UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Third-placeness : supporting the experience of third place with interactive public displays Roberto, Calderon 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


24-ubc_2016_september_calderon_roberto.pdf [ 18.58MB ]
JSON: 24-1.0307293.json
JSON-LD: 24-1.0307293-ld.json
RDF/XML (Pretty): 24-1.0307293-rdf.xml
RDF/JSON: 24-1.0307293-rdf.json
Turtle: 24-1.0307293-turtle.txt
N-Triples: 24-1.0307293-rdf-ntriples.txt
Original Record: 24-1.0307293-source.json
Full Text

Full Text

Third-placenessSupporting the Experience of Third Place with InteractivePublic DisplaysbyRoberto CalderonM.Sc., The University of British Columbia, 2009B.Arch., Universidad Iberoamericana, 2005A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinThe Faculty of Graduate and Postdoctoral Studies(Interdisciplinary Studies)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)July 2016© Roberto Calderon 2016AbstractIn contemporary western cities, socialization often occurs in locations witha mix of public and private characteristics. Oldenburg defined these settings as“Third Places” because they provide a space of conviviality in between the pri-vacy of home and the rigidity of work. Coffee shops and pubs are some of theprototypical Third Places providing a welcoming and neutral atmosphere for con-versation that is essential to community development. Consumer computing andtelecommunications have impacted how we socialize with each other and use ThirdPlaces. This brings about the question of how technology can support Third Placesor if technology has a role at all in these settings.We propose an alternative paradigm called “Third-placeness” defined as a stateof socialization, of which a Third Place is a physical embodiment. Third-placenessarises when information is uncensored, which minimizes inequalities and differ-ences, and is characterized by low barriers to information access, regularity, light-heartedness and comfort. We identify aspects of Third-placeness and study howa particular type of technology, interactive public displays, could affect these as-pects. Through our observations and lessons learned we identify social, public, andphysical characteristics of interactive public displays that could support aspects ofThird-placeness. Our research contributes a framework, the Sociality, Publicityand Physicality Framework, that organizes aspects and requirements of designinginteractive public displays for Third-placeness. It also describes a way in which tocommunicate about these designs and a way such designs can be approached.iiPrefaceThe motivation and inspiration for the concept of “Third-placeness” came froma number of sources. Chapter 3 is based on discussions that were born from thefirst Workshop on Human Computer Interaction for Third Places (HCI3P) [1] andthe second Workshop on Human Computer Interaction for Third Places (HackingHCI3P) [2]. I was responsible for organizing and chairing both workshops. Fieldactivities studying Parisian Third Places in HCI3P [1] yielded a collaboration paperwhere we stated the need to revisit the concept of Third Places [3]. I contributedbackground knowledge on the emergence of the concept of Third Place, reexam-ination of Oldenburg’s original characteristics of Third Place and data from fieldobservations. The need to discuss the role of technologies for Third Place Commu-nities was born from a collaboration with Dr. Vania P. de Almeida N. and Dr. JuniaAnacleto to organize the First Interdisciplinary Workshop on Communication forSustainable Communities [4]Chapter 4 is based on research conducted in UBC’s Human CommunicationTechnologies Lab with Dr. Sidney Fels and Dr. Junia Anacleto, along with researchconducted in UBC’s Media And Graphics Interdisciplinary Centre (MAGIC) un-der Dr. Sidney Fels, Dr. Junia Anacleto and Dr. Rodger Lea. In the research pre-sented I was responsible for designing the experiments, conducting co-design ses-sions, developing the technology, conducting ethnographic observations, applyingquestionnaires, conducting interviews and analyzing the data. An earlier prototypeof the interactive garden was co-developed and tested with fellow student AndreO. Bueno [5]. Underlying technologies used in the prototype to allow multi-displayinteractivity were co-developed with fellow student Ricardo A. Perez-de-Almeida,Dr. Michael Blackstock and Andre O. Bueno [6, 7]. The research goal of measuringhow differences in game play affect social capital is a continuation of preliminarywork conducted in collaboration with student Marcos Alexandre [8, 9].iiiRelevant PublicationsChapter 5 is based on research conducted in collaboration between UBC’s Me-dia And Graphics Interdisciplinary Centre (MAGIC) and UFSCAR’s AdvancedInteraction Laboratory in Brazil under Dr. Sidney Fels, Dr. Junia Anacleto andDr. Rodger Lea. Ethnographic observations for this chapter were conducted withfellow student Joˆnatas Oliveira [10, 11], of which I was responsible for designingthe ethnographic approach, writing behavioural and architectural notes and analyz-ing data gathered. Similarly, the ddesign, development and deployment of the pro-totoype was also conducted in collaboration with fellow student Joˆnatas Oliveira[12], for which I was responsible for designing the experiment, developing thetechnology and analyzing the data gathered.Numbers for UBC Research Ethic Board’s Certificates of Approval for all re-search reported in this thesis are listed in Table 1.Certificate Name Start EndH10-01651 Community Displays 2010/07/12 2015/02/13H14-00394 Interactive Community Displays 2014/03/06 2015/03/06H14-01620 HCI3P 2014/06/30 2016/06/03Table 1: UBC Research Ethics Board’s Certificates of Approval for all researchreported in this thesis.Relevant PublicationsThis work yielded several publications and presentations through conferenceproceedings, workshop organization and participation, and collaboration publica-tions. Collaboration publications were born from key collaborations with otherresearch groups.Conference ProceedingsR. Calderon, M. Blackstock, R. Lea, S. Fels, A. Bueno, and J. Anacleto, “RED:A framework for prototyping multi-display applications using web technologies,”in Proceedings of the international symposium on Pervasive displays, PerDis ’14,(New York, NY, USA), pp. 148–153, ACM, 2014ivRelevant PublicationsR. Calderon, M. Blackstock, R. Lea, S. Fels, A. Bueno, and J. Anacleto, “Support-ing conversation and community interaction with a table-top community gardenapplication,” in Proceedings of the international symposium on Pervasive dis-plays, PerDis ’14, (New York, NY, USA), pp. 190–191, ACM, 2014R. Calderon, S. Fels, and J. Anacleto, “Towards understanding the effects of pub-lic displays on the functioning of a Brazilian health-care centre,” in Proceedingsof INTERACT 2013, 2013R. Calderon, R. Lea, M. Blackstock, and S. Fels, “Developing cross-display ap-plications using the really easy displays (RED) framework,” in 2013 Internationalsymposium on Pervasive displays, (Mountain View, California), 2013R. Calderon, S. Fels, J. Anacleto, and J. Leite de Oliveira, “Towards supportinginformal information and communication practices within a Brazilian healthcareenvironment,” in CHI ’13 Extended abstracts on Human factors in computingsystems, CHI EA ’13, (New York, NY, USA), pp. 517–522, ACM, 2013R. Calderon, S. Fels, J. L. de Oliveira, and J. Anacleto, “Understanding NUI-supported nomadic social places in a Brazilian health care facility,” in Proceed-ings of the 11th Brazilian symposium on Human factors in computing systems,IHC ’12, pp. 76–84, Porto Alegre, Brazil: Brazilian Computer Society, 2012R. Calderon, S. Fels, R. Lea, and O. Neumann, “Harvesting communal interactionwith public displays through place-dependent community displays,” in Workshopon Large displays in urban life. CHI2011 Workshop, 2011Workshop OrganizationR. Calderon, S. Fels, J. Anacleto, N. Memarovic, and T. Thompson, “HackingHCI3P: second workshop on human computer interaction in third places,” in Pro-ceedings of the 2014 companion publication on Designing interactive systems,DIS Companion ’14, (New York, NY, USA), pp. 195–198, ACM, 2014R. Calderon, S. Fels, and J. Anacleto, “Workshop on human computer interactionin third places,” in CHI ’13 Extended abstracts on Human factors in computingsystems, CHI EA ’13, (New York, NY, USA), pp. 3167–3170, ACM, 2013vRelevant PublicationsIWCSC ’10: first interdisciplinary workshop on Communication for sustainablecommunities, SIGDOC 2010, (New York, NY, USA), ACM, 2010Collaboration PublicationsN. Memarovic, S. Fels, J. Anacleto, R. Calderon, F. Gobbo, and J. M. Carroll,“Rethinking third places: contemporary design with technology,” The journal ofcommunity informatics, vol. 10, no. 3, 2014A. Bueno, J. C. Anacleto, R. Calderon, S. Fels, and R. Lea, “ICT to support com-munity gardening: a system to help people to connect to each other in real life,”in Proceedings of the 2014 companion publication on Designing interactive sys-tems, DIS Companion ’14, (New York, NY, USA), pp. 133–136, ACM, 2014R. A. Perez de Almeida, M. Blackstock, R. Lea, R. Calderon, A. F. do Prado, andH. C. Guardia, “Thing Broker: a Twitter for things,” in Proceedings of the 2013ACM conference on Pervasive and ubiquitous computing, UbiComp ’13, (NewYork, NY, USA), pp. 1545–1554, ACM, 2013viTable of ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiPreface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiiTable of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiList of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiList of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiiAcknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiv1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Research Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . 31.2 Research Approach . . . . . . . . . . . . . . . . . . . . . . . . . 41.2.1 Literature Review . . . . . . . . . . . . . . . . . . . . . 51.2.2 Development of the Concept of Third-Placeness . . . . . 51.2.3 Supporting Aspects of Third-placeness with Technology . 61.3 Summary of Contributions . . . . . . . . . . . . . . . . . . . . . 71.4 Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92.1 Third Places and their Communities . . . . . . . . . . . . . . . . 92.1.1 Aspects of Places . . . . . . . . . . . . . . . . . . . . . 102.1.2 Aspects of Third Places . . . . . . . . . . . . . . . . . . 112.1.3 Aspects of Social Capital . . . . . . . . . . . . . . . . . 152.2 Interactive Public Displays . . . . . . . . . . . . . . . . . . . . . 17viiTable of Contents2.2.1 Aspects of Displays . . . . . . . . . . . . . . . . . . . . 182.2.2 Aspects of Public Displays . . . . . . . . . . . . . . . . 202.2.3 Aspects of Interactivity in Public Displays . . . . . . . . 232.3 Related Work on Technology for Third Places . . . . . . . . . . . 262.3.1 Interactive Public Displays in Third Places . . . . . . . . 262.3.2 Social Capital and Network Analysis in HCI . . . . . . . 292.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313 Third-placeness and the Sociality, Publicity, Physicality Framework 333.1 Third-placeness . . . . . . . . . . . . . . . . . . . . . . . . . . . 333.1.1 Motivation for Revisiting the “Third Place” . . . . . . . . 333.1.2 Revisiting Oldenburg’s Third Place . . . . . . . . . . . . 363.1.3 Aspects of Third-placeness . . . . . . . . . . . . . . . . 413.2 Sociality, Publicity and Physicality Framework . . . . . . . . . . 423.2.1 Framework Design Dimensions . . . . . . . . . . . . . . 443.2.2 Framework Design Circle . . . . . . . . . . . . . . . . . 463.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 474 Study of Supporting Aspects of Third-placeness With Interactive Pub-lic Displays in a Coffee Shop . . . . . . . . . . . . . . . . . . . . . . 504.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 524.2 MySeedlings Design . . . . . . . . . . . . . . . . . . . . . . . . 524.2.1 Components of Prototype . . . . . . . . . . . . . . . . . 544.2.2 Game Conditions . . . . . . . . . . . . . . . . . . . . . 624.3 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654.3.1 Recruitment and Participant Overview . . . . . . . . . . 654.3.2 Experimental Design . . . . . . . . . . . . . . . . . . . . 664.3.3 Measures . . . . . . . . . . . . . . . . . . . . . . . . . . 664.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 694.4.1 Social Capital Surveys . . . . . . . . . . . . . . . . . . . 694.4.2 Social Network Analysis . . . . . . . . . . . . . . . . . . 694.4.3 Characteristics of Third-placeness . . . . . . . . . . . . . 754.4.4 Surveys and Informal Interviews . . . . . . . . . . . . . 77viiiTable of Contents4.4.5 Ethnographic Observations . . . . . . . . . . . . . . . . 804.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 844.5.1 Sociality . . . . . . . . . . . . . . . . . . . . . . . . . . 874.5.2 Publicity . . . . . . . . . . . . . . . . . . . . . . . . . . 874.5.3 Physicality . . . . . . . . . . . . . . . . . . . . . . . . . 884.5.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . 885 Study of the Impact of Information Visualization on Community Reci-procity in a Workplace . . . . . . . . . . . . . . . . . . . . . . . . . 905.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 925.2 Ethnographic Observations . . . . . . . . . . . . . . . . . . . . . 925.3 Ticketing System Design . . . . . . . . . . . . . . . . . . . . . . 995.3.1 Prototype Components . . . . . . . . . . . . . . . . . . . 1005.3.2 Visualizations Used . . . . . . . . . . . . . . . . . . . . 1015.4 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1035.4.1 Recruitment and Participant Overview . . . . . . . . . . 1035.4.2 Experimental Design . . . . . . . . . . . . . . . . . . . . 1055.4.3 Measures . . . . . . . . . . . . . . . . . . . . . . . . . . 1065.5 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1085.5.1 Network Analysis . . . . . . . . . . . . . . . . . . . . . 1085.5.2 Semantic Analysis . . . . . . . . . . . . . . . . . . . . . 1125.5.3 Questionnaire and Interviews . . . . . . . . . . . . . . . 1135.6 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1145.6.1 Sociality . . . . . . . . . . . . . . . . . . . . . . . . . . 1155.6.2 Publicity . . . . . . . . . . . . . . . . . . . . . . . . . . 1165.6.3 Physicality . . . . . . . . . . . . . . . . . . . . . . . . . 1165.6.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . 1176 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1196.1 Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1206.2 Initial Design Guidelines . . . . . . . . . . . . . . . . . . . . . . 1236.2.1 Sociality . . . . . . . . . . . . . . . . . . . . . . . . . . 1236.2.2 Publicity . . . . . . . . . . . . . . . . . . . . . . . . . . 124ixTable of Contents6.2.3 Physicality . . . . . . . . . . . . . . . . . . . . . . . . . 1256.3 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1266.4 Directions for Future Work . . . . . . . . . . . . . . . . . . . . . 1276.5 Lessons Learned . . . . . . . . . . . . . . . . . . . . . . . . . . 1286.6 Concluding Comments . . . . . . . . . . . . . . . . . . . . . . . 131Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132AppendicesAppendix A – Background Review Taxonomies . . . . . . . . . . . . . 154Appendix B – Surveys and Interviews . . . . . . . . . . . . . . . . . . . 160Appendix C – Research Methodologies . . . . . . . . . . . . . . . . . . 173xList of Tables3.1 Third-placeness: Aspects of Third-placeness . . . . . . . . . . . . 403.2 Third-placeness: Dimensions of Third-placeness . . . . . . . . . 423.3 Framework: Interactive public displays and communities . . . . . 443.4 Framework: Modules . . . . . . . . . . . . . . . . . . . . . . . . 453.5 Framework: Design Dimensions . . . . . . . . . . . . . . . . . . 463.6 Framework: Sociality, Publicity and Physicality . . . . . . . . . . 484.1 MySeedlings: Individual game play payoff-matrix . . . . . . . . . 634.2 MySeedlings: Communal game play payoff-matrix . . . . . . . . 644.3 MySeedlings: Favours game play payoff-matrix . . . . . . . . . . 644.4 MySeedlings: Participant overview . . . . . . . . . . . . . . . . . 664.5 MySeedlings: Deployment schedule and treatments . . . . . . . . 664.6 MySeedlings: Instruments used . . . . . . . . . . . . . . . . . . . 674.7 MySeedlings: Density and reciprocity coefficients of favours . . . 714.8 MySeedlings: Centrality of favours . . . . . . . . . . . . . . . . . 724.9 MySeedlings: Density and reciprocity coefficients of gifts . . . . . 734.10 MySeedlings: Centrality of gifts . . . . . . . . . . . . . . . . . . 745.1 C.A.I.S. Third-placeness characteristics observed . . . . . . . . . 955.2 C.A.I.S. Participant overview . . . . . . . . . . . . . . . . . . . . 1045.3 C.A.I.S. Research instruments used . . . . . . . . . . . . . . . . . 1055.4 C.A.I.S.: Deployment schedule and treatments . . . . . . . . . . . 1065.5 C.A.I.S. Analysis of density and reciprocity coefficients . . . . . . 1105.6 C.A.I.S. Analysis of centrality . . . . . . . . . . . . . . . . . . . 1115.7 C.A.I.S. Summary of text analysis . . . . . . . . . . . . . . . . . 112xiList of Figures2.1 Background: A Parisian street with several social public places . . 132.2 Background: Displays found in technology-prone public spaces . 192.3 Background: Examples of display content found in a public space 212.4 Background: Displays, their content and types of interactions . . . 243.1 Framework: Sociality, Publicity, Physicality Framework design circle 494.1 MySeedlings: Seedlings Cafe . . . . . . . . . . . . . . . . . . . . 514.2 MySeedlings: Interaction flow storyboard . . . . . . . . . . . . . 554.3 MySeedlings: Large display application . . . . . . . . . . . . . . 564.4 MySeedlings: Test conditions in the large display application . . . 574.5 MySeedlings: Mobile application; signin and people pages . . . . 594.6 MySeedlings: Mobile application; private and public profile pages 604.7 MySeedlings: Prototype deployed . . . . . . . . . . . . . . . . . 614.8 MySeedlings: Social capital measures . . . . . . . . . . . . . . . 704.9 MySeedlings: Centrality of favours . . . . . . . . . . . . . . . . . 724.10 MySeedlings: Centrality of gifts . . . . . . . . . . . . . . . . . . 744.11 MySeedlings: Comparison of reciprocity of favours and gifts . . . 754.12 MySeedlings: Third-placeness measures . . . . . . . . . . . . . . 764.13 MySeedlings: Perception of prototype . . . . . . . . . . . . . . . 774.14 MySeedlings: Feelings of belonging to MySeedlings . . . . . . . 784.15 MySeedlings: Observed customers and regulars of Seedlings cafe 814.16 MySeedlings: Physical appropriation . . . . . . . . . . . . . . . . 834.17 MySeedlings: Observed interactions with MySeedlings’ display . 854.18 MySeedlings: Usage, checkins and favours . . . . . . . . . . . . 865.1 C.A.I.S. Clemente Ferreira . . . . . . . . . . . . . . . . . . . . . 935.2 C.A.I.S. Existing information relaying . . . . . . . . . . . . . . . 945.3 C.A.I.S. Existing sociability while working . . . . . . . . . . . . 96xiiList of Figures5.4 C.A.I.S. Impromptu spaces of sociability . . . . . . . . . . . . . . 975.5 C.A.I.S. Ticketing prototype . . . . . . . . . . . . . . . . . . . . 1015.6 C.A.I.S. Individual-centric visualization . . . . . . . . . . . . . . 1025.7 C.A.I.S. Community-centric visualization . . . . . . . . . . . . . 1035.8 C.A.I.S. User training . . . . . . . . . . . . . . . . . . . . . . . . 1045.9 C.A.I.S. Prototype installed . . . . . . . . . . . . . . . . . . . . . 1075.10 C.A.I.S. Tasks submitted and resolved . . . . . . . . . . . . . . . 1095.11 C.A.I.S. Post-deployment questionnaire results . . . . . . . . . . 113xiiiAcknowledgementsThis work would not have been possible without the infinite support from manypeople. I owe my deepest gratitude to my supervisors Dr. Sidney Fels and ArchitectOliver Neumann. Their guidance into the world of Human Computer Interactionand Third Places have been essential to this work.I am deeply grateful to Dr. Kellogg S. Booth and Architect Ronald Kellettfor their immense support and endless insight on the research presented in thisdissertation.I am also grateful to those that, in one way or another, contributed to the re-search; in particular Dr. Rodger Lea, Dr. Michael Blackstock and Dr. Junia Ana-cleto.xivChapter 1IntroductionThird Places are defined by Oldenburg [15] as physical locations between thehome (first place) and work (second place) that provide a welcoming and com-fortable atmosphere where people can socialize. Examples of such places includecoffee shops, bars, bowling alleys and beauty parlours. Third Places allow peoplefrom different walks of life to come together, and through conversation, create andmaintain social links that are essential to community building. As Putnam [16]argues, this allows people to build trust in their communities and create strongercommunities. Thus, Third Places are essential to achieving many social, culturaland political goals in society [16, 17]. The concept of Third Place has its roots inthe concept of “public sphere”, defined by Habermas as a space where differentopinions come together precisely because they are different [17]. Oldenburg foundthat some places promote an enjoyable “public sphere”. These places are charac-terized by having a low profile, socio-political neutrality, inclusiveness, accessibil-ity and accommodating environments that promote playfulness and conversationamong regulars.Today, information and communication technologies, which are woven into thefabric of every day life [18], have increased the roles that information has in howwe socialize and how we experience places around us. This has had an effect onhow we experience the “public” and Third Places. For example, today’s coffeeshop – a prototypical Third Place in Oldenburg’s view – leverages small tablesto accommodate private media consumption, power outlets to control seating dy-namics, and complementary and accessible wireless to attract customers. Visitedby telecommuters, coffee enthusiasts, neighbours, students, or on-the-go visitors,modern coffee houses play the roles of Internet-hub, meeting place, coffee dis-tributor and study room simultaneously. Inside these places conversation is eitheraugmented or replaced by laptop computers, tablets, mobile phones and situated1Chapter 1. Introductiondisplays. Yet, Oldenburg’s concept of Third Place does not mention many of theroles that digital technologies play on its contemporary function. This is becausethe concept was coined in the late 1980s and was inspired by physical places fromaround the world, such as the American tavern, the English public house or theFrench cafe´, that were often romanticized in American culture. At the time, in-formation and communication technologies had not yet had a strong influence insociety. Today, however, information and communication technologies have af-fected how we experience the home and the workplace, how we work and, morerecently, how we socialize with each other. Our contemporary experience of be-ing in a Third Place constantly bridges between physical and virtual space, andbetween synchronous and asynchronous communication modalities.Over the last decade, research in the field of Human Computer Interactionhas investigated how technology could support and enhance the function of ThirdPlaces. Researchers have recognized the importance of place [19] and Third Places[20–22] in designing technology for public spaces. They have also looked at how toleverage interactive technology to promote sociability [23–25] or to support com-munity functioning [22, 26, 27]. Today, this research has become even more rel-evant, as information and communication technologies have become a prominentpart of, and are often used to complement, Third Places. As computing becomesmore accessible and telecommunication networks more pervasive, the importanceof the roles of information and communication technologies in shaping the wayswe socialize, and the places in which we do so, will only increase.In this dissertation we articulate an updated concept of Third-placeness thatreflects how the Third Place functions in today’s world, as well as some of theroles that digital technologies (particularly interactive public displays) can play inhow we socialize in such places. We argue that the experience of being in a ThirdPlace can be defined in terms of behavioural characteristics of interpersonal inter-action leading to social capital, aspects of informational governance that promoteequality and accessibility, and physical characteristics that promote regularity andlightheartedness.We call the collection of these qualities “Third-placeness” and posit that bytaking into account the experience of being in a Third Place in terms of informa-tion characteristics we can better design technologies to support the generation of21.1. Research Goalssuch an experience. To corroborate this approach we conducted two observationalstudies using interactive public displays as tools to capture how different char-acteristics of information technologies could support different aspects of Third-placeness. Specifically, we introduce interactive public displays in two differentsettings where Third-placeness could arise: a community of student volunteers ata coffee house and a group of professionals at a workplace.1.1 Research GoalsThe overarching goal of this work is to document how technology could sup-port the experience of Third-placeness. It is driven by the thesis that informa-tion technologies