure," which he observed tended to give people the opportunity to engage in solo leisure activities like television.16 Of course the extent to which this was true twenty years ago has been dwarfed by the current state of affairs, where it's no longer remarkable to see a public space full of people literally all of whom are preoccupied with media or games on their mobile phones. Barring some unexpected large-scale social change of course, surely this trend towards increasingly individualized and enveloping media engagement will continue.

The urban–rural divide in particular is increasing

In social science it is frequently difficult to come by reliable numerical data that captures people's thoughts and feelings. Many surveys extrapolate from a relatively small sample group to make claims about larger populations, but the cost of doing larger surveys is quite significant. Fortunately, we have a survey mechanism in place in which millions of American adults voluntarily participate on a regular basis: elections.

There were 137 million participants17 in the U.S. 2016 general election, representing adult Americans from every walk of life. Geographical analysis of the election returns is based on an extremely wide-ranging dataset, and sheds light on the political disposition of every part of the country. While using political data to make claims about social division is an inherently risky maneuver, there are some obvious alignments between many people's political and social preferences, a fact which national campaigns take careful advantage of.

The elections metric that is most revealing about the state of social division across the rural/urban divide is a simple one: using counties as a grouping mechanism, it's easy to examine each local area's returns and measure the difference in votes that the 1st and 2nd place candidates had in that area. Any county with a greater than 20% disparity between first and second place (i.e. any county that's 60/40 or more split) can be called a "landslide" county, in Bill Bishop's language.18 In a de-facto two-party country in which both parties vie aggressively for the median voter, the finding that 60% of voters now live in "landslide" counties illustrates the emerging divisions quite clearly. In fact, according to Bishop's analysis, in 2016 80% of voters who live in "rural and remote" areas lived in landslide counties. This is a huge increase; in 2004 the equivalent metric was 59%.19 Conversely, some of the country's most-landslide counties are in cities; Washington, D.C., for instance, divided 90/4.20

Evidence-based ways of closing the social gap

Best practices for creating social connections

A variety of techniques have been explored for the purpose of creating positive social connections between strangers. Some of these are mentioned above in the context of Epley and Schroeder's research (connecting commuters on public transit) and Sandstrom and Dunn's research (encouraging people to chat with their barista), among others. Both these lines of study found efficacy in connecting strangers and producing a greater sense of positive affect among the experimental subjects. However, the experiments these represent were contrivances with informed participants—in this way they're not representative of the sort of social intervention between strangers at the heart of the present research.

So while there are certainly lessons to be learned from the studies mentioned above, further research conducted between uninformed participants is especially relevant.

Creating interpersonal closeness {this section probably needs relocated}

Given two unacquainted people, how might they be brought to have some measure of interpersonal closeness? While the outcome is notoriously difficult to measure, research has shed light on some good answers. A study conducted by Aron et al.21 aimed to test a proposed procedure for achieving "closeness" between study subjects, where closeness had the wonderful definition of "'including the other in the self'—creating an interconnectedness of self and other."22

The general procedure the researchers were testing for generating closeness was "sustained, escalating, reciprocal, personalistic self-disclosure."23 This seems like an failsafe recipe for generating closeness, as the basis of meaningful relationships is mutual trust and disclosure. It is, however, not necessarily easy to force people into the emotional crucible of sharing their innermost thoughts, and one imagines that some of the study subjects were made uncomfortable at times discussing deeply personal matters with people they'd just met, e.g. "when did you last cry in front of another person? By yourself?" and "of all the people in your family, whose death would you find most disturbing? Why?"24 Quickly building rapport through mutual self-disclosure like this may be effective at generating "closeness," but it is far too top-down and potentially emotionally burdensome to impose on strangers in a public space who are not explicitly electing to participate in research.

Experimental interactive objects in the public space

This thesis research has the intention of producing several interactive pieces in public space, and there is prior social science work that is well-aligned with this goal in particular. In addition to basic research in social psychology, there is a rich stream of work coming from the interactive museum field that pertains to the present thesis.

Installing an interactive object in a public space

A recent study by Balestrini et al.25 covers nearly exactly the same subject area as this thesis research: an installation in a bus stop in an urban setting in Mexico which aimed to create social interactions among strangers. The researchers devised a setup designed to foster interactions by guiding users through a simple pathway of engagement. Two large red buttons with internal lights were set up on small podiums about three meters apart. If one of the buttons was pushed, the other one would start blinking, indicating that it needed to be pushed to complete the activity. The spacing of the podiums and timing of the buttons meant that two people would need to cooperate to complete the button-pushing task.

Once triggered by the two pushes in rapid succession, the system would then reward users by telling them a lighthearted joke through a loudspeaker, which is why the system is called the "Jokebox" by the researchers. (The jokes were randomly selected from a list of jokes that had been judged through a survey to be popular among locals.) The system was very effective at creating excitement and enthusiasm among users, even producing very interesting ancillary social structures around the system that developed spontaneously. For instance, the researchers report three users they describe as "champions" of the Jokebox, insofar as they would recruit others to try the system out, explaining as necessary, evangelizing about its fun and value. An additional effect they observed is the so-called "honeypot," where once a pair of people has engaged with the thing and appear to be having fun, others are attracted to come and try it for themselves, frequently forming a chain of successive interactions after the first one has begun.26

  1. Goffman, Erving. Behavior in Public Places. The Free Press of Glencoe: New York, New York. 1963.

  2. Goffman p. 126.

  3. Levine, Robert V., Stephen Reysen, and Ellen Ganz. "The kindness of strangers revisited: a comparison of 24 US cities". Social Indicators Research. vol. 85. 2008. pp. 461–481.