could support communication and socialization practices that,in turn, support social, public and physical aspects of Third Places – collectivelyreferred to as “Third-placeness”. To achieve this, we first develop the conceptof “Third-placeness” to reflect the characteristics of the contemporary experienceof being in a Third Place, in which we posit that technology plays a key role.We develop the concept of Third-placeness by reviewing the literature on ThirdPlaces and on social capital. We organize different aspects of the concept of Third-placeness into a conceptual framework and use this framework to distill currentknowledge on interactive public displays, a technology which is commonly foundin Third Places, and pinpoint characteristics of said technologies that could be usedto support or encourage aspects of Third-placeness. We then use this frameworkto design and evaluate technology prototypes for two observational studies: oneto observe the feasibility of supporting Third-placeness behaviours, and anotherto measure how a single aspect of Third-placeness is affected by one aspect ofinformation technology.Goal 1. Develop an updated concept of Third Place, “Third-placeness”, re-flecting contemporary Third Places and technologies. We review the literatureon Third Places and social capital and define the concept of “Third-placeness”.Third-placeness takes into account the increasing roles that technologies play insocialization and the functions of contemporary Third Places. We review knowl-edge on an information technology commonly found in Third Places, interactive31.2. Research Approachpublic displays, to identify important characteristics of information technologiesthat could support or encourage key aspects of Third-placeness. We distill thisknowledge ito a conceptual framework – the Sociality, Publicity and PhysicalityFramework – that organizes aspects of Third-placeness and their possible relation-ship to characteristics of information technologies, particularly interactive publicdisplays.Goal 2. Document what characteristics of interactive public displays couldsupport or encourage aspects of “Third-placeness” in real world scenarios.We apply the proposed concept of Third-placeness and the Sociality, Publicity andPhysicality Framework to design and evaluate technology prototypes into two dif-ferent communities. We develop a social gardening game for a community of stu-dents running a non-profit coffee shop and a task ticketing system for a communityof professionals at a workplace. We document how the introduction of these tech-nologies affects behaviours that could support different aspects of Third-placenessin each community.1.2 Research ApproachThe research goals are met through observational studies that allow us to gaininsight into how aspects of Third-placeness could be supported by different charac-teristics of information and communication technologies. To support our approachwe examine current literature on Third Places and social capital in order to in-dentify the key aspects of Third-placeness. We then review current knowledge ofinteractive public displays and related research to identify characteristics of saidtechnologies that could support aspects of Third-placeness. Informed by the fieldof design cognition we categorize the identified aspects of Third-placeness andcharacteristics of technology that could support them in a conceptual framework –the Sociality, Publicity and Physicality Framework.We use this framework to guide the design and evaluation of two prototypes inorder to document how some characteristics of information technologies could sup-port Third-placeness. One study examines the feasibility of supporting behavioursthat can lead to the experience of Third-placeness. The other study focuses on41.2. Research Approachmeasuring how a particular aspect of technology – how information is presented– can directly affect Third-placeness. We use ethnography, social capital, and so-cial network analysis as measuring tools. Each study serves different purposes inour research. The study presented in Chapter 4 allowed us to capture a snapshotof a prototypical “Third Place” in order to identify how the introduction of tech-nology affects behaviours that could support different aspects of Third-placeness.The study in Chapter 5, situated in a workplace, provided us with a more con-trolled environment within a well-documented community, in which to test howone particular aspect of technology (information visualization) could affect Social-ity (Social Capital type of reciprocity) and Publicity (ownership) and Physicality,all aspects of Third-placeness.1.2.1 Literature ReviewWe review existing literature on interactive public displays, Third Places andthe concept of social capital. The literature review of interactive public displayslooks at physical and spatial aspects of displays and their relationship to both peo-ple and places, aspects of content such as privacy and accessibility, and aspectsrelated to how people author and use content on interactive public displays. Thereview on Third Places gathers knowledge on what defines a “place” and what gov-ernance, location and ambience characteristics make it a “Third Place”. We reviewthe literature on the concept of social capital, to understand the norms and typesof social connections that bind communities together. Finally, we look at relatedresearch in the field of Human Computer Interaction that has looked at using tech-nology to support Third Place communities; specifically, we focus on interactivepublic display research. The literature review can be found in Chapter 21.2.2 Development of the Concept of Third-PlacenessBased on this review, we revisit Oldenburg’s definition of “Third Place” toreflect the growing roles that contemporary information and communication tech-nologies have in how we communicate, socialize and experience a Third Place. Weposit that the experience of being in a Third Place can be detached from architec-tural and geographical constraints. By analyzing characteristics of contemporary51.2. Research ApproachThird Places and technologies commonly found in such places we articulate theconcept of “Third-placeness”.We organize aspects of Third-placeness in three domains: “sociality”, “pub-licity” and “physicality”. The sociality dimension gathers aspects of interpersonalsocialization through the concept of social capital, particularly reciprocity, densityand centrality. The publicity dimension gathers aspects of governance, particu-larly information ownership, accessibility and privacy. The physicality dimensiongathers aspects that relate to physical aspects of technology and places.We revisit the literature review on interactive public displays to gather charac-teristics of interactive public displays that can support aspects of Third-placeness.We compile key characteristics as a list that can help the design of interactive publicdisplays for Third-placeness. We call this the Sociality, Publicity and PhysicalityFramework. A discussion of third-placeness and the framework can be found inChapter 31.2.3 Supporting Aspects of Third-placeness with TechnologyWe designed a social network and table-top gardening game for a communityof student volunteers at a coffee shop to observe if technology could support be-haviors that can support aspects of Third-placeness. We also design a task ticketingsystem for a community of professionals at a workplace to measure in more detailhow differences in information visualization affect reciprocity within said com-munity. For sociality aspects we measure interpersonal socialization, particularlyreciprocity, and use social capital and social network analysis as methodologies.For publicity aspects we measure governance aspects of equality, accessibility andownership and use questionnaires and ethnographic observations. Regarding phys-icality, we document how perceptual qualities of technology affect adoption – likeregularity and ligheartedness – through ethnographic observations.Our studies suggest that different characteristics of technology encourage dif-ferent structures of communication, or support different social behaviors that canhave an impact on how people communicate with each other. We observed, throughsocial network analysis, questionnaires, participant interviews and ethnographicobservations that these differences could affect on how people experience sociabil-61.3. Summary of Contributionsity. For example, the way in which they maintain interpersonal relationships, takeownership of information shared within a community, or approach newcomers toa group. We found that these variations of the experience of sociability had thepotential to lead to different experiences of Third-placeness. Full details of thesestudies can be found in Chapter 4 and Chapter 5.1.3 Summary of ContributionsContribution 1: An updated view of “Third Place” relevant to contemporaryplaces and technologies called third-placeness. We revisit Oldenburg’s defini-tion of “Third Place” and articulate an updated concept called “third-placeness”.Third-placeness is independent of physical and temporal constraints and emergesas a quality of interpersonal interaction. It is defined by socialization practices thatsupport social capital, aspects of information governance that promote equalityand accessibility, and physical characteristics that promote regularity and light-heartedness. Third-placeness is independent of the architectural or geographicallimitations originally proposed by Oldenburg, which allows the experience of be-ing Third Place to arise anywhere that socialization occurs. The need to revisit theconcept of Third Place has been discussed at several workshops [1, 2] and yieldeda collaboration journal paper [3].Contribution 2: Observational studies of the impact of interactive public dis-plays on different aspects of third-placeness. We designed and deployed asocial game at a student-run coffee shop [5–9] and a task ticketing system at aworkplace [10–12]. We used ethnography and social network analysis to measurechanges in communication, third-placeness and social capital. We also documentedhow people adopt and adapt the technologies introduced into their communities.Contribution 3: Articulation of key characteristics of interactive public dis-plays to support third-placeness resulting in the Sociality, Publicity, Physical-ity Framework. We synthesize key aspects of third-placeness and characteristicsof interactive public displays that can support said aspects of third-placeness. Wedevelop a categorized list of characteristics and a visual diagram outlining design71.4. Outlinedimensions, design ontologies and characteristics of interactive public displays thatcan help designers approach the design of interactive public displays to supportthird-placeness.1.4 OutlineWe begin by reviewing related literature on interactive public displays, ThirdPlace and social capital in Chapter 2. This same chapter presents a review of relatedwork on technologies for Third Place, social capital and network analysis in thefield of Human Computer Interaction.The following chapter (Chapter 3) presents the development of the concept of“third-placeness” and the “Sociality, Publicity and Physicality Framework” con-sisting of a list and a diagram of categorized characteristics important to designinginteractive public displays for third-placeness.We then present two observational studies. Chapter 4 presents the study of in-troducing an interactive table-top garden and private social network into a student-run coffee shop. We observe whether or not behaviours promoting different as-pects of third-placeness can be promoted in a prototypical Third Place. Chapter5 presents the introduction of a task ticketing system designed to support a singleaspect of third-placeness in a workplace. In this study we focus on measuring howcontent visualization affects interpersonal reciprocation within the community. Itis important to mention here that the presented studies were originally conducted inreverse order. However, we present the coffee shop study first as it helps the readerunderstand how technology affects many aspects of Third Places, setting the scenefor the more focused discussion on measuring reciprocity in the workplace study.Chapter 6 summarizes our work, provides details on the contributions of thisresearch and provides initial design guidelines. Finally, taxonomies used to catego-rize the literature review can be found in Appendix A – Background Review Tax-onomies. Surveys and interviews used in this research can be found in Appendix B– Surveys and Interviews. Ethnography and social network analysis methodologiesused in both studies can be found in Appendix C – Research Methodologies8Chapter 2BackgroundThere is growing interest from the Human Computer Interaction (HCI) commu-nity in studying the role of “place” [19] and “Third Place” [20–22] when designingtechnology for communities. Previous research has looked at how to promote so-ciability [23–25] and how to support community functioning [22, 26, 27] leading toa growing body of work on designing interactive public displays for Third Places.This chapter reviews the literature that is common to this area of research. We re-view the literature on interactive public displays to understand how to design saidtechnologies for the context of Third Places. We review what makes Third Placeswhat they are and from where the concept originates. We also review the conceptof social capital which we will use to measure changes in community functioning.Finally, we review related HCI work on using interactive public displays to supportThird Places and the use of social capital and network analysis.2.1 Third Places and their CommunitiesSome researchers, including Oldenburg [15, 28], Putnam [16], and Habermas[17], argue that it is in places where conversation occurs that human societies areshaped and woven together. Oldenburg calls these places “Third Places” to sep-arate them from dwelling and working places. Third Places, like pubs, coffeehouses or book stores, allow people to come together and exchange informationthrough conversation. Putnam argues that this allows people to build trust in theirlocal and national social networks, leading to better communities. According toOldenburg, “Third Places” are characterized by a low profile, social and politicalneutrality, inclusiveness, accessibility, being accommodating, conversation, and aplayful ambience that promotes returning “regulars”. Three aspects are recurrentin the literature regarding communities in Third Places: the concept of “place”,92.1. Third Places and their Communitiesthe concept of “Third Place” itself, and tools to understand community function-ing, like “social capital”. The following subsections will introduce each of theseconcepts.The first subsection, “Aspects of Place”, introduces the concept of “place” ascentral to the human experience, specifically in society. It presents the concepts oflocation, ensemble and insideness that are key to describing and categorizing thedifferent places of our lives.The second subsection, “Aspects of Third Place”, introduces the concept of“Third Place” and presents a brief history of its origins in 17th century coffeehouses. We categorize characteristics of Third Places through the concepts of gov-ernance, location and appearance, and ambience.The third subsection, “Aspects of Social Capital”, discusses an approach tounderstanding community functioning by adding value to social interactions andsocial links. We discuss differences in types of social connections, types of socialcapital, and types of capital reach and gain.A visual guide to each subsection and the concepts discussed can be found intaxonomies presented in Appendix A – Background Review Taxonomies.2.1.1 Aspects of PlacesThe concept of “place” is the intersection of four types of space: pragmatic,perceptual, existential and architectural [29]. Pragmatic space is the primitive andinstinctive (without reflection) interaction with space. Perceptual space is the ego-centric experience of space confronted by each individual. Existential space is theinner structure of space formed by experiencing the world as members of a group.Architectural space is the creation of space through thought, and planning, fol-lowing values (theory and style) and restrictions (building codes). According toLukermann [30] places are human and cultural experiences defined by a uniquecombination of pragmatic, perceptual, existential and architectural space in con-stant re-creation by the groups that experience it. Characteristics that are key tothese four spaces are their physical setting (location), the group of characteristicsand human activities that are performed within such a location (ensemble), and thesense of sharing such space with other people (insideness).102.1. Third Places and their CommunitiesLocation. The “location” of a place can be described in terms of internal arrange-ment of features (site) and external connectivity (situation) [30]. While places haveboth site and situation they are often skewed towards one or the other. Places withstrong external location serve as connectors, while places with internal location areoften isolated or have rich internal arrangements that serve as boundaries.Ensemble. Each place is a unique combination of physical characteristics, ac-tivities and character that renders them unique [30, 31]. Physical characteristicsinclude natural aspects like topography, weather and fauna, and man made ele-ments like constructions and their spatial configurations. Human activities includeroutines or events that people conduct in a place. Character, or also called “geniusloci”, is formed of elements that compose the ambience, atmosphere and percep-tion of a place in a culture. According to Seamon [31], character is what givesplaces their sacredness and humanity.Insideness. The meaning of places in our lives is determined by how much webelong to them. In other words, how much we are “insiders” of them [29, 32].Through various grades of insideness, places around us take on different meanings[33]. There are four types of insideness. —“Existential insideness” allows the ex-perience of a place’s significance without a deliberate and self conscious reflection.“Behavioural insideness” occurs when we deliberately attend to the qualities andthe activities of a place, for example when becoming familiar to a new place. “Em-pathetic insideness” is a deliberate attempt of an outsider to be open to a place inorder to understand it. “Vicarious insideness” is second-hand involvement with aplace through imagination, for example in paintings, films or music.2.1.2 Aspects of Third PlacesOldenburg [15] defines Third Place as the space in-between the first place(home), which dedicated to the experience of private live, and the second place(work), which embodies productivity and economic purpose. While often an op-tion for the gregarious and adventuresome [15], Third Places allow many othertypes of people to meet and, through conversation, build social links essential to112.1. Third Places and their Communitiessocial, cultural and political goals [16, 17]. Any physical place that has the ingredi-ents to promote social interaction can be a Third Place. These ingredients are: a lowprofile, social and political neutrality, inclusiveness, accessibility, being accommo-dating, conversation, a playful ambience and regulars. These could be found at aparlour, a bowling alley, a book store or coffee house. Similar concepts are pro-posed in the literature. Bhabha proposes “third space” as a fluid space, formed asa result of dialogue and contradiction between interpersonal interaction of socialand political groups [34]. Soja proposes the concept of “Thirdspace” as an alterna-tive to the dualism between home and work. In Soja’s definition Thirdspace takesinto account place, location, locality, landscape, environment, home, city, region,territory and geography [35] to shape a fluent experience of space and cultures.Of the concepts discussed, Oldenburg’s definition of third place is the most widelyaccepted and is a growing concept in the field of Human Computer Interaction[22, 26, 36, 37].The need for a “Third Place” in between the privacy of the home and the public-ity of work has its origin in the evolution of the concept of “public”. In particular,it is born from the institution of the 17th century “coffee house”, which had a largeimpact on the development of Western political, financial, scientific and literarythinking [38]. Coffee houses were introduced to the Western world in 1650 in Ox-ford, England [39]. At that time there was no daily publications, so people gatheredin these places to share and gather information. It was the homeliness and rusticnature [40] of these coffee houses that attracted visitors from all walks of life. Infact, Ellis reminds us that one penny was all that was needed by a man, rich orpoor, to gain entry to a world where the Government’s Gazette would be read outloud in search of a democratic assembly [39].At the end of the 17th century the democratic coffee house made way for theclub and the privately-owned coffee house. Although exclusive, the private cof-fee houses of the 18th century served as key backstage players to the developmentof the concept of public [41]. They became information centres where the edu-cated bourgeoisie gathered to sketch the concept of a “democratic public” throughconversation and periodicals that were either written or recited out loud.A romanticized version of these coffee houses was first introduced in NorthAmerica during the 1920s [42] and grew in adoption during the 1950s. During122.1. Third Places and their Communitiesthe 1990s, market saturation forced coffee-house owners to focus on providingspecialty spaces that would cater to different niches [42, 43]. This led to a wide se-lection of urban coffee houses, highly accessible, offering low prices and a regularcliente`le. It was these coffee houses that attempted to replicate 17th century coffeehouses and focused on providing spaces for conversation [37, 44] in cities whereSuburban developments separated living from working quarters. During this timeOldenburg coined and defined the term “Third Place”. The key characteristics of“Third Places” can be organized into three categories: governance, location andappearance, and ambience.Figure 2.1: Illustration of a Parisian street with many types of social public spaces.Governance. Third Places are defined as a neutral ground where different groupsof people, with similar or opposing views, come together to socialize. Such neu-132.1. Third Places and their Communitiestrality is born of the fact that these places are not “owned” by a particular groupnor tied to a particular authority. Places with high neutrality allow anybody to feellike an insider (Figure 2.1-A). This is not the case in low neutrality places, whichare owned by a particular group or require membership (Figure 2.1-B). High inclu-siveness is found in places that avoid implying exclusion through the use of designelements such as open windows, terraces, and tables that pour onto public space(Figure 2.1-C). Places with low inclusiveness limit who can have access, throughhigh prices, requirements, dress codes, and architectural design to communicate aseparation from public space (Figure 2.1-B). Another related aspect of governanceis “inclusiveness”, which means allowing people from different groups to bridgeacross their differences. Places that have a high level of inclusiveness may also bedescribed as having insideness, or “rootedness”. Places with high rootedness allowtheir visitors to appropriate the place easily, allowing them to use it as a pivot pointfrom which to organize comings and goings. Individuals have the unconscioussense that such a place is un-mediated [45], or that it can serve as a secure pointfrom which to look out on the world [29].Location and appearance. Location and appearance are the physical qualitiesof Third Places. Accessibility and inconspicuousness are important location andappearance qualities that make a place welcoming and inviting. Accessible placesare located close to work or home and are always ready to serve people’s need forsociability and relaxation [15]. Places with high accessibility seem to be alwaysopen and are often found at midpoints between work and home, or close to publicamenities. Places with low accessibility are more difficult to reach or endpointsinstead of midpoints between important places. In terms of appearance, a veryimportant aspect of Third Places is their low profile appearance, characterized bymodesty and plainness. A low profile discourages pretentiousness and creates amore inclusive and neutral place that people feel inclined to appropriate. Highinconspicuousness is often promoted by vernacular or plain design and low costdecor that announce the affordability of the place [15].Ambience. The “ambience” of a Third Place is characterized by familiarity, play-fulness and conversation [15, 41]. Familiarity, or the existence of ‘regulars’ that142.1. Third Places and their Communitiesmake people feel at home, can only be beneficial if it is accompanied by trustwor-thiness [39] and the acceptance of new faces. Places with high familiarity encour-age regulars to welcome newcomers and are permeated by a sense of trust. Play-fulness is essential to casual public interactions, allowing a comfortable distancebetween individuals [41]. Places with high playfulness promote chatter, loudness,laughter and spontaneity (Figure 2.1-E) where people can freely engage in socialplay. Places with low playfulness often limit loud patrons, chatter, laughter andspontaneity (Figure 2.1-F).Conversation. Finally, for Oldenburg, the main activity of Third Places is con-versation. It is through conversation that people can fully utilize the governance,physical qualities and ambience of Third Places. Conversation allows for informa-tion to flow between people and allows the creation of social links.2.1.3 Aspects of Social CapitalThird Places host and support strong communities [15] because they providea place where people can build and nurture the social links that tie communitiestogether [16]. A sense of a community’s strength can be captured by measuringits “social capital”. This concept originated between the 19th and the mid 20thcenturies [46], when nations and corporations began applying new methods to in-crease productivity, reduce uncertainties and counteract the increasing power oflabour unions. It was posited then that social cohesion between groups of peoplecould be used as a value against less organized competitors. Hanifan [47] coinedthe term social capital in the early 1900s to describe the accumulation of tangiblesocial assets (good will, fellowship, sympathy, and social intercourse) that mem-bers of a community can use to enhance their lives. The concept was more deeplyadopted by new-liberalism during the mid 20th century to capture the economicvalue of social relationships [48]. Aspects of social capital in the literature includecapital type, connection type, capital reach and capital gain.Capital type. Researchers [16, 46, 48] argue that the most important types ofsocial capital when studying communities are “trust” and “reciprocity”. Trust is152.1. Third Places and their Communitiesthe belief in other people’s reliability, truthfulness or ability, that allows peoplenot to confirm every social transaction when socializing. Trustworthiness allowsgroups to accomplish more than a comparable group without trust [48] becausepeople do not have to watch each other’s back. In a similar fashion, strong socialnetworks often involve mutual obligations and strong norms of ‘reciprocity’ [16].Reciprocity can be understood as “favour debt slips” payable in the future eitherby the receiver of the favour or the community. Reciprocity encourages people todo things for each other in the expectation that someone will return the favour inthe future.Connection type. Of all the dimensions across which social capital can vary, themost important distinction to make is that of connection type, which can be “bridg-ing” or “bonding” [16]. Some communities reinforce exclusive identities and ho-mogeneous groups (bonding links), for example fraternal organizations. Othergroups include different people from different social groups (bridging links), forexample youth service groups. The decision about whether a community oper-ates with either bridging or bonding connectivity is sometimes made by choiceand other times out of necessity. Bridging connectivity generates broader iden-tities and reciprocity, while bonding connectivity bolsters self reflection [16]. InThird Places, bridging connectivity allows newcomers to socialize with long termregulars, while bonding connectivity strengthens the existing links between placevisitors increasing rootedness.Capital reach. Reciprocity can be “particular” or “generalized”. Particular reci-procity describes the case of a person doing something for someone in exchangefor something else from that person. Generalized reciprocity describes the caseof a person doing something for someone without expecting anything in return, inthe confident expectation that someone else will do something for him or her inthe future. The generalized reach of social capital makes communities more effi-cient because people are not required to balance every social exchange instantly[16]. When generalized reciprocity reach is embedded in dense networks of socialactivities, opportunism and malfeasance are reduced.162.2. Interactive Public DisplaysCapital gain. Social capital can used to build better schools [47], more produc-tive organizations [48] and more politically engaged societies [16]. Specifically, itcan be used for either personal gains, like obtaining a better job or gaining the trustof a community, or it can be used for collective gain, for example in the case ofmaintaining a community garden where gains can be used by all members of thecommunity.2.2 Interactive Public DisplaysHuman beings are social animals that interact with each other by exchanginginformation. Historically, visual communication has been used because of its ca-pability to withstand the effects of time and its ability to reach groups of people.Displays (devices used to provide sensory manifestations of information [49]) areamong tools most commonly used to achieve this. The origin of the relationshipbetween displays and the concept of “public” can be traced back to early neolithiccave paintings, which required advanced technologies and group manufacturing[50, 51]. Manifestations of displays evolved from wall paintings to carved architec-tural structures, engraved wooden tables [52], and mechanically printed pages thatboth individualized the consumption of information [53] and played an importantpart in creating the contemporary understanding of “public” during the 17th and18th centuries [15, 39]. More recently, large scale printing, like the iconic Amer-ican Billboards [54], 1920s neon lighting [55], the Television during the 1940sand the Personal Computer during the 1980s blurred the line between private andpublic life [56]. New kinds of displays, paired with growing telecommunicationnetworks have become tightly interwoven with daily activities, increasing the rolethat information plays on our lives [57]. We are particularly interested in displaysfound in contemporary Third Places, where conversation is either augmented, orreplaced, by new kinds of displays called “interactive public displays”, such aspersonal computers, tablets or mobile phones. The following subsections discusscurrent knowledge of these kinds of displays and the concepts and characteristicsthat are important when designing them for public usage.The first subsection, “Aspects of Displays”, gathers knowledge of the physical,spatial and perceptual capabilities that characterize different types of displays. It172.2. Interactive Public Displayspresents the concepts of geometry, reach, resolution and perceptual qualities com-monly used when designing such technologies.The second subsection, “Aspects of Publicity” discusses how content is createdand accessed by people through displays, particularly in a public context. It willpresent the concepts of ownership, accessibility and privacy commonly used in theliterature to describe content creation and usage.The third subsection “Aspects of Interactivity” discusses characteristics thatmake public displays interactive, specifically the activities that people perform inorder to create such content and how they relate to the meaning and experience ofusing such displays. This section will present the concepts of information repre-sentation, synchronicity, authorship, embodiment and awareness.The relationship between each subsection and the concepts introduced can begrasped visually through taxonomies presented in Appendix A – Background Re-view Taxonomies.2.2.1 Aspects of DisplaysThe main aspects of displays found in the literature include their geometry,reach, resolution and perceptual capabilities.Geometry. Displays have planar (Figure 2.2-A), curved or hemispheric shape.While planar displays are pervasive, curved and hemispheric displays can providewider viewing angles that can allow for simultaneous consumption of content bymany viewers [58, 59]. Both curved and hemispheric displays can be formed byarranging planar displays together, however this approach can result in lack ofcoherence [60]. Display scale is important and is often referred to as mobile (2” to10”), situated (10” to 30”) or architectural (greater than 30”). Situated displays aremore commonly used in public spaces as they are cheaper and versatile [26], whilearchitectural displays are costly and difficult to install [61]. Research shows thatvertical displays (Figure 2.2-A) entice public sharing and consumption of content[22, 25, 26], while horizontal displays on tables or floors tend to encourage morecommunal interaction with content [23]. Slanted displays have existed in kiosksfor over 20 years [62] and can be considered more common today due to personal182.2. Interactive Public DisplaysFigure 2.2: Illustration of various types of displays that could be found intechnology-prone public places: (A) wall-mounted displays, (B) kiosks, (C) mo-bile phones and (D) wall-sized displays.tablets and phones.Reach. Displays can be found at body reach (at arm’s length), within body reach(closer than arm’s length; Figure 2.2-C) and at place reach (out of arm’s reach; Fig-ure 2.2-A). While displays at place reach allow for public consumption of content,within-body and at-body reach allow for increasing levels of private engagementwith content [63–65].Resolution. There is a proven relationship between spatial resolution and vi-sual fatigue [66] that is affected by the distance at which the display is observed[60], particularly in public space. Up-close interaction with displays often requireshigher resolution to avoid eye strain(90 dots per inch and greater [66]), when beingobserved at a near-perpendicular angle (Figure 2.2-C). At longer distances (start-ing at arm’s length, or 20 to 40 inches [60]) displays can have lower resolutionand maintain content coherence by making content such as text larger at the ex-pense of privacy loss (Figure 2.2-D). For temporal resolution, higher frame refreshrates (>60fps) allow for movement to be perceived as smoother and clearer (Fig-192.2. Interactive Public Displaysure 2.2-C), but lower frame rates are often sufficient for displays placed at greaterdistances (Figure 2.2-A) [60]. Similarly, response times that are lower than 8 mil-liseconds provide quick transition between content and higher perceived quality invideo [67]. As people move away from the display this correlation seems to bemore flexible, and lower resolutions can be used.Perceptual capabilities. Three concepts are key to perceptual capabilities of dis-plays: contrast, brightness, and viewing angle. Brightness is often an importantfactor in public and open spaces, as daylight and artificial light can have a detri-mental effect on display perception [61]. Situated displays often have medium(approximately 120 cd/m2) brightness capabilities, while larger architectural dis-plays use higher brightness levels (approximately 1000 cd/m2) [68] to account forambient lighting and wider viewing angles. In contemporary displays clarity at dif-fering viewing angles is affected by a combination of manufacturing process, thematerials of the display and display brightness. Situated and architectural displays(Figure 2.2-A, Figure 2.2-D) must be designed taking into account wide viewingangles as content is often viewed from different locations. Displays closer to ob-servers can sacrifice this as they are often used by just a few people and held inoptimum position (Figure 2.2-C).2.2.2 Aspects of Public DisplaysFor Hanna Arendt, “public” is the act of making one’s self available to othersby disclosing information to strangers [69]. The separation of public and privatehas its roots in ancient Greece and the distinction between activities that relatedto the maintenance of life (or the home) and those common to the group to whichone belongs (work, among others). Yet, this ancient public sphere was exclusiveto those in command of the household. Encounters between the humanistic aris-tocratic society and the bourgeois intellectuals in places like the coffee houses,salons and tischgesellshaften (table societies) during the 18th century began brew-ing a new concept of “public” [17] defined by the disregard of status, a domainof common concern and inclusiveness. During the 18th century, “being in public”meant the act of people becoming “individuals” by disclosing themselves to others202.2. Interactive Public Displaysthrough conversation. This concept of public was originally only part of the elite ofthe 18th century, until the French revolution of 1789, when the state began policingall aspects of civic life in an attempt to control the rising bourgeoisie [70]. This“omnivoyance” made publicity an inescapable aspect of civic life.Displays have played an important role in the development of the concept ofpublic. During the 18th century, press-printed periodicals were essential in fuellingdiscussion [39], while hand-out bills and billboards were used in state propagandaduring the 1930s and 1940s [71]. Television [72, 73] reshaped the boundaries be-tween public and private in both public (saloons, bars) and domestic (the livingroom [74–76]) places. Contemporary displays, such as mobile displays, both cen-tralize and privatize information [77–80], while situated public displays affect howinformation is compiled and distributed in public space [81, 82]. Furthermore,telecommunication technologies allow for simultaneous and distributed accessibil-ity [24–26, 63] of content. In the literature of such displays, three concepts arekey to understanding this relationship between information and the public sphere:ownership, accessibility and privacy.Figure 2.3: Illustration of different types of content one could find in a public space:(A) Photographs, (B) abstract representations, (C) video content, or (D) text.212.2. Interactive Public DisplaysOwnership. The literature categorizes content as being owned individually orcollectively [83]. Individually owned content has become one of the most preva-lent types of content in modern displays [84]. With the increase of telecommunica-tion networks in public spaces, the practice of sharing individual content has beencarried into public space. Public displays often serve as vehicles for this purpose[26, 85] (Figure 2.3-A). Collectively owned content [86], or content produced aspart of a collective experience [86, 87] (Figure 2.3-B), has been previously investi-gated as a means to encourage social interaction in public space [87], often leadingto inter-personal interaction and a sense of community [36]. This occurs becausecollective content creation often leads to a perception of collective ownership [88].Accessibility. The term “accessibility” is often used to to describe the level towhich an individual can access and modify content and displays themselves. Openaccessibility allows content to be accessed by anyone, and information is disclosedin its entirety (Figure 2.3-B). Filtered accessibility, by contrast, allows only a se-lected group of people to access the display or content. This can be either a groupof people or a single individual (Figure 2.3-C). Previous research argues that acces-sibility can be promoted through display size [89], positioning of display, contentsize [63], and how such content can be manipulated. While gestural interactionoften promotes open accessibility, touch-enabled displays allow people to interactwith content in a private or semi-private manner [63]. Large-scale displays oftenpromote open accessibility, while body-scale displays promote filtered accessibil-ity of content [90] (Figure 2.3-C).Privacy. Regarding “privacy” in public displays, Palen et al. [91] define it as aself-regulatory human act in a spectrum between openness and closeness of con-tent. Privacy is important because it helps prevent the disclosure of intimate andsensitive information to a public sphere that is, today, considered tightly related tosurveillance and control [70, 92, 93] within Western cultures. Thus, privacy is not astatic enforcement of global rules, but is shaped by trying to find a balance betweenbeing “in public” and maintaining human status as “individuals” when in public.Public displays serve as mediators for information disclosure by associating con-tent to different types of identities. The use of a truthful identity allows people to222.2. Interactive Public Displaysdisclose their true self by means of their real name, age, gender, and sometimesgeographic location. Research has shown that associating content with individu-als makes information more attractive [94]. For example, by using Online SocialServices to leverage people’s identities to promote inter-personal interaction [95].In reality, however, people often use masked representations of themselves, likenicknames and avatars, that allow for a more careful exposure of their identity inpublic spaces [96] (Figure 2.3-D). An even further step into identity shielding is theuse of anonymous or abstracted visualizations (Figure 2.3-B). Abstract visualiza-tion of content has proven effective at providing useful information to communitieswithout compromising private information [97].2.2.3 Aspects of Interactivity in Public DisplaysThe technical definition of “interactivity” is a timely control over the com-puter process placed in the hands of the user through rapid computation [98, 99].However, a more contemporary understanding of “interactivity” is that both parties(human and machine) respond to each other in a constant flow of bi-directionalmessages. Scholars argue that this informational exchange can be considered “lin-guistic”, because it requires understanding each piece of information in terms of allmessages previously received [99]. Moreover, this linguistic nature can sometimeappear as intentional and social, because the inner workings of the computer arenot evident to humans [98].Interactive public displays have their origin in in public kiosks created in the1970s, through the pairing of computers with Television sets [100]. Newer displaymanufacturing technologies, like liquid crystals and light emitting diodes, alloweddisplays with less weight to be hung on walls, and placed in floors and tablesduring the 1990s [101]. More recently, technologies like touch tracking and bodymotion tracking have made interacting with public displays more pervasive [102–104] and have further increased our sense that we engage in dialogue with displays.Important characteristics of such interaction found in the literature include contentrepresentation, synchrony, authorship amplification, embodiment and awareness.232.2. Interactive Public DisplaysFigure 2.4: Displays often found in Third Places can be (G) paper-based displays,(J) situated digital displays and (D) mobile displays. These displays can contain(A) iconic information, like images; (B) symbolic representations, for example a“like” Facebook button or (H) a digital calendar; (C) or indexical or abstract infor-mation, like (I) circles to represent a community. People often use these devicesthrough (D) interactions involving mobile devices connected to situated displays,(E) gesture-based interactions with situated displays, (F) interactions with touch-enabled displays, (K) private interactions involving mobile devices, or (L) periph-eral or passive interactions with displays.Representation. The theory of signs (entities representing or signifying thingsto the observer [105]) and can be a helpful tool in understanding the meaningscontent can take. There are different types of content. “Iconic content” exhibitsthe quality of the object it signifies [105, 106], like photographs and videos, andis often used in interactive public displays [26]. “Symbolic content” is associatedwith its signified object by conventional use, like the character @ to representa relationship and a “thumbs-up” image to represent approval. The meaning ofsymbolic content evolves as conventions evolve [107]. “Indexical content” dependson its context, as its meaning depends on the situation in which it is found, likevisual content paired with unrelated text [108], content modified by other people242.2. Interactive Public Displays[22] and abstract representations [97].Synchrony. Synchronous interaction supports instant consumption [109] of in-formation and activities. Previous research [36, 110] finds that synchronous inter-action often leads to ‘performances’ (Figure 2.4-F) that capture a watching public.However, synchronous interactions are rare due to embarrassment and often smallaudiences [85]. Asynchronous interaction significantly increases the span of timeduring which actions performed by a person with a public display (Figure 2.4-D)may be observed by other people [85] (Figure 2.4-E). Asynchronous interactioncan reach broader audiences [22], promote content collaboration over long periodsof time [108], promote information permanence [111] and allow temporal sharing[94].Authorship amplification. Observing others interact with displays is intrinsic tothe experience of interactive public displays [112]. Because being in “public” isultimately revealing ourselves to others [17] we can understand authorship throughthe concept of amplification. “Hidden authorship amplification” separates actionsfrom content, making the authors of such content hard to discern, for examplein multi-display interactions (Figure 2.4-D). “Revealed authorship amplification”occurs when observers are aware of who is willingly interacting with the contentbut the details of such interaction are not made the focal point of attention (Fig-ure 2.4-E). When authorship is amplified the actions themselves are made the fo-cus of attention, for example when displays are arranged as a theatre [113], intouch-enabled situated displays (Figure 2.4-F) or when using body gesture inter-faces [114]. While some people enjoy amplified authorship, research quotes pub-lic embarrassment as a common barrier to these kind of authorship amplifications[111].Embodiment. When people immerse themselves in conversation they ultimatelyembody information conveyed and received [115]. Interaction with public displays,understood as a dialogue between human and machine, is no exception. Fels [116]proposes four types of embodiment: response, control, reflection and belonging.Most contemporary interactive public displays are designed to “respond” to in-252.3. Related Work on Technology for Third Placesstructions by producing content as required, for example a message board (Figure2.4-G). When a display or its content becomes indispensable to interaction betweenpeople, for example a digital community calendar [111] (Figure 2.4-H), a “control”embodiment occurs, because the display feels like an extension of the people thatuse it. When a display provides ambient information leading to a deep communalunderstanding of the people that use it [97, 117], a “reflection” embodiment arises.Finally, if such information leads to behavioural effects on people [117] it willbecome a “belonging” embodiment, because people are embodied by the display[118, 119] (Figure 2.4-J).Awareness. The literature [65] often refers to three levels of awareness: “direct”(Figure 2.4-F), “focal” (Figure 2.4-K) and “peripheral” (Figure 2.4-L). Finke etal. [64] argue that interaction with public displays implies a transition betweenall three levels of awareness as people flow between the roles of by-stander andactor [120]. As people change from being observers to becoming participants theyperform different activities affecting other aspects of interactive public displays,like authorship amplification and embodiment [108].2.3 Related Work on Technology for Third PlacesThe subsections below present related HCI research on interactive public dis-plays for Third Places, and research that has previously used social capital andnetwork analysis to understand communities.2.3.1 Interactive Public Displays in Third PlacesResearch in HCI has stressed the need to better understand the relationshipbetween technology and community functioning. Previous research has empha-sized the need to design for groups rather than individuals [121], to consider thelong-term social impact of technology and to focus on social engagement [122]when designing for communities. Research has also found that information abouta group’s social structure can be used to increase collaboration and engagement[123, 124].262.3. Related Work on Technology for Third PlacesSpecific to interactive public displays, previous work has documented thatmany displays in public settings are often ignored [61] due to inadequate selectionsof display type, size, reach and positioning. Static content and wrongly selectedprivacy also contribute to people often ignoring public displays. More recently theincreasing adoption of touch-enabled situated displays has provided ways to en-courage interaction with public displays. Research by Brignull et al. [65] proposesa categorization of interactions with displays into the groups of peripheral aware-ness, focal awareness, and direct interaction. Finke et al. [64] argue that whenpeople flow through spaces that contain displays they transition from bystanders,to spectators, to actors, requiring different interaction mechanisms, content typesand privacy gradients [63]. One often-cited drive for interaction with public dis-plays is what Brignull et al. [65] call the “honey-pot effect” that arises when severalpeople interact with a public device and draw attention to it by doing so. Studieslike “City Wall” [125] document this effect and argue that conflict managementthat arises from multi-user interaction with displays plays an important role in pro-moting social behaviours between users. In such cases, where some people interactwith the display while others curiously watch, the act of “performing” in such aninteraction is an important factor [112] that can lead to either fear of public shameor performance enjoyment.Growth in the number of portable devices with the capability to connect towireless networks, particularly laptop computers and mobile phones, has allowednew ways to interact with content in public places. Kaviani et al. [89] note that pri-vate devices can offload some of the cognitive barriers of public displays throughhigher perceptual capabilities and fine-grained privacy levels. When used in con-junction with situated large displays, mobile displays can promote social interac-tion by leveraging global positioning [126], collaborative creation of public content[82], and personal content mined from online social networks [90]. These strate-gies have previously been investigated at an urban scale in the city of Oulu, Finland[90], at an area level with the ActiveCampus project [127], and at different ThirdPlaces [23, 25].Research by Churchill et al. [21, 94] has explored complementing existingcommunication technologies used by communities of a coffee shop (email, onlinechat rooms and community spaces) with large situated displays. The researchers272.3. Related Work on Technology for Third Placesfind that most people are passive consumers of content produced by few membersof the group. They propose designing displays as “neutral spaces” for infrequentuse. Research by Olsson [83] proposes an alternative to promoting engagementwith displays by designing content as a collective contribution. He proposes thattechnology should be motivated by the need to nurture existing social relationships,sharing of common knowledge or individual perspectives on common matters, andcreating group memories. Olsson argues that this differs from the common per-ception of “public content”, which is often motivated by a desire for publicity orpublic recognition.Technology designed to promote interpersonal interaction in Third Places in-cludes Yoon et al.’s ‘FishPong’ [23], Hosio et al.’s ‘Social Surroundings’ [95],Scheible and Ojala’s ‘Mobilenin’ [36] and Ohara’s ‘Jukola’ [110]. FishPong ex-plores casual interaction with virtual fish on a table-top display to promote interper-sonal conversation between strangers. Social Surroundings attempts to relieve theparadigm of ‘isolation-by-choice’ by allowing people to bring parts of their digitalrepresentations from online social networks onto large displays situated in ThirdPlaces, aiming to encourage people to spend more time in such places and interactwith each other face to face. Mobilenin explores deploying an interactive multi-display application in a bar, consisting of a multi-track music video projected on alarge display and several mobile phones. The video features a virtual singer whoseactions (clapping, playing guitar, singing) can be voted on by participants via theirmobile phones. In a real-world deployment Scheible and Ojala learned that jointauthorship can help with appropriation of content and social involvement. Jukolais a similar project that uses situated touch-enabled displays, mobile phone ap-plications and a website to vote on the music being played at a bar. Researcherslearn that voting using displays, both situated and mobile, leads to face-to-face ne-gotiation and device sharing that supports conversation. In Jukola, a networkedconnectivity between devices and feedback on previous selections allowed peopleto make social inferences about co-located people, creating a sense of community.Research investigating the introduction of technologies to Third Places overextended periods of time includes work by Farnham et al. [85] and Churchill etal. [22]. Farnham has argued that people looking to connect with others, searchingfor a sense of community, or previously attached to a place, are more likely to use282.3. Related Work on Technology for Third Placestechnologies in Third Places. Similarly, Churchill has found that creative digitalcontent, like scribbles, serves to complement artifacts in Third Places that promoteserendipity, like suggestions books. Both research groups argue that the abilityof situated displays to cross online and offline socialization boundaries is moreuseful to communities than digitizing previously existing content of Third Places,like menus or newsletters. Research looking specifically at promoting face-to-face interaction includes work by Izadi and Brignull [24, 65, 111] (Dynamo) andTaylor et al. [27] (Wray Photo Display). Dynamo explores co-located sharing ofmedia using a situated large display. Wray Photo Display allows residents of avillage community in England to design and print digital postcards using a situateddisplay. Both studies learn that technology is often adapted by a community as ittakes different meanings for different people.2.3.2 Social Capital and Network Analysis in HCIThird Places are often described [15, 16] as places that can foster social capi-tal, particularly trust and reciprocity, between members of a community. Thus, it isimportant to be able to understand how social capital is affected when technologyis introduced to a community. This subsection reviews the literature of using so-cial capital to understand community functioning and tools, surveys and networkanalysis often used to do so.There is no single definition of social capital [128] and there are many instru-ments used by researchers to quantify it. In the literature [129, 130], social capitaloften relates to understanding how people invest in social relations and manageexpected returns. Often-measured characteristics include civic participation in vol-untary associations, norms of mutual aid and reciprocity, and levels of interper-sonal trust. Many of the existing social capital instruments use the social surveyapproach, with individual responses aggregated. These surveys [131–134] includequestions for commitment and participation in civic groups, management of tieswith friends and strangers over long periods of time and perception of institutions.Lochner [128] points out that such surveys need to be complemented with ‘in-trinsic’ measures that are directly observable features of a community, becausesocial capital depends more on day-to-day interactions between people than on so-292.3. Related Work on Technology for Third Placescial, economic or cultural factors. Western et al.[135] propose a survey aroundintrinsic measures of a community. They categorize social capital as the intersec-tion between norms (normative attributes of a community) and structures (physicalstructure of the community network), focusing on structural measures like size,capacity, openness, homogeneity, and normative measurements like trust and reci-procity. Another approach found in the literature is to focus on the structural as-pects of social capital. Lin [129] argues that social capital can be understood asassets in a network. He posits that one can measure social capital by observingnetwork resources, network structure and types of ties between nodes. This ap-proach provides an initial understanding of a community’s social capital. However,patterns of interactions and social exchanges, socio-economic differences betweenpeople and the effect of globalization are not captured by this methodology.The recent growth of online networking sites has allowed researchers to com-plement survey and socio-economic instruments to measure social capital with datafrom day-to-day interaction between people. Smith [136] proposes using networkanalysis, which is a tool that has been used for decades to understand the struc-ture and functioning of static networks [137]. Network analysis measures include:“centrality”, or importance of nodes; “density”, or the proportion of links relativeto total nodes; “strength”, or the combination of emotional intensity, intimacy andreciprocity; “clustering”, or the likelihood of association between nodes; “reci-procity”, or the extent to which nodes reciprocate each other; and “homophily”, orthe extent of which nodes are similar or dissimilar to each other.From an HCI standpoint researchers have previously used methods of networkanalysis to better understand patterns of usage in online social networks. Burke etal. [138] have compared types of social activities (direct communication, passivecontent consumption, and broadcasting) on Facebook, finding that direct person-to-person exchanges are associated with increases bridging social capital. Jung etal. [139] have also analyzed Facebook usage and found that bridging and bondingsocial capital do not significantly predict whether or not people will respond tofavour requests. They observe that frequently asking for help is often associatedwith a higher number of responses, likely by keeping reciprocal relationships alive.Finally, Liu et al. [140] have proposed that an individual’s personality, social capitaland online social networking data can be used to understand such individual’s real-302.4. Summarylife bridging social capital. They observe that empathy and being conscious ofone’s community influence how people create real life inter-personal relationships.They show how network structure, measured using network analysis, can predicthow people use social networks and utilize social capital in real life. They arguethat pairing network analysis with standardized surveys and usage data can be anefficient way to quantify and predict a group’s social capital.2.4 SummaryDisplays are among the tools most commonly used to exchange information,and vary according to physical, spatial and perceptual characteristics. When inpublic, displays have to take into account ownership, accessibility and privacy inorder to support and enable the act of making one’s self available to others by dis-closing information. Some of these displays can be called interactive when theyallow content to be modified according to its representation, synchrony, authorshipamplification, embodiment and awareness. Interactive public displays are becom-ing more common in places where socialization often occurs. One such type ofplace is what Oldenburg refers to as the “Third Places”, or places that are nei-ther home nor work, where socialization occurs. Third Places are characterizedby aspects of location, ensemble and insideness that define them as ’places’, andthrough the fact that they exhibit aspects of governance, location, appearance andambience that make them low profile, neutral, inclusive, accessible and accommo-dating to conversation and play. A tool often used to understand how communitiesin these places function is the concept of “social capital”. Aspects of social capitalinclude capital type, connection type, reach and type of gains attained by using it.Trust and reciprocity are two of the most often quoted social capital types found inThird Places.In the literature, there is a growing interest within the field of HCI in designingtechnologies for Third Places. The need to understand how technologies can sup-port communities has been raised by several researchers [21, 122, 123, 141], andseveral teams have explored using displays to promote and support socialization[25, 61, 65, 82, 127] in urban public spaces. Particular interest in introducing dis-play technologies in Third Places has been previously discussed by a select group312.4. Summaryof researchers [22, 85] and several prototypes have been built and evaluated inThird Places [22–24, 36, 95, 110]. Most of these prototypes involve using technol-ogy to enable people within the same community to share media and content amongthemselves, or use the situated technologies as tools to support asynchronous com-munication with people who may visit the place at a later point in time. Severalstudies recount [25, 27] that often people adopt technologies in unexpected waysto serve their already existing socialization practices, and in some cases enjoy aperceived increase in sociability. While this is often only reported in an anecdotalway, the HCI community is beginning to gain a better understanding of ways inwhich this increase in sociability can be quantified.The use of the concept of social capital to better understand community func-tioning is relatively new within the field of HCI [136, 138, 139, 142]. This is mainlybecause the methodologies and tools to measure social capital are still evolving[129] in the field. Researchers have used a wide variety of theoretical and practicalapproaches [131–134] to quantify social capital, but it is only recently [128, 135]that measuring tools have been standardized across disciplines. The increased pop-ularity of online social networking sites provided an abundance of social network-ing data, allowing researchers to more easily explore social interaction using socialcapital as a tool. Tools to measure social capital include network analysis, stan-dardized surveys, usage data and ethnographic observations. Used in conjunction[140], these tools can yield a better understanding of how people use technologiesin social situations and how technology affects socialization.32Chapter 3Third-placeness and the Sociality,Publicity, Physicality FrameworkWe revisit Oldenburg’s concept of “Third Place” to reflect contemporary ThirdPlaces and the growing use of technology in socialization practices. We presenta new concept called “Third-placeness” to denote a state of human socializationcharacterized by information accessibility and adaptability, regularity, comfort andlightheartedness. Furthermore, Third-placeness is independent from the architec-tural and geographical constraints originally proposed by Oldenburg in his conceptof Third Place. We build this concept by revisiting Oldenburg’s Third Place con-cepts presented in the first section of this chapter.We also revisit the literature presented in Chapter 2 to gather characteristicsof interactive public displays that could support aspects of Third-placeness. Wepresent design dimensions that help designers organize the various aspects thatcome into play when designing technology for Third-placeness. This leads to theSociality, Publicity and Physicality Framework. We present the framework in thesecond section of this chapter.3.1 Third-placeness3.1.1 Motivation for Revisiting the “Third Place”The concept of “Third Place” was coined in the late 1980s by Oldenburg [15]and was inspired by the characteristics of physical places where face-to-face con-versation had historically played a key role, such as the American Bar, the En-glish Public House and the French Cafe´. In the second half of the 20th cen-tury, the growth of American cities, which favoured the automobile and suburban333.1. Third-placenessdwellings, had the effect of increasing the geographical distance between workingand dwelling places. This adversely affected the places “in between” where peoplehad historically met to socialize. Oldenburg pointed out the need to bring back such“Third Places” and discussed the key aspects that characterized them. Throughoutthe subsequent decades his books have inspired many followers to reinvigorate orcreate Third Places in settings like book shops, bowling alleys, community gardensand even prisons [28].Today, Third Places are still key to urban societies, as they provide communi-ties with physical spaces for levelling socialization. For example, in the case of arecent shop threatened with closure by the local government in Vancouver, Canada,its owner argued that “it’s great to build all [those] condos, and build higher density[dwellings], but where are these people supposed to go to meet their neighbours?”[143]. However, the functioning of contemporary Third Places bears little resem-blance to the Third Places described by Oldenburg in his books. The increasedpervasiveness of digital technologies – like online social sites, mobile phones andtelecommunication networks – has impacted how people socialize with each otherand how Third Places are used in confronting ways. For example, free WiFi offeredby some coffee houses encourages people to leave their home or work to visit theseplaces. Simultaneously, this technology is used by the same people to avoid humaninteraction while in said places [144]. Some coffee shops have decided to removeor ban technologies altogether, defining them as “distracting” [145] or leading topatrons “sitting all day long on one cup of coffee, blocking tables. [Resulting inplaces where] nobody [is] talking” [146]. Patrons themselves argue that “a coffeeshop should be a place to meet your friends and hold conversations and cultivateideas instead of ... sticking your head in a laptop” [147]. Yet, the presence of tech-nology in Third Places has only increased in the last decade. In a field study ofParisian Third Places [3] we observed that technology, commonly found in theseplaces, plays a key role in their function. For example, people are often occupiedand interacting with mobile phones instead of interacting with people nearby. ThirdPlaces, themselves, make an extensive use of technology to craft an online pres-ence through review sites like “yelp”, wayfinding applications like “google maps”,landing pages, online menus and social media accounts operated by place ownersor patrons. These virtual aspects of Third Places often extend to the physical world343.1. Third-placenessthrough URLs, QR codes, and signs on tables and windows. In some cases thesedigital aspects have a direct effect on the function of the places themselves. Forexample, online sites like “restovisio.com” allow people to choose a Third Placethrough a discrete menu of social functions and desired types of experiences, whichdirectly affects the neutrality and levelling nature of these places.This contradictory relationship between technology and the Third Place can betraced back to the origins of the concept itself. Oldenburg’s definition of ThirdPlace has its roots in the concept of “public” defined by Habermas as “the siteof collective performance that brings together those who are different from oneanother precisely because they are different...” [17]. Habermas’s concept of publicwas shaped by the architectural and functioning qualities of 17th century coffeehouses (one large room with one large table) that gave way to vivid conversationssupported by the media of the time – gazettes read aloud and passionately discussedbetween strangers[39]. It was these coffee houses filled with buoyant conversationbetween people from different classes that inspired many of the characteristics ofthe concept of Third Place. However, when the concept was forged, informationand communication technologies had not yet had a strong influence on society andwere not often found in Third Places. Media found in Third Places has changeddramatically over the last couple of decades. For example, books and noticeboardshave been replaced by laptop computers and wall-mounted televisions. These, andmany other digital technologies, are now woven into the “fabric of everyday life”[18] and allow people to experience a type of “public sphere” that extends outsidethe walls of the Third Place.The layout and functioning of many contemporary Third Places bear little re-semblance to the Third Places of 30 years ago. For example, in a contemporarycoffee shop tables are often designed smaller to accommodate single patrons ratherthan groups, power outlets are carefully placed to control seating dynamics, andwireless technologies are adopted as a desirable feature to attract customers. In-formation found online about a Third Place on reviewing or wayfinding sites candetermine the social activities of a place and its clientele. Yet, when we study anddesign technologies for the Third Place, we rely on Oldenburg’s original defini-tion, outlined before information and communication technologies had the effectthey have exerted on these places. It is important, then, that we revisit what it353.1. Third-placenessmeans to experience a Third Place today, and how technology can support suchan experience. Thus, we will revisit Oldenburg’s concept of Third Place in thefollowing subsections.3.1.2 Revisiting Oldenburg’s Third PlaceSome technologies that have closely affected Third Places include telecom-munication networks, mobile computing and online social networking sites. Thegrowth of wireless communications and the increased adoption of mobile phonesprovide people with uninterrupted access to information regardless of place andtime. The growth of online social networking sites in the last decade has madethe creation and maintenance of social relationships both easier and more perva-sive. Some people [145–147] argue that the growth of these technologies has had adetrimental effect on the places in which people have historically met to socialize.However, research in the field of Human Computer Interaction has shown that saidtechnologies could make use of online content to complement face-to-face social-ization [25], provide a platform for communal and asynchronous interaction [85]and support existing community practices [27].We revisit Oldenburg’s concept of Third Place to include the informationaland interactive roles that some of these technologies play in the functioning ofmodern Third Places. First, we propose that the concept can be detached from itsarchitectural and geographical constraints, and that it is independent from its natureas intermediary between the home and the work realms. Through this approach wecan pinpoint the aspects that promote the experience of Third Place, particularlythe one that arises through conversation, sociability and a sense of homeliness. Wecall this approach “Third-placeness” and define it as:The qualities of interpersonal interactions and contextual informa-tion that support the creation and maintenance of social capital.The key distinction between Oldenburg’s concept of Third Place and the con-cept of Third-placeness is that Third-placeness represents a ‘state’, while ThirdPlace represents a physical location. To compile the key aspects that shape Third-placeness we revisit each characteristic proposed by Oldenburg in his definition of363.1. Third-placenessThird Place. In the paragraphs below we propose alternatives in terms of informa-tion and communication practices for each one of these characteristics. Table 3.1summarizes our discussion.Neutral ground. For Oldenburg, Third Places provide a neutral ground wherepeople from different opinions and life views are welcomed. The key to being aneutral ground lies in unmediated and uncensored flow of information reflectingopposing opinions and life views. When free speech is encouraged, people fromdifferent walks of life can find a neutral ground on which to voice their opinions.Thus, neutrality can be defined in terms of the promotion of an uncharted space(virtual or physical) where information flows without censorship.Leveller. Oldenburg’s Third Place is characterized by making no social distinc-tions between people. Such a place may be referred to as a “leveller” due to itsability to level the gaps between social classes. The term “social class” was firstused during the rise of the bourgeoisie that took place in the 17th and 18th cen-turies, and in which 17th century coffee houses played an important role. It wasthese coffee houses, that made no distinction between class or origin, that allowedfor a bourgeois revolution to take place. The term “leveller” is tightly coupled withthis stratification of society and the early history of the concept of Third Place. Wepresent an alternative to the concept of being a “leveller” which broadens the termto include the minimization of any type of inequality between those that experienceThird Places, not just differences in social class. Whether the difference is social,political, economic, cultural, linguistic or geographic, people would be allowed tointeract with each other in a seamless and natural way.Conversation. For Oldenburg, face to face conversation is the main activity ofThird Places. This holds true in many of today’s Third Places. Today, however,this co-located conversation is often augmented with other types of conversation.Technology provides mechanisms that allow for dialogue to take place in a mannerthat is asynchronous and independent of geographical location. While some canargue that this detracts from the original purpose of Third Places we propose thatby allowing conversation to take different forms (synchronous or asynchronous,373.1. Third-placenessco-located or remote) socialization is encouraged and allowed to extend in bothtime and space. Spatially, and temporally, augmented conversation can in factextend the reach of the experience of Third Place.Accessibility and Accommodation. Oldenburg argues that Third Places are easyto access, both geographically and socially. They are close to home, not far fromwork, and in between them. This characteristic has a strong connection to theAmerican urbanism and society that inspired Oldenburg to create the very conceptof Third Places. However, technology has had a great effect on the boundaries ofThird Places. Mobile phones and wireless networks often allow visitors to reachoutside the physical boundaries of the place and even extend socialization prac-tices across time, through asynchronous communications. Online representationsof Third Places, for example on online social networks, allow for conversations tooccur independently of the physical characteristics of their hosting Third Places.Moreover, the separation between work and home has become less important withthe increased adoption of nomadic work and telecommuting. We propose, instead,that accessibility and accommodation in contemporary places are promoted wheninformation and conversation themselves are accessible. We argue that promot-ing accessibility and accommodation can be done by lowering adoption barriers,promoting information adaptability, and encouraging accessibility to information.Regulars. According to Oldenburg’s discussion of Third Place, regulars shapethe “tone” of a place. More importantly, it is when regulars welcome new peoplethat the true experience of Third Place arises. We propose that promoting and en-couraging regularity allows for the experience of “having a place” in a communityby allowing people to become insiders [29] of the experience of Third Place. Pro-moting such regularity can be achieved when people are allowed to experience aThird Place repetitively with ease.Low Profile. For Oldenburg, Third Places have certain physical characteristicsthat avoid “pretentiousness”. They are frugal, comfortable and relaxing. What iskey to this characteristic is that the experience of Third Place broadcasts its accessi-bility and levelling nature through its physicality and design. Oldenburg’s concept383.1. Third-placenessof Third Place cites physical design and geographic location as incarnations of thisbroadcast effort. Technologies can extend how Third Places showcase their ac-cessibility, and more importantly how they allow such accessibility to take place.Visual design and interaction flows that promote genuine and clear interactions canallow people to feel comfortable and promote a low-profile experience.Playfulness. Oldenburg specifies that the atmosphere in Third Places is light-hearted and conducive to play. Patrons are allowed, and encouraged, to raise theirvoices, be boisterous, and laugh out loud. The key to this characteristic is inter-personal interactions that allow for cheerful and carefree socialization. In con-temporary Third Places, where socialization occurs both face-to-face and virtually,playfulness might not be initially evident. However, encouraging playful interac-tions through technology could promote the experience of such a place towards theperception that it encourages playfulness.A home away from home. According to Oldenburg, Third Places should pro-vide places of comfort and familiarity that promote the feeling of “being at home”in public space. This characteristic has its origins in the concept of Third Placebeing linked to a suburban lifestyle and the need for a comfortable environment forsocialization outside of the home. In contemporary Third Places, online aspects ofa Third Place can actually occur at home, for example interacting with your localbar community through online social networks while at home. Thus, we argue thatrather than promoting “a home away from home”, the experience of Third Place ischaracterized by a strong sense of familiarity and comfort that can occur anywhere,even in the home itself.Characteristic Oldenburg’s Third Place Third-placenessNeutral Ground Third Places provide a neutralground where people with differ-ent opinions and life views arewelcomed.Unmediated and uncensored flowof information provides a neutralarena for free speech in Third-placeness.393.1. Third-placenessCharacteristic Oldenburg’s Third Place Third-placenessLeveller Third Places make no social dis-tinctions between people.Minimizing any type of inequality(for example political, economic,geographic or linguistic) encour-ages interaction between peopleregardless of differences.Conversation Face-to-face conversation is themain activity of Third Places.Providing mechanisms for con-versation – synchronous or asyn-chronous, co-located or remote –is key to Third-placeness.Accessibility andAccommodationPlaces are easy to access, both ge-ographically and socially.Accessibility to information isimportant to experience Third-placeness.Having regulars Regulars shape the “tone” ofa place. Regulars that wel-come newcomers are key to ThirdPlaces.Encouraging recursive interactionwith information could promoteregularity in Third-placeness.Low profile Third Places have physical char-acteristics and design that avoid“pretentiousness”. They are fru-gal, comfortable and relaxing.Information that promotes com-fort and accessibility could en-courage Third-placeness.The mood is playful The conversation in the ThirdPlace is light, and conducive toplay. A playful, loud, and con-versation filled atmosphere is al-lowed and encouraged.Interaction with information islighthearted, promoting a cheer-ful and carefree atmosphere, inThird-placeness.A home away fromhomeThird Places should provideplaces of comfort and familiaritythat promote the feeling of “beingat home” in public space.Information that promotes famil-iarity and comfort could encour-age Third-placeness.Table 3.1: Aspects of Third-placeness, compared to Oldenburg’soriginal characteristics of Third Place.403.1. Third-placeness3.1.3 Aspects of Third-placenessWe have revisited Oldenburg’s original definition of the concept of Third Placein terms of characteristics of the flow of information and interpersonal communi-cation. We argued that the Third Place can exist independently of its geographicaland architectural constraints, and consists of an experience of human interpersonalinteraction that supports the creation of strong social connections. We call thisapproach “Third-placeness” to denote that said characteristics of the experienceof Third Place can emerge anywhere where human socialization occurs. Third-placeness is not bound to architectural or geographical constraints and it is evokedwhen people socialize with each other, mainly through informational exchanges –most commonly conversations. This interpersonal interaction (synchronous, asyn-chronous, co-located or remote) is unmediated and uncensored, which minimizesall types of inequalities and differences between the people that participate in saidsocialization. In the state of Third-placeness there are few barriers to accessingand acting on information. Finally, Third-placeness is characterized by qualities ofcomfort and lightheartedness that are often strengthened by regularity and repeti-tion.These aspects of Third-placeness can be organized in three domains: “social-ity”, “publicity” and “physicality”. The sociality domain gathers structural andnormative aspects of social links that enable interpersonal socialization. We usethe concept of social capital, particularly looking at reciprocity, density and cen-trality, to model these aspects. The publicity domain gathers aspects of governance,specifically information ownership, accessibility and privacy, that enforce the typeof social interaction that characterizes the experience of Third Place. Finally, thephysicality domain gathers aspects that relate to physical characteristics of technol-ogy and places that support the experience of Third Place, specifically regularity,comfort and lightheartedness.In our research we measure sociality aspects using standardized social cap-ital questionnaires [135] and social network analysis. Aspects of publicity aremeasured with our own designed questionnaires and ethnographic observations,specifically shadowing and field notes. Finally, aspects of physicality are docu-mented using the ethnographic methodologies of shadowing and field notes. Table413.2. Sociality, Publicity and Physicality Framework3.2 presents each domain, its corresponding aspects and the methodologies used inthis research to measure them.Sociality Publicity Physicality(Structures and norms of so-cial links that enable interper-sonal socialization)(Rules of interpersonal so-cialization that enforce theexperience of Third Place)(Physical aspects of tech-nology and places that sup-port the experience of ThirdPlace)Aspects MeasuredReciprocity, Density, Cen-trality. (Conversation)Equality (neutral ground,and leveller), Accessibility(accessible and low profile) ,Ownership.Regularity (regulars), Com-fort (home away from home),Lightheartedness (playfulmood).MethodologiesSocial Capital Question-naires [135] and SocialNetwork Analysis.Questionnaires and Ethno-graphic Observations.Ethnographic Observations.Table 3.2: Aspects of Third-placeness can be organized in three dimensions: “so-ciality”, “publicity” and “physicality”. In parenthesis we present Oldenburg’s orig-inal Third Place characteristics. We outline methodologies used in this study tomeasure each aspect.3.2 Sociality, Publicity and Physicality FrameworkDesigning interactive public displays for socialization requires managing as-pects of Third-placeness, displays, publicity, interaction and social capital. Man-aging such complexity can be a daunting task. The field of design cognition,pioneered by Rowe [148], focuses on outlining systematic problem-solving pro-cedures and methodologies applicable to different fields [149, 150]. There aremany approaches to design [150]: participatory [151–153], user-centric [154–156],cognitive-engineering [157–159], rationale-based [160, 161], design as complex-ity management [162, 163], design as taxonomy [164–166], and theoretical [167–169]. In practice, designers apply methodologies from different approaches [170]to manage the complexity of their particular problems [171]. One way that design-ers manage this is through the use of ontologies and dimensions. These are toolsthat allow for the composition and decomposition of information [171–173], and423.2. Sociality, Publicity and Physicality Frameworkthe organization of different aspects of design problems [149, 174].Ontologies and dimensions are tools that are highly effective at organizingcomplexity and allowing designers to communicate their approaches to their peers.Ontologies help to specify and formalize concepts of a domain [174, 175] and pro-vide a common language that helps to bridge differences in methodologies andapproaches in any practice of design. Within an ontology, definitions associateparts of a problem (for example classes, relations or functions) with formal axiomsthat constrain interpretation and use of such entities [174]. This allows for mod-ularization of problems [176] that support the scalability of information queryingand reasoning of a design problem. This reduces complexity, improves context-awareness and promotes re-usability. Ontologies often formalize the quantitative,qualitative and conceptual aspects of a domain.Design Dimensions allow designers to manage quantitative, qualitative andconceptual aspects of a design simultaneously by allowing them to deconstructproblems and their aspects systematically [149, 177]. Dimensions allow the quickde-construction and re-construction of problems throughout the iterations of a de-sign problem [178], and may be situational, organizational or physical.Characteristics of interactive public displays and Third-placeness can be con-solidated into a single schema with three dimensions: aspects regarding to Social-ity, aspects regarding Publicity and aspects related to Physicality. The “Sociality”,dimension collects characteristics of technology and aspects of Third-placenessthat relate to social interaction, specifically how people create social capital andhow technology enables this through interactivity. The “Publicity” dimension col-lects characteristics of technology and aspects of Third-placeness that relate to theexperience of a public sphere characterized by familiarity, playfulness and con-versation, specifically how technology can support it through rules of ownership,accessibility and privacy. The “Physicality” dimension collects characteristics ofinteractive public displays and aspects of Third-placeness that relate to the tangi-ble experience of socialization when in public, particularly how technology is usedand adopted in the experience of Third Place. Table 3.3 presents each of thesedimensions.We can inform this initial schema of aspects of interactive public displays andThird-placeness dimensions. Dimensions provide a proven methodology to help433.2. Sociality, Publicity and Physicality FrameworkSociality Publicity PhysicalityInteractivePublicDisplays(IPDs)How people use inter-active public displaysfor inter-personal so-cialization (Interactiv-ity)How interactive publicdisplays enforce social-ization that supports apublic sphere of Third-placeness (Publicity)Physical and technicalcharacteristics ofdisplays that sup-port Third-placeness(Physical, Technical)Third-placenessStructures and norms ofsocial links that enableinter-personal social-ization. (Reciprocity,Density, Centrality)Rules of inter-personalsocialization that en-force the experience ofThird Place. (Equal-ity, Accessibility, Own-ership)Physical aspects oftechnology and placesthat support the expe-rience of Third Place.(Regularity, Comfort,Lightheartedness )Table 3.3: The Sociality, Publicity, Physicality Framework organizes aspects ofInteractive Public Displays and Communities into Sociality, Publicity and Physi-cality.manage the large number of constraints and definitions of our problem. Restruc-turing the above categorization (Table 3.3) into problem dimensions we obtain theSociality, Publicity, Physicality Framework Design Dimensions.3.2.1 Framework Design DimensionsWe use three modules to manage the complexity of designing interactive pub-lic displays for Third-placeness (Table 3.4). A conceptual module contains criteriathat describe the design phenomena as such through their intrinsic features. Thatis, how a system functions and what guidelines tie parts of the system together. Aqualitative module describes dependencies, reciprocal or hierarchical, and relation-ships between sub-parts of a system. A quantitative module contains informationthat helps define parts of a system as “a display” or “a place”; it includes geomet-ric and material properties, technical characteristics limited by manufacturing andphysical relationship to the environment of a system.Each module can then be mapped to a design dimension, leading to three de-sign dimensions (Table 3.5) that categorize approaching the design of interactivepublic displays to support Third-placeness. An organizational dimension encirclesconstraints and relationships that define how sub-parts of the design will be tied to-gether. Here we find aspects that facilitate the conceptual analysis of a problem bysupporting variations in the organization of the problem’s sub-parts [179]. They in-443.2. Sociality, Publicity and Physicality FrameworkConceptual Module Qualitative Module Quantitative ModuleAbstract criteria of displaysand their content: Expectedfunction and actions of dis-plays and their content. (In-teractivity)Dependencies afforded bycontent: Reciprocal andhierarchical relationshipsbetween people and content.(Publicity)Properties of displays: Ge-ometries and scales of systemparts. (Geometry Aspects)Physical characteristics ofdisplays: Materials used tocreate the system (Technicaland Perceptual Characteris-tics)Contextual criteria of Third-placeness: Guidelines defin-ing how communities are tiedtogether. (Linking Aspects:Social Capital, ConnectionType, Reach and Gain.)Connections enforcedby governance of Third-placeness: Rules thatenforce social connectionsbetween people. (Aspects ofGovernance.)Limits imposed by Third-placeness: Relational, or de-limitant. (Physical aspects:Appearance and Ambience.)Table 3.4: Modules: Conceptual Module, Qualitative Module and QuantitativeModule.clude constraints that define a conceptual binding within parts of a design througha cognitive meta-structure. In our case, meta-structures that define interactivity andinter-personal socialization through cognitive models are part of this dimension.A situational dimension encircles the functional and anthropological aspectsthat situate the design as part of human social governance. A situational dimen-sion gathers functional, economic, climatic, cultural, cognitive, psychological andergonomic aspects of interactive public displays and Third-placeness that shapea particular design problem. Specifically, aspects that specify dependencies andconstraints between entities and describe possible connections, dependencies andlimits when participating in a community or using technology. In our case aspectsthat relate to the functioning and governance of third place and the public spacebelong to this dimension.A physical dimension encircles aspects that relate to the physical characteris-tics of the design objects, and its sub-parts. A physical dimension gathers geomet-ric, dimensional, production specific, and material aspects that shape a particulardesign or experience of Third-placeness. Specifically, it collects basic informationof a design that allows a design to be defined metrically, in terms of scale or geom-453.2. Sociality, Publicity and Physicality Frameworketry, or in terms of physical size and scale. In our case geometrical and materialaspects of technology, and appearance and ambiance aspects of Third-placenessare gathered in this dimension.Organizational Situational PhysicalConceptual Aspects Qualitative Aspects Quantitative Aspects(Sociality) (Publicity) (Physicality)Interactive Public DisplaysInteraction Aspects: Interac-tivity.Functional Aspects: Public-ity.Geometrical Aspects: Physi-cal.Material Aspects: Technicaland Perceptual.Communities and Third-placenessLinking Aspects: Social Cap-ital, Connection Type, Reachand Gain.Anthropological Aspects:Governance.Physical Aspects Appearanceand Ambience.Table 3.5: Design Dimensions: Organizational, Situational and Physical.3.2.2 Framework Design CircleThe resulting framework organizes aspects and requirements of designing in-teractive public displays for Third-placeness. It also describes a way in whichto communicate about this design and a way such a design can be approached(through the use of dimensions). Table 3.6 presents the Sociality, Publicity, Physi-cality Framework. As already discussed, a design of interactive public displays forThird-placeness can be described via three dimensions, each describing the social,public, and physical aspects that form such a system. By walking through each ofthe requirements (abstract criteria, dependencies, properties, physical characteris-tics, contextual criteria, connections and limits) a designer makes sure to take intoaccount many of the aspects that come into play when designing interactive pub-lic displays for Third-placeness. In particular, designers can tackle (1) aspects ofsociality, that include structural and normative aspects of social links that enableinterpersonal socialization, (2) aspects of publicity, that include rules that enforcethe types of social interaction that characterize the experience of Third Place, and(3) aspects of physicality, that include physical aspects of technology that support463.3. Summarythe experience of Third Place.We provide a visualization of the Sociality, Publicity, Physicality Frameworkcalled the Sociality, Publicity, Physicality Framework Design Circle (Figure 3.1)that can be used to represent the aspects presented by our framework in a visualmanner. Around the radius of the circle, the reader will find each of the threedimensions proposed by the framework: Sociality, Publicity and Physicality. Theouter circle indicates the design dimensions and aspects (outlined by Table 3.6) thatfall under each of these three dimensions. The inner circle indicates the particularaspects of Third-placeness (white background) and the characteristics of interactivepublic displays (grey background) that fall within each dimension or aspect.3.3 SummaryWe revisit Oldenburg’s concept of “Third Place” to reflect contemporary ThirdPlaces and the growing use of technology in such places. We present the concept of“Third-placeness” defined as a state of human socialization that is independent ofthe architectural and geographical constraints of classical Third Places. We arguethat Third-placeness is evokwhen people socialize with each other, mainly throughconversation, and that it is characterized by equality, accessibility, lightheartednessand regularity. These aspects can be organized in three dimensions: “Sociality”,“Publicity” and “Physicality”. The Sociality dimension gathers aspects of interper-sonal socialization through the concept of social capital, particularly reciprocity,density and centrality. The Publicity dimension gathers aspects of governance,specifically information ownership, accessibility and privacy. The Physicality di-mension gathers aspects that relate to physical aspects of technology and placeswhere the experience of Third Place occurs.We revisit the background on interactive public displays to map how they canpromote characteristics of Third-placeness. We present dimensions that help or-ganize the various aspects that come into play when designing technologies forThird-placeness. We call this the Sociality, Publicity and Physicality Framework,and it consists of a list of key aspects and a diagram that can be used by designersas a guide to designing technologies to support Third-placeness.473.3. SummarySociality Publicity PhysicalityHow people use interactivepublic displays for inter-personal socializationHow interactive public dis-plays enforce socializationthat supports a public sphereof Third-placenessPhysical and technical char-acteristics of displays thatsupport Third-placenessConceptual Aspects Qualitative Aspects Quantitative Aspects(Organizational dimension) (Situational dimension) (Physical dimension)Interactive Public DisplaysAbstract criteria Dependencies PropertiesInteraction aspects Functional aspects Geometrical aspects• Interactivity: Representa-tion, Synchronicity, Author-ship, Embodiment, Aware-ness.• Publicity: Ownership, Ac-cessibility, Privacy.• Physical: Geometry(Shape, Scale, Orientation),Reach.Physical CharacteristicsMaterial aspects• Technical: Resolution, Per-ceptual Capabilities.CommunitiesContextual criteria Connections LimitsLinking aspects Anthropological aspects Appearance aspects• Social Capital Type: Trust,Reciprocity.•Governance: Neutrality, In-clusiveness, Rootedness.• Accessibility, Inconspicu-ousness.• Social Capital ConnectionType: Bridging, Bonding.• Ambience: Familiarity,Playfulness, Conversation.• Social Capital Reach andGain: Particularized, Gener-alized.Table 3.6: The Sociality, Publicity and Physicality Framework.483.3. SummaryConceptualAspects(AbstractCriteria)LinkingAspects(ContextualCriteria)FunctionalAspects(Dependencies)AnthropologicalAspects(Connections)AppearanceAspects(Limits)MaterialAspects(PhysicalCharacteristics)GeometricalAspects(Properties)PublicityPhysicality Sociality                Representation, synchronicity, authorship,     embodiment, awareness.Bridging or bondingtype of Social Capital.Particularized, generalizedreach of Social Capital.Geometry (shape, scale,orientation) and reach of tech.Resolution and perceptual capabilities of technology.Accessibility, inconspicuousness,familiarity and playfulness.Ownership, accessibilityand privacy of content. Neutrality, inclusiveness,rootedness and regularity.Social Capital type:trust or reciprocity.Figure 3.1: Sociality, Publicity, Physicality Framework Design Circle. This visu-alization can be used by designers to target the criteria in the Sociality, Publicity,Physicality Framework. In gray we outline aspects of Interactive Public Displays,in white aspects of Third-placeness.49Chapter 4Study of Supporting Aspects ofThird-placeness With InteractivePublic Displays in a Coffee ShopWe study the feasibility of inducing behaviours in a community that could sup-port several of the aspects of Third-placeness presented in 3. In collaboration withmembers of a student-run coffee shop, we co-designed a high-fidelity prototypecalled MySeedlings. MySeedlings is a table-top garden and private social networkdesigned to support the creation of social capital by encouraging people to interactreciprocally in order to take care of the garden. To understand how technologywould affect Third-placeness, we compared how two different designs affected theways in which people socialized with each other.We used the Sociality, Publicity and Physicality framework to inform the de-sign of the prototype presented here and to guide our measurements.Specifically, regarding Sociality we measured how characteristics of author-ship and awareness of information could affect interpersonal interactions betweenpeople. Similarly, for Publicity we documented how differences in informationownership and accessibility could affect perception of inclusiveness and rooted-ness, and regularity of usage. Finally, regarding Physicality, we documented howsome material and physical characteristics of technology, like display positioning,could affect accessibility and familiarity. Table 4.6 presents the instruments usedto measure each of these aspects of the Sociality, Publicity and Physicality Frame-work.50Chapter 4. Study of Supporting Aspects of Third-placeness With Interactive Public Displays in a Coffee ShopFigure 4.1: Seedlings Cafe main room. A couch can be seen facing the chimney.To the left of the chimney people can enter a second room often used for quietstudying. To the left one can see floor-to-ceiling windows, and to the right peoplecan access the bar and entry hall. MySeedlings can be seen in the centre of thephotograph.514.1. Background4.1 BackgroundSeedlings Cafe is a student-run coffee shop serving low-cost food and drinks.The coffee shop is located in the same building that hosts the Graduate StudentSociety building at the University of British Columbia. The coffee shop is en-tirely run by volunteers and is visited by students and faculty on a regular basis.Many of these visitors are closely acquainted with the staff or are volunteers them-selves. Seedlings Cafe has an inviting and cozy atmosphere permeated by chatterand laughter, especially during lunch hours (12pm to 3pm). The coffee shop lay-out consists of a small entry hall and two large rooms. When entering the coffeeshop through the entry hall people are welcomed by a bar and kitchen where theycan order non-alcoholic drinks or food. A four-seat table, often occupied by staffmembers, is situated in this space. Continuing forward, visitors can access the firstroom which hosts a decommissioned chimney, a couch, and tables to seat around50 people. Behind the chimney the second room hosts tables and private seatingfor 30 patrons. This second room is often used for studying and reading. Our studytook place in the first room, following advice from the staff, as it hosts the mostlively atmosphere in the coffee shop (Figure 4.1).4.2 MySeedlings DesignThe design of the prototype was born out of a co-design process with membersof this community. Co-design sessions were conducted during which researchersand volunteers discussed possible uses of technology to sustain or enhance com-munity activities. During these sessions volunteers suggested the use of a privatesocial network that visitors and volunteers could use to connect to other people inan asynchronous manner. The researchers suggested the addition of a game com-ponent on a gardening theme in order to support recursive interaction. During oursessions it was discussed that this theme fit particularly well with the concept ofthe cafe and participants felt it would appeal to those who were visiting the coffeeshop.A medium fidelity prototype was deployed as a technology probe [180] togather feedback on the design of the application, the game, and the choice of tech-524.2. MySeedlings Designnologies. A final prototype called MySeedlings was then developed further with acommittee of key volunteers.MySeedlings is a game and online social network that consists of a situatedLCD display and a table-top garden. People can use MySeedlings by visiting amobile application on their phone, tablet, or laptop. The game is played by ac-cumulating weekly points earned when watering the table-top garden using themobile application or by interacting with other people. Each week the three peoplewith the highest week points are awarded perks like free meals and drinks. Peoplecan also send private messages or virtual gifts, with no game value, to each otherusing MySeedlings.We used the Sociality, Publicity and Physicality framework to inform and di-rect our design decisions. We informed the design of game dynamics through theSociality dimension, in particular, we implemented two types of authorship (indi-vidual and collective) in the way points were calculated to be either earned by anindividual or the community. The aim behind using these two different types ofauthorship was to help us examine the difference in the degree of Third-placenessexperienced when each was deployed. We hoped to find out which authorship(individual or collective) would better serve to support the experience of Third-placeness.Similarly, we leveraged the Publicity aspects of ownership and accessibility toinform when and how to share information with the community. This was achievedby using the open accessibility and communal ownership of a public display tobroadcast game winners, with the aim of allowing newcomers to access such in-formation without a significant barrier. Research shows that making informationaccessible can make people more likely to participate in these kind of interactions.Therefore, we leveraged the private ownership and filtered accessibility of mobiledisplays to promote individual ownership of actions taken by people when usingthe prototype, e.g. messaging people.Finally, we informed our choice of technologies with characteristics outlinedby the Physicality dimension of the framework. We leveraged the public reachof situated displays to encourage open accessibility of information needed by thecommunity, like usage instructions and winner announcements. Similarly, we usedthe personal reach supported by small personal devices, like mobile phones, to534.2. MySeedlings Designsupport individual ownership of actions and anonymity when interacting with theprototype, for example to water the garden. This design decision aimed to avoidembarrassment related with the amplification of authorship often associated withusing large displays in public, and therefore to remove a potential barrier to partic-ipation.4.2.1 Components of PrototypeMySeedlings is composed of (1) a large display application, (2) a mobile webapplication, and (3) an actionable table-top garden. A typical usage scenario isexemplified by the storyboard in Figure 4.2: (1) Potential users passing by canspot the display and obtain information on how interact with the display. (2) Aspeople approach the display people can read more about the daily interactions andlearn about the purpose of the game. (3) Scanning the QR Code or visiting a URLallows people to access a mobile application. (4) Once logged in people can tap ona button that says “Water garden”. (5) When the display detects a person has tappedon the “Water garden” button, a gauge is updated to reflect this interaction and amessage reading “Thank you (FirstName) for watering the garden!” is displayed.(6) The surround of the table-top garden begins to glow and a small quantity ofwater is sprayed onto the garden.Large display application. The large display application is displayed using atwenty four inch (24”) monitor mounted on a wooden box hiding a portable com-puter. It consists of: (A) A status area containing interaction instructions, gamestatus and winners; (B) A noticeboard, labelled “Notes Board”, containing mes-sages posted through the mobile application. Each post shows a profile picture ofthe poster and text; (C) A list of available applications, only showing “GardeningApp”; (D) A QR Code with visual instructions to scan it; (E) A URL that could beused to access the mobile web application in a large font (Figure 4.3).Status area on large display application. The status area (Figure 4.3-A) con-tains: (1) a six seconds video demonstrating how to water the table-top gardenthat is dynamically hidden when someone accesses the mobile application to avoid544.2. MySeedlings DesignFigure 4.2: Storyboard exemplifying MySeedlings’s interaction flow. (1) Potentialusers passing by see how to interact. (2) When they approach the display they canlearn more about the purpose of the game. (3) People visit a URL. (4) Users tap ona “Water garden” button. (5) The large display shows their interactions. (6) Andthe table-top garden is watered.554.2. MySeedlings DesignFigure 4.3: Large display application: (A) status and main application area, (B)notice board, (C) list of applications, (D) QR Code, and (E) URL.distractions; (2) a diagram depicting visual instructions to show that tapping on“Water garden” results in watering the table-top garden; (3) a leaderboard display-ing a list of the top six users, showing their profile picture and accumulated pointsfor the week in descending order; (3) a list of winners showing only the profilepictures of the three users with the highest number of points accumulated duringthe previous week; (4) a list of the latest garden-watering events as text in the formof “FirstName LastName WeekDay Day Month”; (5) and according to each gamecondition, the following information: In the “individual” design (Figure 4.4-A),a vertical gauge showing “Poor, Moderate and Optimal” and a message in a bluebox reading “Each participation earns you 1 water point”. In the “communal” de-sign (Figure 4.4-B) a vertical gauge with the values “Planted, Grown, Harvested”and a message in a blue box reading “20 more participations to earn 20 harvestpoints”. Visual differences between each design condition are minimized as muchas possible to reduce the design of the large display layout interacting with ourmeasurements. Subsection 4.2.2 presents each game condition in detail.564.2. MySeedlings DesignA. Individual Condition.B. Communal Condition.Figure 4.4: Test conditions in the large display application574.2. MySeedlings DesignMobile application. The mobile application can be accessed by scanning theQR code (Figure 4.3-D) or visiting a displayed URL (Figure 4.3-E). It consists of alanding page, a game status page, and a page listing participants in the game. If theuser taps on the name of a participant, a page allowing the user to interact with thatperson. When a user visits the mobile application for the first time, they are pre-sented with a welcome message and a quick introduction explaining the purposeof MySeedlings as a game to win perks at Seedlings Cafe (Figure 4.5-A). Relevantinformation regarding the Ethics of participation in the study is made available, aswell as a way to contact the researchers. A button to “Sign in using Facebook” pro-vides an easy way for users to create an account. Once logged in, and thereafter,users are presented with their profile page (Figure 4.6-A) showing a breakdown ofpoints earned during the week, a blue button to water the garden and a list of re-ceived “favour” and “gift” points. According to each test condition, the followinggame information mirroring the display application is also presented: In the “in-dividual” condition, a vertical gauge showing “Poor, Moderate and Optimal” anda message saying “Each participation earns you 1 water point”; In the “commu-nal” condition, a vertical gauge with the values “Planted, Grown, Harvested” anda message reading “20 more participations to earn 20 harvest points”. Through thenavigation bar users can visit a list of the game participants, each list item contain-ing the participant’s picture, real name and week points (Figure 4.5-B). By tappingon a participant’s name or profile picture, users can access their page (Figure 4.6-B) showing that person’s points earned, buttons to send help or gift points, and aform through which a private message can be sent.Interactive table-top garden. The last component of MySeedlings consists of acoffee table (Figure 4.7) that has a white box encasing a grass and herbs gardenon the top and hiding a small water pump, a light bulb, a water reservoir, andthree inches of soil. When people visit their profile page in the mobile application(Figure 4.6-A) and tap on “Water garden”, a hidden computer actuates both thelight bulb and water pump for 10 seconds. Water flows out of a tube and visiblyirrigates the garden. Simultaneously, the bulb lights up and makes the white boxglow. The table-top garden was placed between the large display application andan preexisting blue couch, often used by customers of the Seedlings Cafe.584.2. MySeedlings DesignA. Application Signin page. B. List of registered people.Figure 4.5: Mobile application, signin and people pages, here shown the completeapplication as it would look when a user scrolls down594.2. MySeedlings DesignA. User Profile. B. Person public profile.Figure 4.6: Mobile application, private and public profile pages, here shown thecomplete application as it would look when a user scrolls down.604.2. MySeedlings DesignFigure 4.7: MySeedlings prototype deployed.614.2. MySeedlings Design4.2.2 Game ConditionsOur game design was informed by Hamari et al. [181] and aimed at enhanc-ing our prototype with motivational affordances to promote behavioural outcomes.According to Hamari et al. motivational affordances that stem from game design,which have psychological outcomes that can promote a desired behaviour is a validapproach to technology gamification. This is not the only position regarding gam-ification. Other researchers, for example Deterding et al. [182], argue that affor-dances implemented in gamification approaches have to be the same as the onesused in games, regardless of their expected outcomes. However, we take the for-mer approach and focus on implementing affordances that stem from a gaming dy-namic to promote an expected behaviour. In particular we implement two differentdynamics of collecting points by performing a repetitive action: an individual dy-namic where points are collected by individual actions, and a collective dynamicwhere points are collected when several members of the community collectivelyperform an action. While we expect differences in the resulting behaviours of eachgame design we are aware that gamification, as a phenomenon, is more complexthan many studies originally assume [183], and thus our observations are cautiouswhen analyzing the relationship between the game design and the measured out-comes.The game play of MySeedlings was designed to foster the creation of socialcapital when socializing. When engaging in social interaction, especially in a stateof Third-placeness, social value can take many forms. We are particularly inter-ested in “reciprocity” as a social value, that is, how much people are willing toreciprocate actions and how this is affected by collaborative game play. Thus, My-Seedlings was designed around value gained when reciprocating actions.During game play, MySeedlings users were able to perform the following ac-tions: (1) Water the garden by tapping on a button labelled “Water garden” (Figure4.6-A); (2) Send or respond to a favour request by tapping on buttons labelledrespectively, “Request Help” or “Fulfill Help Request” on a user’s page (Figure4.6-B); (3) Send a gift by tapping on a button labelled “Send Gift” on a user’s page(Figure 4.6-B); Or (4) send a private message to another user via that user’s page(Figure 4.6-B). Users earned points from watering (1) and receiving favours (2).624.2. MySeedlings DesignThese points were tallied on a weekly basis, allowing users to compete for foodand drink perks. Gifts (2) and private messages (3) are interactions that hold novalue when computing points. Private messages were introduced by request fromthe community, and gifts were introduced as a control measurement to be comparedagainst “favours”.Two game play modes were implemented and compared: “individual” and“communal”, with points awarded in a different way for each. The individual play-ing mode awarded points for watering in a manner favouring individual game play,and the other in a manner favouring reciprocal game play.In individual game play, users are awarded one point every time they tap on“Water garden”, which can be done only once per day. This makes the maximumgain available to participants the number of days of their participation in the game,or MaxGain = D. During this mode of game play, users were able to increasetheir water points by tapping on “Water garden”, regardless of whether or not otherparticipants were watering the garden. Table 4.1 presents the payoff matrix forwatering points during the individual condition, showing the points received byplayers A and B respectively. Differential scoring depends on one player beingmore active than the others in their watering activity.(B) Do Nothing (B) Water Garden(A) Do Nothing 0,0 0,D(A) Water Garden D,0 D,DTable 4.1: Individual game play payoff-matrix representing gains obtained by eachplayer’s strategy (A,B) with the best outcome for player A can take highlighted ingrey. D represents number of days.In communal game play, users need to collaborate in a pooled effort in orderto gain points. This is in contrast to individual game play, where players dependon the inactivity of other participants in order to gain a points differential. In acommunal game play, participants can tap on “Water garden” once per day to col-laborate on a predetermined fixed goal, or “pool”, in order to obtain communalpoints. Everyone that participates in the effort, regardless of the number of timesthey participated, is awarded the value of the pool, MaxGain = P . During thisgame play people will only increase their water points if the communal goal is634.2. MySeedlings Designachieved. The communal goal is set as the number of members of MySeedlings.Table 4.2 presents the payoff matrix for watering points during the communal con-dition.(B) Do Nothing (B) Join Effort(A) Do Nothing 0,0 0,0(A) Join Effort 0,0 P,PTable 4.2: Communal game play payoff-matrix representing gains obtained byeach player’s strategy (A,B) with the best strategy player A can take highlighted ingray. P represents the number of players in the pool.Favours (2) is a component present in both conditions where users can requestpoints from each other. When the other participant chooses to “reply” to a favourrequest, both the requesting and the responding user is awarded a point. Partic-ipants can send only one favour request per user per day. Therefore, the maxi-mum possible gain from favours is the number of days on which a user partici-pates, multiplied by the number of participants (minus the user) in the game, orMaxGain = D ∗ (N − 1). For example, person A is awarded a point either whenhe or she fulfills a request sent by person B, or when a person B fulfills a requestsent by A. Table 4.3 presents the payoff matrix for issuing and resolving favours,called “Help Points” in MySeedlings.(B) Request (B) Reply(A) Request 0,0 (N-1)*D,(N-1)*D(A) Reply (N-1)*D,(N-1)*D 0,0Table 4.3: Favours game play payoff-matrix representing gains obtained by eachplayer’s strategy (A,B) with the best strategy player A can take highlighted in gray.D represents number of days, N number of players in the game.Finally, weekly points are computed by summing, regardless of the game playcondition, watering points and favour points, or in terms of the equations providedabove: TotalPoints = WateringPoints+ FavorPoints. From a payoff pointof view, the payoff matrices in tables 4.1 or 4.2 are combined with the matrixin Table 4.3. For example, a player playing the individual game play choosingnot to do anything will gain 0 points per day, but might choose to reply to all644.3. Methodologypossible favour requests adding a maximum possible gain of D ∗ (N − 1), thatis TotalPoints = 0 + D ∗ (N − 1). A player choosing to water the gardenevery day and replying to all possible favour requests will get a TotalPoints =D + D ∗ (N − 1). Thus, during the individual condition users have a greaterchance of winning the game by focusing on watering the garden each day. Bycontrast, during the communal condition users have a better chance of success byfocusing on creating alliances of favours with particular individuals, as the pooledeffort results in every participant in the pool earning points.4.3 MethodologyMySeedlings was deployed at the Seedlings Cafe for 4 weeks (31 days). Eachweek three “winners” were selected and compensated with a voucher valid for onefree meal or drink at Seedlings cafe. Each week the points were reset to zero and anew leader board was created.4.3.1 Recruitment and Participant OverviewEfforts were made to recruit participants from multiple ethnic groups and agegroups. However, non English speakers were excluded. Participants were recruitedvia flyers pinned to previously existing community boards at Seedlings Cafe. Infor-mation regarding the research, including a notice of consent and details on data col-lection and anonymity for people willing to participate when visiting MySeedlings,were available via the MySeedlings’s mobile application, publicly accessible via aURL.Twenty nine (29) people participated in our trial. Seven (7) participants wererecruited via flyers while twenty two (22) participated by signing in through themobile application. In total, nine (9) participants had not visited Seedlings Cafeprevious to participating in the experiment. Of all participants, 52% were female(15) and 48% male (14). Participants were assigned to one of two groups (GroupA, or Group B) according to the experimental design presented in Table 4.5. Wepresent an overview of all participants in Table 4.4.654.3. MethodologyGroup N M F AgerangeAgeMeanAgeSDVisitsCafeFlyerRecruitGroup A 20 9(45%)11(55%)19-33 23.95 3.76 14(70%)4 (20%)Group B 9 5(55%)4(44%)20-28 25 2.73 6(66%)3 (33%)Total 29 14(48%)15(52%)19-33 24.27 3.46 20(69%)7 (24%)Table 4.4: Participant overview.4.3.2 Experimental DesignOur experiment followed a counter-balanced within-subjects design. Partic-ipants were assigned to a group (Group A or Group B) according to when theysigned up for the experiment. One of two conditions, either Individual or Commu-nal, was applied during each one of the four weeks of deployment, in a counter-balanced manner. Subsection 4.2.2 elaborates on the differences of each treatment.Table 4.5 presents our design. Twenty (20) people participated in Group A, whilenine (9) participated in Group B. We compared game conditions, Individual ver-sus Communal, to understand how game play affects different aspects of Third-placeness, in particular “sociality” aspects.Week 1 Week 2 Week 3 Week 4Individual Communal Communal IndividualGroup AGroup BTable 4.5: Deployment schedule and treatments applied.4.3.3 MeasuresFor each condition of the experiment we documented social capital, Third-placeness and technology adoption using surveys, interviews, ethnographic obser-vations and prototype usage data. Table 4.6 presents the instruments used in thestudy.To document social capital, a survey adapted from Western et al. [135] (Ap-664.3. MethodologyInstrument Time Applied MeasurementAspects of SocialityLikert Survey. Post-deployment (comparingusers versus non-users).Exclusion, agency, communityparticipation, particularized andgeneralized trust, communityspirit, place attachment, formalreciprocity.Favours. Per week. Density, reciprocity and cen-trality.Gifts. Per week. Density, reciprocity and cen-trality.Aspects of PublicityLikert survey. Pre-usage and post-usage. Communal, regulars, welcom-ing, leveller, playful.Likert survey. Post-usage Belonging, exclusiveness andsocial agency.Aspects of PhysicalityLikert survey. Pre-usage, post-usage for users.Post-deployment non-users.Enjoyment and intrusiveness.Informal interviews. Post deployment (users versusnon-users).Familiarity, perception, rate ofusage.Ethnographic notes. During deployment. Adoption of prototype.Table 4.6: Instruments used.674.3. Methodologypendix B – Surveys and Interviews) was used to measure informal structures (feel-ing of exclusion and particularized agency), formal structures (participation incommunity and particularized trust), informal norms (generalized trust) and formalnorms (community spirit, place attachment, and formal reciprocity). Furthermore,we analyzed the resulting networks of gifts and favours, on a week-by-week basis,for network density, reciprocity and centrality. In the case of reciprocity measures,values reported are Pearson Correlations whose value lie in a range −1 < r < 1.To allow for statistical analysis, reciprocity measures were Fisher-transformed toz-scores and used to test for statistical significance. In the case of centrality mea-sures, we performed student t-tests to determine statistical significance.Characteristics of Third-placeness were measured using a Likert scale survey.The survey measured: sense of community, feeling like a regular, sense of a wel-coming environment, allowing for levelling interactions, and sense of playfulness(Appendix B – Surveys and Interviews). Aspects of publicity and participationwere measured with two separate Likert scale surveys (Appendix B – Surveys andInterviews). The first survey measured how the prototype was perceived (enjoy-ment and intrusiveness) and was applied at the beginning and at end of the exper-iment. The second survey measured sense of belonging, exclusiveness and socialagency after using the prototype, and was applied as an exit questionnaire to partic-ipants. Surveys were complemented with an informal interview aimed at gatheringanecdotal data on the perception of the technology.We performed ethnographic observations that comprised documenting visitsto the coffee shop, noting dialogues and recording observed interactions with thetechnology. Documented interactions with technology included: glancing at thesituated display, approaching the display, talking about or using the technology,among others.Finally, it is important to note that Likert scale measurements for social capi-tal, Third-placeness and adoption are presented as-gathered and discussed to teaseout trends. We don’t report t-test analyses because these instruments gather non-parametric data that make it difficult to achieve statistically significant conclusions[184, 185].684.4. Results4.4 ResultsThe following sections present our findings on social capital, characteristics ofThird-placeness and adoption patterns.4.4.1 Social Capital SurveysFigure 4.8 presents our findings regarding social capital. Regarding informalstructures (exclusion and particularized agency) of social capital, MySeedlingsusers perceived themselves to be part of an homogeneous group, willing to helpother users of MySeedlings outside of personal interests. By contrast, visitors ofthe coffee shop that did not use the prototype tended to be more aware of differ-ences between themselves and fellow visitors of the coffee shop, and rated them-selves to be less likely to help strangers outside of personal interest. This alignswith our measurements regarding informal norms (particularized trust), where My-Seedlings users were more likely than non-users to give monetary help ($1) to peo-ple they had not previously met. Analysis of formal structures (participation incommunity) showed that both users and non-users of MySeedlings were willingto participate in the community. Measures regarding formal norms (communityspirit, place attachment, generalized trust and formal reciprocity) showed that peo-ple who used MySeedlings felt attached to the virtual place and showed a senseof trust towards other users. By contrast, non-users of MySeedlings were moreneutral in their feelings of attachment and generalized trust, but were more likelyto have a strong sense of community spirit. We have no strong evidence that thesedifferences are due to the use of technology, and could have occurred because peo-ple interested in MySeedlings were more gregarious. Further studies with a largersample are needed.4.4.2 Social Network AnalysisIn MySeedlings, “favours” (Table 4.3) are one-to-one requests that can bereplied to in order to earn points. In our social network analysis, favour requestsnot replied to are considered directed arcs from a node representing the personrequesting the favour to a node representing the person receiving the request. Sim-694.4. ResultsCompletelyDisagree (1)Disagree (2)Neutral (3)Agree (4)CompletelyAgree (5)ExclusionParticularized AgencyCommunity ParticipationParticularized TrustCommunity SpiritPlace AttachmentGeneralized TrustFormal ReciprocityMeanMySeedlings UsersSeedlings Cafe Visitors (Control)Figure 4.8: Social capital measures.ilarly, fulfilled favour requests are modelled as two directed arcs connecting the twonodes involved in opposite directions. MySeedlings’s “gifts” are sent in a unidi-rectional way from an issuer person to a receiving person, thus they are considereda directed arc from a node representing the person sending the gift to a node rep-resenting the person receiving such gift. We compare the favour and gift networksresulting from the individual and communal game conditions (Section 4.2.2).FavoursA communal condition resulted in a slightly higher reciprocity coefficient (p =0.722 and p = 0.592) than an individual condition (p = 0.496 and p = 0.409)(Table 4.7). However, using a Fisher r-z transformation to analyze for statisticaldifference we found no statistical significance when comparing treatments eitherwithin test groups or between them.Regarding density, we measured an apparent difference between Group A andGroup B (Table 4.4). As seen in Table 4.7 Group A (N=20) showed a lower densityratio (a¯ = 0.268 for individual condition, a¯ = 0.339 for communal condition) than704.4. ResultsGroup B (N=9) (a¯ = 0.833 for communal condition, a¯ = 1.805 for individualcondition). The smaller size of Group B may have played a role in promoting ahigher density, as people can reach a larger percentage of the group in a shortertime span.Week N a¯ p σp pminPerfect reciprocal ... ... 1 ... − a¯1−a¯Week 3 (Communal) 9 0.833 0.722 0.095 -5.000Week 2 (Communal) 20 0.339 0.592 0.159 -0.513Week 4 (Individual) 9 1.805 0.496 0.232 2.241Week 1 (Individual) 20 0.268 0.409 0.152 -0.366Areciprocal ... ... 0 ... − a¯1−a¯Perfect areciprocal ... ... −1 ... −1Table 4.7: Density and reciprocity coefficients of favours. Values are arranged indescending order for p. For comparison, values are: perfect reciprocal (p = 1),areciprocal (p = 0) and perfect antireciprocal (p = −1). We provide the standarddeviation (σp) when any link from the network is removed, and minimum expectedreciprocity coefficient (pmin) if there existed no reciprocal links.Using the resulting graphs for each game condition and test group we comparedactor degree, closeness and betweenness centrality using a student’s t-test (Table4.8 and Figure 4.9). Regarding degree centrality, we observed no statistical signif-icance either between treatments or between groups. For closeness centrality, wefound a statistical significance between treatments (t(27) = 2.7875, p = 0.0096between Week 2 and Week 4, and t(27) = 2.8546, p = 0.0082 between Week1 and Week 3). However, we also found a statistical significance between treat-ment groups (t(27) = 2.9983, p = 0.0058 between Week 1 and Week 4, andt(27) = 2.3937, p = 0.0239 between Week 2 and Week 3) suggesting that themeasured differences on closeness centrality were not due to the game play treat-ments applied. Our analysis of betweenness centrality showed only a statisticaldifference when comparing both test groups for the communal condition (t(27) =1.9642, p = 0.0599 between Week 2 and Week 3) suggesting that differences ingame conditions had no effect on betweenness centrality.714.4. ResultsDegree Closeness BetweennessWeek N µ σ µ σ µ σ1 Indiv. 20 0.0500 0.1288 0.2660 0.4732 0.0001 0.00062 Comm. 20 0.0500 0.1063 0.3440 0.4888 0.0008 0.00273 Comm. 9 0.1110 0.0902 0.8140 0.4899 0.0069 0.01344 Indiv. 9 0.1110 0.1536 1.4600 1.6703 0.0049 0.0140Table 4.8: Centrality of the favours requested and replied to each week. We presentmean (µ) and standard deviation (σ) values. 0 0.5 1 1.5 2 2.5Week 1 Week 2 Week 3 Week 4Meanp=0.005p=0.008p=0.009p=0.02p=0.05Degree CentralityCloseness CentralityBetweenness CentralityFigure 4.9: Centrality analysis of favours (statistically significant differencesshown).724.4. ResultsGiftsRegarding reciprocity, Group A (Week 1 and Week 2) had reciprocity ratios ofp = 0.126 during the individual condition and p = 0.145 during the communalcondition. This contrasted with Group B (Week 3 and Week 4) which showed a-reciprocal ratios (p = −0.025 for the individual condition, p = −0.052 for thecommunal condition). However, this difference was not statistically significantwhen analyzed using a Fisher r-z transformation (Table 4.9). Regarding density,Group A was more tightly connected (a¯ = 0.126 for the individual condition,a¯ = 0.121 for the communal condition) than Group B (a¯ = 0.041 for the individualcondition, a¯ = 0.069 for the communal condition).Week N a¯ p σp pminPerfect reciprocal ... ... 1 ... − a¯1−a¯Week 2 (Communal) 20 0.121 0.145 0.116 -0.137Week 1 (Individual) 20 0.126 0.122 0.085 -0.144Areciprocal ... ... 0 ... − a¯1−a¯Week 4 (Individual) 9 0.041 -0.025 0.025 -0.043Week 3 (Communal) 9 0.069 -0.052 0.034 -0.074Perfect areciprocal ... ... −1 ... −1Table 4.9: Density and reciprocity coefficients of gifts. Values are arranged indescending order for p. For comparison, values are: perfect reciprocal (p = 1),areciprocal (p = 0) and perfect antireciprocal (p = −1). We provide the standarddeviation (σp) when any link from the network is removed, and minimum expectedreciprocity coefficient (pmin) if there existed no reciprocal links.We compared actor degree, closeness and betweenness centrality of gifts usinga student’s t-test (Table 4.10 and Figure 4.10). We found no statistical significancebetween game conditions either for degree or for closeness centrality, suggestingthat our game play treatments had no measurable effect on centrality. We were notable to compare betweenness centrality because means yielded perfect data, that isµ = 0.734.4. ResultsDegree Closeness BetweennessWeek N µ σ µ σ µ σ1 Indiv. 20 0.0500 0.1805 0.0711 0.1740 0 02 Comm. 20 0.0500 0.1510 0.0748 0.2360 0 03 Comm. 9 0.1110 0.2514 0.0496 0.1039 0 04 Indiv. 9 0.1110 0.2222 0.0231 0.0443 0 0Table 4.10: Centrality of the gifts sent each week. We present mean (µ) and stan-dard deviation (σ) values. 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14Week 1 Week 2 Week 3 Week 4MeanDegree CentralityCloseness CentralityBetweenness CentralityFigure 4.10: Centrality analysis of gifts (no statistically significant differenceswere found).744.4. ResultsComparison of Favour and Gift ReciprocityUsing a Fisher r-z transformation we found no statistical significance betweenreciprocity of favours and reciprocity of gifts (Figure 4.11). Thus, we conclude thatthe apparently higher reciprocity values obtained during the communal conditionwere not related to differences between favours and gifts nor due to game playtreatments applied. 0 0.2 0.4 0.6 0.8Week 1IndividualGameplayWeek 2CommunalGameplayWeek 3CommunalGameplayWeek 4IndividualGameplayReciprocityFavours ReciprocityGifts ReciprocityFigure 4.11: Comparison of reciprocity coefficients of favours and gifts (no statis-tically significant differences were found).4.4.3 Characteristics of Third-placenessSurveys applied to each group, both at the beginning (N=10) and at the end(N=10) of the trial, allowed us to document characteristics related to aspects ofThird-placeness, particularly, being leveller, being welcoming and being commu-nal. (Figure 4.12).Regulars. Users, both at the beginning and at the end of their participation,seemed neutral about identifying themselves are regulars.754.4. ResultsCompletelyDisagree (1)Disagree (2)Neutral (3)Agree (4)CompletelyAgree (5)Communal Regulars Welcoming Leveler PlayfulMeanBeginning of trialEnd of trialFigure 4.12: Third-placeness measures.Leveller We asked participants if MySeedlings allowed them to interact withother participants they would not have otherwise interacted with. During the firstfew days of their participation users were neutral about this statement. However,after using the technology for two weeks most participants agreed this was the case.Playful Before their participation, users felt neutral about the technology beingplayful. However, at the end of their participation some users agreed this was thecase.Welcoming. At the beginning of their participation, users mostly disagreed aboutMySeedlings being welcoming. However, at the end of their participation mostusers agreed that MySeedlings felt welcoming.Communal. Participants were asked if taking care of the MySeedlings table-top garden felt like a communal effort. At the beginning of their participation,users had differing answers (some disagreeing and some agreeing). However, afterusing the prototype for two weeks, most of the participants agreed this effort was764.4. Resultscommunal.4.4.4 Surveys and Informal InterviewsA survey was applied to MySeedlings users (N=10) and coffee shop visitorsthat did not use the technology (N=10). We observed that neither users nor coffeeshop patrons found the technology to be intrusive to the coffee shop. Over time,people that used the MySeedlings found the technology more enjoyable, whilethose that did not use MySeedlings felt neutral about our prototype being enjoy-able. Neither users nor non users of MySeedlings found the technology distractingor intrusive to their regular experience of the place (Figure 4.13).CompletelyDisagree (1)Disagree (2)Neutral (3)Agree (4)CompletelyAgree (5)Enjoyable IntrusiveMeanUsers Beginning of TrialUsers End of TrialNon-Users End of Trial (Control)Figure 4.13: Perception of prototype, enjoyment and intrusivenessA second survey applied to MySeedlings users (N=10) showed that most par-ticipants felt a stronger sense of belonging to the coffee shop than to the digitalMySeedlings community. While the majority of MySeedlings users felt the tech-nology was inclusive, fewer than half of the participants said they would be willingto meet other users of MySeedlings in the future (Figure 4.14).We applied an exit interview to all users (N=29) of MySeedlings (Table 4.4).774.4. Results12345678910BelongingMyseedlingsBelongingCafeExclusiveness SocialAgencyNumer of ResponsesCompletely DisagreeDisagreeNeutralAgreeCompletely AgreeFigure 4.14: Feelings of belonging,exclusivity and social agency after using theprototype.Two participants (7%) commented they felt bored of MySeedlings after the firstweek. One user said that “at first it was fun, but after a while [he] was bored of it,just watering the garden everyday”, the other participant commented that “after thefirst day it was all the same, so [he] stopped playing”. A large percentage of peo-ple, 62% (18), commented they had originally thought of MySeedlings as a socialnetwork, but later perceived it as a “game” and found it more engaging and fun.For example, in one case a participant said “I thought I was signing up for anotherFacebook. But the points stuff made it better, not as boring”. When asked whichgame play people preferred, and why, 79% of people (23) said they did not havea preference. Of the total participants, 13% (4) preferred the individual conditionwhile 7% (2) preferred the communal condition. Comments made by participantsthat preferred the individual condition included that it allowed a “more instanta-neous way of getting points and giving points”, and a more direct way to knowtheir actions “actually watered the plant”. Allowed them to earn “bigger payoffs”and that it felt “less like a ’team’ game”. Those that preferred the communal condi-tion liked it because “it made it more important for other people to be participating784.4. Resultsin the game”, and “it felt more like [people] were working together”.In addition to our exit interviews, we interviewed a total of 26 non-participantvisitors of the coffee shop (14 women, 12 men), all of whom were students agedbetween 23 to 36. The interviews took place at the end of the fourth week of de-ployment, after the experiment had come to a conclusion (Table 4.5). Regardingfamiliarity with the prototype, all visitors (26) identified the name “MySeedlings”when a researcher asked if they were familiar with the name. All visitors wereaware of the public display application, but only 76% (20) knew that the displaywas linked to the table-top garden. Perception of the prototype was mostly posi-tive. When asked for their thoughts about MySeedlings, 50% (13) of people saidthey enjoyed or liked the system, defining it as “nice to have”, “interesting”, “nice”,or “exciting”. Of the remaining 50%, 15% (4) of people said they found it “dis-tracting”, 7%(2) said they found it “annoying” and 26% (7) were impartial, forexample using phrases like “it’s OK”, “I don’t have any particular thoughts aboutit”, or “it’s just there”. Only 30% (8) said that they would have liked the display toremain permanently in the coffee shop. Regarding interviewees self reported rateof usage of MySeedlings, 46% (12) said that they had used the seating areas next tothe display or table-top garden, 11% (3) said they had used the table-top garden towrite or eat, 26% (7) said that they had visited the URL of the mobile applicationand only 7% (2) said they had signed into MySeedlings to use it. Regarding thislow rate of self-reported usage, one person simply commented he “[had] no timefor [watering the garden], [he had] a lot of work to do”, another person explainedshe wouldn’t water the garden because she “[didn’t] know how to do it” and an-other person commented that “it [looked] complicated. [so she] wouldn’t use it.”Of those people that had created an account, one person commented that the gamewas not engaging enough to her and that although she “logged in to try it out andplayed for a couple of days, [she] forgot about it”. One other person commentedhe “tried it, but I didn’t know it was a game [and he] thought it was a one time offthing.”Surveys and interviews helped us document opposing views regarding the tech-nology. While some people, regardless of whether they used the technology or not,were optimistic about MySeedling’s purpose and usefulness, others found it dis-tracting or viewed it as a temporary addition to the coffee shop. These observations794.4. Resultshighlight that the introduction of an artifact into a Third Place can have differentmeanings to different people. Our ethnographic observations, outlined below, seemto imply that the complex nature of this coffee shop, with its many activities andsub groups of people, could be the cause of these differences.4.4.5 Ethnographic ObservationsWe carried ethnographic observations to document aspects of “publicity” and“physicality” of Third-placeness. Particularly, aspects of governance and technol-ogy adoption promoted by the introduced technology. A researcher visited thecoffee shop for 13 days during the four weeks of deployment, for an average of 3hours per day, totalling 43.5 hours of observation. The researcher sat on a tablenext to MySeedlings and took notes on a laptop and occasional photographs usinga mobile phone. Notes included the number of people entering the coffee shop, thenumber and nature of interactions with the display, incidences of evident usage ofMySeedlings, the number and nature of conversations regarding MySeedlings andincidences of unexpected usage of the system components. To minimize impacton the scenarios observed, the researcher avoided technical support (for example,fixing arrangement of display or fixing lack of connectivity) during observationhours, acted as a visitor of the coffee shop and interacted with the community whenneeded. However, due to the tight-knit nature of the community it is expected thatthe presence of the researcher was noticed by coffee shop patrons. This might haveimpacted how visitors of the coffee shop used the display while a researcher waspresent and taking notes.Observed Experience of Visiting Seedlings CafeMost of the people that visit the Seedlings coffee shop are regulars or volun-teers, and have previously met each other (Figure 4.15). This leads to playfulnessand familiarity that promotes Third-placeness for many patrons. Banter and rau-cous laughter are not only allowed but encouraged and treasured. The followingaccount from our observational research detailing an event in the coffee shop illus-trates its ambience and character.804.4. ResultsSome time in the morning one of the staff members came to the mainarea asking “does anybody have a screwdriver? Someone is lockedin the washroom.” A patron promptly gave her a multi-tool. Someminutes after, the staff member came back and boisterously said: “It’sall right! He’s fine, it was [John], Whooo!!!”. Some people laughedwhile others cheered and clapped. As the staff member left we couldall hear the banter in the kitchen as they laughed at the fact that awindow had to be unscrewed to rescue a person trapped in the toilet. 0 20 40 60 80 1001 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17People(Individual Condition) (Collaborative Condition) (Individual Condition)CustomersRegularsFigure 4.15: Observed customers and regulars of Seedlings cafe. We counted asregulars those people who visited the coffee shop at least twice during a calendarweek.Initial Adoption PatternsWord of mouth played a leading role in communicating the purpose of ourprototype and encouraging people to try it for the first time. On several occa-sions, visitors to the coffee shop were observed asking other patrons the purposeof the table-top display. People that had previously used MySeedlings would often814.4. Resultsverbally explain the technology and encourage them to try it. This interaction isexemplified by the following recorded dialogue:[A couple just sat down on the couch next to the garden, the woman begins readingthe large display.]WOMAN: Wow, this is very pretty.MAN: What is it?WOMAN: Oh this is the seeds thing! You can water this grass.You’re supposed to use a phone, you click on the phone andwater goes out here [pointing at the garden].MAN: It would be nicer if it had some pretty flowers.WOMAN: They don’t grow in there, some moss would be better,or chives maybe. Want to try it?MAN: I’m fine. [About 30 minutes later the couple have settled down andare working on their computers. The man takes out his mobile phoneand tries the prototype, water begins to flow, irrigating the garden.]MAN: That is cool!WOMAN: Wow. [Anna] would love this.Long-Term Adoption PatternsThe introduction of MySeedlings in the coffee shop was welcomed with curios-ity (20 people registered in the first 2 days). However, the novelty of the technologyquickly fade off. We originally designed the table-top garden to be able to placebelongings, food and drinks. But overtime people used it more as a table than aninteractive garden (Figure 4.16). Most of the observed interactions with the tech-nology were carried out by individuals or small groups of people (3 or less). Bycontrast, larger groups of people that sat around the garden or close to the displayoften focused on talking to each other. People that discussed MySeedlings withina group, often tended to interact with the technology at a later time.Story-telling played an important role in shaping how other people perceivedand used the technology. On several occasions, newcomers would ask out loud824.4. ResultsFigure 4.16: People appropriated parts of the prototype to fit the functioning of thecoffee shop. For example, the table-top garden was used as a coffee shop table.834.5. Discussion“What is that!?”, prompting personal stories from people in the vicinity regardingtheir own experience of using MySeedlings. These stories promoted curiosity innewcomers and shaped how the community understood the technology. For ex-ample, people often referred to MySeedlings as “the garden” rather than a socialnetwork, furniture piece, or interactive display. When people talked about “thegarden” they often referred to it as “something that is going on at [the coffee shop]Seedlings”, positioning the technology as a long-term characteristic of the coffeeshop rather than a one-time introduction of technology. For example, in one con-versation between two patrons, one asked “have you tried the garden” to which theother replied “MySeedlings?, I haven’t tried it yet, but I want to. Maybe tomor-row”.Finally, comparing the number of observable physical interactions (approach-ing the display, publicly watering the garden, Figure 4.17) with usage of the tech-nology in data logs (watering the garden, requesting and replying to favours, send-ing gifts, etc., Figure 4.18) we observed a discrepancy suggesting that people usedthe prototype either as passive observers or from outside the walls of the coffeeshop. Three of our participants from Group A and Group B had never visited thecoffee shop prior to our study, but after using MySeedlings began vising the cof-fee shop on a regular basis. In an interview with one of these participants shecommented that “playing MySeedlings [made her] want to [visit] the coffee shopmore, [she] felt [she] was taking care of the garden, so [she] came in more often,to check on it”.4.5 DiscussionThrough this study we documented whether or not technology can induce be-haviours that support aspects associated with our concept of Third-placeness. Weobserved the introduction of a prototype consisting of a digital display, a mobileapplication and an interactive table-top garden into a community of students thatregularly visit a University coffee shop. Specifically, we introduced a table-top gar-den with an associated private social network, designed to support the creation ofsocial capital by encouraging people, through gamification, to interact reciprocallyin order to take care of a table-top garden.844.5. Discussion 0 2 4 6 8 10 12 14 16 181 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17People(Individual Condition) (Collaborative Condition) (Individual Condition)Approaching to useApproaching to readGlances when entering placeTalking about displayFigure 4.17: Observed interactions with MySeedlings’ situated display.854.5. Discussion 0 10 20 30 40 50 60 708 15 22 31Number of 'Water garden' tapsNumber of Favours requestedNumber of Gifts sentFigure 4.18: Usage of MySeedlings per day, checkins, favours and gifts.864.5. DiscussionWe measured aspects of Sociality, Publicity and Physicality in the expectationthat the introduction of the technology would promote these aspects to a certain de-gree. Furthermore, we measured how patterns of socialization (reciprocity, density,centrality) were affected by different game designs implemented in the technology.4.5.1 SocialityRegarding the measured aspects of Sociality (reciprocity), social network anal-ysis showed that differences in game play (individual vs communal) had no sig-nificant effect on how people used MySeedlings. Network analysis of favourshowed that although a community-centric game play seemed to increase reci-procity of favours, this effect was not statistically significant. Similarly, neitherdensity nor centrality of favours seemed affected by game play. We documenteda slight change in closeness centrality, throughout the duration of our technologydeployment and across test conditions and participant groups, possibly due to ex-ternal and here unmeasured factors. Network analysis of gifts showed game playhad no statistical effect on reciprocity, density or centrality.4.5.2 PublicityIn relation to aspects of Publicity (ownership, accessibility and privacy of con-tent), our analysis of social capital suggests that some people that used the My-Seedlings prototype saw themselves as part of a group with similar goals. Usersof MySeedlings showed an increased sense of social capital, when compared tovisitors of the coffee shop who did not use the technology. It is our belief thatthe increase in social capital that was observed in participants of MySeedlings, ascompared to non-users of the technology, was the result of increased aspects ofThird-placeness encouraged by promoting a levelling and welcoming virtual spacecentred around regularity and playfulness. From our observations, it is unclearwhether or not these behaviours are correlated to the design of the technology in-troduced and a further study is needed to understand this relationship.874.5. Discussion4.5.3 PhysicalityRegarding aspects of Physicality (accessibility and familiarity), through sur-veys and interviews we found differing opinions regarding the technology. Somepeople seemed to enjoy the prototype and found it non-intrusive to the coffee shop,while others found it intrusive or were neutral to the technology. While most vis-itors of the coffee shop were aware of the technology deployed, perceived com-plexity and apparent lack of purpose discouraged some people from signing in.Our ethnographic observations showed that regulars of the coffee shop, who wereexposed to MySeedlings repetitively, quickly accepted the technology as part of thecoffee shops background and furniture. We also noted that MySeedlings was moreuseful, in terms of allowing community interaction, to non-regulars of the coffeeshop, as it provided a way for newcomers to engage with a community, althoughnot necessarily the one visiting the coffee shop.4.5.4 ConclusionThe technology we introduced seemed to have encouraged some behavioursthat could support aspects of Third-placeness. While we observed a relationshipbetween our design and these behaviours, we have no evidence that the character-istics of the technology we measured are the cause of these behaviours. People thatused the technology seemed to have experienced characteristics of social capital,like community spirit, place attachment and trust in other members of the digitalcommunity. However, these measurements could have been due to factors externalto our measurements. We documented how certain aspects of the technology weremost useful to newcomers searching for a way to participate in the community. Forexample, the open accessibility of the situated display provided newcomers with atalking point through which to engage in face to face conversations with regularsof the coffee shop. Similarly, the privacy supported by the mobile application al-lowed those less social by nature to participate in the game in private, while stillhaving access to a digital form of sociability, like virtual gifts and conversations.In addition, we learned that while some participants voiced the fact that technol-ogy encouraged them to visit the coffee shop more, several people argued that thetechnology was just another background component of the experience of being at884.5. Discussionthe coffee shop, and at times distracted from this experience. We learned that thecomplexity of a real-world Third Place is a difficult environment in which to mea-sure the relationship between the design of technology and behaviours that supportThird-placeness. Unpredictable changes in composition of community members,sudden changes in community goals and the sheer number of activities performedconcurrently in a public place are just a few of the factors we experienced that in-troduced complexity in our study. The experiment presented in Chapter 5 providesa different approach in which many of these factors were mitigated by choice ofsetting or experimental design.89Chapter 5Study of the Impact ofInformation Visualization onCommunity Reciprocity in aWorkplaceWe learned in the previous study (Chapter 4) that promoting a welcoming andleveling environment could encourage behaviours – like trust, community attach-ment and community spirit – that could help support the experience of Third-placeness. However, our study found it challenging to connect how our designof technology was linked to the documented experiences of Third-placeness. Welearned that the setting of said study – a student coffee shop – introduced a highlevel of complexity and external factors that made drawing a connection betweentreatments and measurements difficult.This study therefore focuses on understanding how interactive public displaysaffect one such aspect, namely interpersonal socialization. Looking at one aspectsimplifies the process of observing connections between characteristics of technol-ogy and aspects of Third-placeness, because it is possible to control for that aspectand directly observe behavioural chances as you modify that one aspect. Its settingin a workplace also allowed us to control and document many factors that previ-ously introduced complexity in our observations in Chapter 4, like unpredictablechanges in community members, changes in community goals and undocumentedactivities performed by visitors of such places. Being a workplace, the communityhas well defined goals and daily practices, along with a well known rate of placeattendance. Furthermore, through documenting the community goals and practices90Chapter 5. Study of the Impact of Information Visualization on Community Reciprocity in a Workplacewe could design a technology that fit better with the community needs.The workplace setting of this study was supported by a research collaborationbetween the Human Communication Technologies Laboratory at The Universityof British Columbia and the Advanced Interaction Laboratory at the Federal Uni-versity of Sa˜o Carlos. This setting allowed us to explore the idea of the existenceof a state of socialization (Third-placeness) outside of conventional Third Places.An Ethnographic study was conducted at the beginning of our study to documentthe possible existence of aspects of Third-placeness in this setting [11]. Our studyrevealed that aspects of Third-placeness emergent through sociability practices spe-cific to the Brazilian culture pervaded work flows in this setting. This sociabilityencouraged storytelling and playfulness that antagonized and challenged a perva-sive, yet unspoken, social and economic stratification of the workplace. In ourstudy, we documented how this sociability encouraged a Third-placeness that al-lowed people in this community to circumvent hierarchies, through conversationand playfulness, in order to get things done efficiently.Following our ethnographic study we designed and introduced a task ticketingsystem into this community of work professionals, in order to support conversa-tion while allowing us to quantify how changes in the design of such technol-ogy affected the functioning of such a community. We documented how changesin information visualization affected interpersonal socialization, specifically reci-procity, density and centrality. Specifically, two types of information visualizationwere compared. An individual-centric visualization presented people as isolatedindividuals, while a community-centric visualization showed frequency of com-munication between members of a connected group.We used the Sociality, Publicity and Physicality framework to inform the de-sign of the technology and to guide our measurements. Regarding Sociality, wemeasured how different types of information representation could affect social reci-procity. For Publicity, we documented how different types of information represen-tation could have an effect on information ownership, ultimately affecting how acommunity experiences inclusiveness. Regarding Physicality, we documented howphysical aspects of displays, particularly scale and reach, could affect accessibilityand familiarity of information.915.1. Background5.1 BackgroundThe Centro de Atenc¸a˜o Integral a` Sau´de (C.A.I.S.) Clemente Ferreira, locatedin Lins, Sa¸o Paulo, is a not-for-profit state-run institution focusing on psychiatricdisorders and neurological damage. The centre’s mission is to contribute to the im-provement of the lives of patients with the guarantee of continuous quality of lifeduring their transition to public life (Figure 5.1). Human interaction is promotedby a wide range of activities inside and outside the centre to encourage patient so-cialization and well-being. The centre’s architectural structure is arranged in fourwings, each about 120 metres in length, and each hosting 20 to 25 rooms. Thislayout is repeated across three storeys, totalling almost 300 rooms. Some of thenon-medical services within by the building include a large kitchen and cafete-ria, pharmacy, cinema, hand-crafting workshop, hairdresser, and transition-phaseapartments. Currently, there are over 800 health care professionals employed andclose to 380 patients are attended to every day. Adults, teenagers and childrenare cared for separately across 8 different care units. Each unit is managed by anadministrative director and employs floor social workers, physiotherapists, nurses,auxiliary nurses, a phonologist, doctors and other administrative support staff to-talling between 15 to 20 professionals. The centre is a small “city” in its ownright, with governance, self-sufficient services and daily routines essential for itsfunctioning.5.2 Ethnographic ObservationsBefore commencing our study we performed ethnographic observations to doc-ument practices existing (Figure 5.2) before the introduction of our technology. Weshadowed 12 professionals during the course of one week. Two researchers fol-lowed each professional during the whole duration of one work shift, documentingand recording their natural comings and goings within the centre. One researcherdocumented the spatial nature of each room or space visited, recorded gesturesand mannerisms and recorded each information exchange where social links oftrust or reciprocity where created or utilized. A second researcher video-taped andphotographed significant activities observed and helped identified culture-specific925.2. Ethnographic ObservationsFigure 5.1: C.A.I.S. Clemente Ferreira, located in Lins, Sa¸o Paulo, Brazil.935.2. Ethnographic Observationsidioms and implied meanings. To gain insight into Brazilian culture, one semi-structured interview was performed at the beginning of the shift to discuss themeaning of “being together”. A second informal interview was performed at theend of the shift to clarify any observations about which we felt we needed a deeperunderstanding.Figure 5.2: A professional verbally relays news information to other co-workers,which leads to a lengthy discussion.Following our observations of all 12 professionals, we documented four re-curring practices of the studied community (Table 5.1). All of these practices(Brazilian sociability, spontaneity of meetings, intermittent storytelling, and infor-mation relaying) equally promoted different aspects of sociability that encouragedThird-placeness in the workplace. In particular we observed the experience of anaccommodating, low-profile, neutral, inclusive workplace that promoted regularsociability and conversation. Said experience of Third-placeness encouraged thecreation of social capital relationships that led to a more efficient workplace, dueto circumvention of hierarchies and reciprocation of workplace tasks.Brazilian sociability. Participants agreed that Brazilian social life is tightly inter-woven into the key settings of everyday life: the workplace, the commute and the945.2. Ethnographic ObservationsTable 5.1: Third-placeness characteristics promoted by characteristics of the stud-ied facility and community.Sociability Spontaneity Storytelling InformationAccommodatingLow profile XNeutral ground X XInclusive X X X XRegulars X X XConversation X X X Xhome. This was clear in our shadowing observations during which we found “hav-ing fun”, or brincadeira, was a common element in the majority of interpersonalinteractions. We observed that almost every case of newly established informa-tion exchange (for example, fetching some medicine from the pharmacy) wouldbegin with an informal greeting (e.g. tudo bem?, or “is everything ok?”) or a joke(e.g. teasing about soccer team affinity). Once this sociable connection was es-tablished, work related activities would be performed as usual (Figure 5.3). Thisseemed to occur, as one participant argued, “because this place [C.A.I.S.] is verytough both mentally and physically, and we need each other for everything, so wetry to keep a good relationship.” This Brazilian sociability is exercised in almostevery interpersonal encounter. As a result, people within the community oftenknow each other, at least as an acquaintance. This allows members of such a com-munity to leverage a network of social connections to perform better at their dailytasks (Table 5.1). Said sociability, however, should be understood within a Brazil-ian context. One participant quoted the phrase uma ma˜o lava a outra, or “one handwashes the other”, to explain the way in which Brazilians help each other if theyexpect to be helped out in return.Spontaneity of meetings. The architectural dimensions of C.A.I.S. are quitelarge, with each wing of the hospital spanning over 100 metres long, and a corebuilding 300 metres in length. Essential services, like the kitchen, administrationand activity rooms are distributed throughout the building or located outside themain building. Walking from one unit to another, in order to perform daily patient955.2. Ethnographic ObservationsFigure 5.3: Professionals briefly joke around about soccer team affiliation beforecarrying out official business.care, can take between 5 and 20 minutes, and professionals make such journeysan average of 6 times per shift. The architectural arrangement of the centre hasforced professionals to be in motion, often coming together in unexpected andunplanned places like common areas and centre-wide services (e.g. pharmacy orkitchen). These places of encounter promote communication between profession-als and allow for social ties to be formed and maintained on a daily basis. In fact,people often seek to perform tasks that require traversing the centre (e.g. pickingup medicine at the pharmacy) to “escape” the architectural limits of their units and,while performing a task, enjoy a quick chat with friends and acquaintances bothinside and outside of their work unit. These experiences of conviviality are ofteninformal, nomadic, and spontaneous. Such spontaneous pockets of sociability havemany of the characteristics of Third-placeness: inclusiveness, playful conversation,filled with regulars and a neutral ground. These nomad pockets of Third-placenessmay occur by doorways to the vegetation surrounding the building, around-the-corner offices providing visual shelter from supervisors, or in common rooms thatprovide a service not found within the professionals’ own units (Figure 5.3 and5.4). These nomad and flexible states of Third-placeness are important to the func-965.2. Ethnographic Observationstioning of C.A.I.S. as they allow for the verbal exchange of information betweenthe architecturally separated units and promote a sense of “togetherness”.Figure 5.4: Professionals escape work by creating an impromptu space for conver-sation at a side door.Intermittent storytelling. We found that pockets of Third-placeness were oftenpromoted by the intermittent telling of work and personal stories. For example,every day at midday the nurses in almost every unit take a 15 to 30 minute break tograb a cup of coffee at their unit’s kitchen. They utilize this time to tell each otherstories that often involve the centre, a patient’s progress or an event in their personallife. Stories flow naturally from personal to work-related and back. This not onlyhappens at lunch or at coffee breaks, but occurs unexpectedly in the aisles of eachunit or in common areas throughout the building during unexpected encounters.For example, in one instance, while carrying out their daily tasks, a professionalbriefly told another a story regarding a recent car accident and how it was related toa mutual colleague in another unit. While the second professional was answering975.2. Ethnographic Observationswith her own side of the story, a third professional joined the conversation anddid the same. The occurrence of such storytelling is facilitated by the geographiclocation of the building and its architectural layout. One explanation behind peopleengaging in face-to-face storytelling during rests, rather than attempting to usetechnology to communicate with the outside world, can be the poor cell phonecoverage within the hospital. As one participant explained: “I gave up checking myphone during breaks.” As a result, we observed that the community has cultivateda high level of verbal conversation between themselves over the years and as oneprofessional jokingly argued “[they] come here to talk. That’s all [they] do.” Thetrend for employees engaging in high levels of social conversation is supported bythe fact that tasks that require only one person, like picking up medicine at thepharmacy, are often performed in groups and the suggestion of ‘going together’ isused as an opportunity to chat.Information relaying. In our interviews it was made clear by several of our par-ticipants that Brazilians from different social strata belong to different communitiesin their personal lives. This stratification of social life permeates the workplace, notthrough social class, but through job hierarchy, and shapes how people create socialconnections within the work space walls. We observed a clear hierarchical distinc-tion between undergraduate-level professionals and graduate-level professionals,and between graduate-level professionals and directors when it comes to socialinteraction. As one participant commented: “[outside work] I am friends withthe director, not close friends, but friends, and I can’t [show] that here, becausepeople think I’ll take advantage of that friendship”. Thus, official mingling oftenoccurs within similarly-educated or equally-ranked professionals. However, unof-ficial participation in storytelling occurring in corridors and public spaces through-out the centre is open to all ranks of the workplace. These pockets of storytellingencourage aspects of Third-placeness (for example, accessibility, neutrality, con-versation) that promote the creation of generalized trust and reciprocity that perme-ates all hierarchy levels across the institution. As a result, when performing dailyactivities, hierarchy divisions are considered flexible by professionals. That is, thetrust built through this intermittent storytelling and sociability allows professionalsto cross hierarchies in order to solve problems. For example, we observed a case985.3. Ticketing System Designwhere, in an informal setting, an auxiliary nurse told some of ther fellow profes-sionals the story of a patient not accepting his food rations. Participating in saidconversation was a phonologist, whose official task is related to cognitive ratherthan care duties. However, after the conversation was over the phonologist tookthe initiative of calling the kitchen to update the patient’s food file. In this examplehierarchy bureaucracy was avoided to solve a simple, but essential, care problem.5.3 Ticketing System DesignWe observed that in C.A.I.S. one of the well-established work practices is theusage of social reciprocity for task completion. People consider solving workplacetasks for each other as doing “favours” for each other, or maintaining a certain levelof reciprocity. According to Putnam [16], acts of reciprocity can be understood asmutual obligations or social payment slips issued when people help each other inthe expectation that they (or someone else in the community) will return the favourin the future.Within the studied community of professionals, ad-hoc socialization has be-come an integral part of efficient delegation and management of tasks, as it pro-motes efficient communication of task status and helps to ensure that tasks arecompleted on time. Completing these tasks is not only seen as an act of officialimportance, but an act of sociability. To understand whether technology designedto support these “favours” has the potential to impact the existing socializationpractices, a medium-fidelity prototype was designed and deployed at one of thewings of C.A.I.S. In particular, we measured how different visualizations of indi-viduals and resolved tasks impacted reciprocity, density and centrality. In carryingout this experiment, we hoped to find that different visualizations would lead tovarying ways of using the technology, which would affect reciprocity, density andcentrality.We used the Sociality, Publicity and Physicality framework to inform and di-rect our design decisions. Regarding Sociality, in order to implement two typesof information visualization we used an indexical representation, or abstraction,of people as circles and inter-personal relationships as lines between those circles.We were interested in understanding how differences in this abstraction could af-995.3. Ticketing System Designfect community reciprocity. In particular we wanted to understand if representinginterpersonal interactions between people could have an effect on how people in-teracted with each other and reciprocated task completion. Regarding Publicity,a design decision was taken to allow tasks to be composed of free text, in orderto be able to analyze tasks for content Publicity, particularly ownership, of ticketcontent. We aimed to understand how this change in ownership and other changesin authorship could affect inclusiveness in the community. Finally, regarding Phys-icality, we decided to use a situated display, the size of a kiosk, that could allowinformation to be visible from a couple of metres away. Similarly, we deployedthe technology in a room where people gathered every day to verbally commu-nicate with each about other daily tasks. With these decisions we aimed to gainan insight as to whether said physical aspects of reach and scale would encourageaccessibility and familiarity of the information being presented.5.3.1 Prototype ComponentsWe followed a user centred iterative process to design a task-managing kioskand compared two visualizations of individuals and resolved tasks. The kiosk con-sisted of a 17-inch display with a mini computer and a wireless access point at-tached to the back of the display. A Portuguese QWERTY keyboard and mousewere provided as peripherals to input text and provide point-and-click interaction(Figure 5.5-A, 5.5-B). The software consisted of a web form through which userscould submit a message up to 250 characters in length, a drop-down list of all em-ployees of the care unit, a drop-down list of previously submitted and unresolvedtasks, and a submit button (Figure 5.5-C). A list of unresolved (Figure 5.5-D) andresolved (Figure 5.5-E) tasks was shown chronologically on the right side of thescreen. A visualization (Figure 5.5-F) of people and resolved tasks was presentedas the main focus of the application. The interaction flow consisted of users select-ing their name from the name drop-down list, and either selecting an unresolvedtask from the unresolved tasks drop-down menu to mark it as resolved, or typinga task to be added to the unresolved list of tasks. The kiosk was installed in thenurses’ room of the care unit, close to the paper-based log books used by profes-sionals to delegate and track important tasks. We chose to locate the technology in1005.3. Ticketing System Designthis room because participants already used this location to exchange informationat the beginning and at the end of their daily shift.Figure 5.5: Ticketing system final setup showing (A) a QWERTY keyboard, (B)a mouse, (C) input form, (D) unresolved tickets, (E) resolved tickets, and (F) acommunity-centric visualization.5.3.2 Visualizations UsedWe created two visualizations: a visualization of people as individuals (calledindividual-centric), and a visualization of people as part of a connected network(called community-centric).In the first visualization, individual-centric (Figure 5.6), each participant is rep-resented as a circle. The radius of each circle is computed using the number ofissues resolved by a person Ri. To account for any bias introduced by spatial po-sitioning, the positioning of each circle was randomized when the application wasinitialized. As people resolved more tasks, a circle labelled with their name in-1015.3. Ticketing System Designcreased in size. Professionals that resolved many tasks had larger circles than theirless active peers.Figure 5.6: Individual-centric visualization (with close ups) showing people (madeanonymous for the purposes of this diagram) as isolated circles.The second visualization, community-centric (Figure 5.7), focused on repre-senting the relationship of each person to other members of the group and thegroup as a whole. Each person was represented as a circle whose radius was com-puted as the ratio of issues resolved by a person, Ri, to all issues resolved by allthe members of the group, R, that is r = Ri/R. In this visualization, the sizeof a circle represented a person’s ranking, rather than their number of tickets re-solved. Therefore, changes in size would only be apparent if a professional solvedconsiderably more tickets than their peers. Circles were then places equidistantlyfrom each other in a circular arrangement, and lines were drawn to represent linkscreated between the person that created a ticket and the person that resolved it. Thewidth of a line connecting circle i and circle j was computed as the ratio of ticketsresolved by person i and issued by person j, that is Ri → j, to all tickets resolvedby person i, that is (Ri → j)/Ri. For example, if Anne submitted a ticket “checklight bulb in washroom” and John marked this ticket as “resolved”, the circle rep-1025.4. Methodologyresenting John would increase in size and a line between Anne and John wouldbe drawn. If all of Anne’s tasks were resolved by John, this line would be thickerthan if Anne’s tasks were resolved on average by many people. A legend was in-cluded on the bottom