  4. Levine et al. p. 467–8.

  5. Levine et al., table 2, p. 471.

  6. Levine et al., p. 474.

  7. Epley, Nicholas and Juliana Schroeder. “Mistakenly Seeking Solitude”. Journal of Experimental Psychology: General (2014). vol. 143, no. 5. pp. 1980–1999.

  8. Epley and Schroeder, p. 1984.

  9. Sandstrom, Gillian and Elizabeth Dunn. “Is Efficiency Overrated?: Minimal Social Interactions Lead to Belonging and Positive Affect”. Social Psychology and Personality Science. vol. 5, no. 4. 2014. pp. 437–442. doi: 10.1177/1948550613502990

  10. Sandstrom and Dunn, p. 439.

  11. Granovetter, Mark. “The Strength of Weak Ties”. American Journal of Sociology. vol. 78, no. 6. May, 1973. pp. 1360–80.

  12. Cwir, David, Priyanka B. Carr, Gregory M. Walton, and Steven J. Spencer. "Your heart makes my heart move: Cues of social connectedness cause shared emotions and physiological states among strangers". J Experimental Social Psychology. vol 47. 2011. pp. 661–4. doi: 10.1016/j.jesp.2011.01.009

  13. Zebrowitz, Leslie, Benjamin White, and Kristin Weineke. "Mere Exposure and Racial Prejudice: Exposure to Other-Race Faces Increases Liking for Strangers of that Race". Social Cognition. vol. 26, no. 3. 2008. pp. 259–275. doi: 10.1521/soco.2008.26.3.259

  14. Putnam, Robert D. “Bowling Alone: America’s Declining Social Capital”. Journal of Democracy. vol. 6., no. 1. January 1995. pp. 65-78. doi: 10.1353/jod.1995.0002

  15. Putnam p. 69–10.

  16. Putnam p. 75.

  17. Wikipedia article "United States presidential election, 2016". accessed 12/12/2016. available,_2016

  18. Bishop, Bill. "Caught in a Landslide". article in "The Daily Yonder". accessed 12/1/16. available at

  19. Bishop, top graphic:

  20. District of Columbia Board of Elections General Election 2016—Certified Results. accessed 12/12/16. available:

  21. Aron, Arthur, Edward Melinat, Elaine N. Aron, Robert Darrin Vallone, and Renee J. Bator. “The Experimental Generation of Interpersonal Closeness: A Procedure and Some Preliminary Findings”. Personality and Social Psychology Bulletin. April 1997. Vol. 23, no. 4. p. 363–377. doi: 10.1177/0146167297234003

  22. Aron et al., p. 364. The first part of this definition is attributed to two prior papers by a subset of the same authors.

  23. Aron et al., p. 364

  24. Aron et al., pp. 374 and 375, respectively

  25. Balestrini, Mara, Paul Marshall, Raymundo Cornejo, Monica Tentori, Jon Bird, and Yvonne Rogers. “Jokebox: Coordinating Shared Encounters in Public Spaces”. Proceedings of the 19th ACM Conference on Computer-Supported Coopertive Work & Social Computing (2016). p. 38–49. doi: 10.1145/2818048.2835203

  26. Balestrini et al., p. 39

  27. Exploratorium, Pier 15, San Francisco, CA 94111.

  28. Hipschman, Ron. Exploratorium Cookbook II; a Construction Manual for Exploratorium Exhibits. revised edition. The Exploratorium: San Francisco. 1986.

  29. artist's website currently has information about prior iterations of bubble devices:

  30. Children's Museum of Pittsburgh, 10 Children's Way, Allegheny Square, Pittsburgh, PA 15212.

  31. further information at the artist's website:

  32. Children's Museum of Pittsburgh, 10 Children's Way, Allegheny Square, Pittsburgh, PA 15212.

  33. artist's website:

  34. Children's Museum of Pittsburgh, 10 Children's Way, Allegheny Square, Pittsburgh, PA 15212.

  35. artists' website:

  36. New York Hall of Science, 47-01 111th St., Corona, NY, 11368.

  37. artist's website:

  38. artists' website:

  39. artists' website:

  40. artist's website:

  41. artists' website:

  42. artist's website:


  44. artist's website:

  45. artists' website:

  46. artists' seemingly nonfunctioning website:; installation featured as part of "Love Has No Labels" campaign by the Ad Council:

  47. see relevant Wikipedia entry

  48. YouTube video link

  49. Balestrini, M., P. Marshall, R. Cornejo, M. Tentori, J. Bird, and Y. Rogers. "Jokebox: Coordinating Shared Encounters in Public Spaces". Proceedings of the 19th ACM Conference on Computer-Supported Coopertive Work & Social Computing (2016). p. 38–49. doi: 10.1145/2818048.2835203

  50. artists' website:

  51. artists' website: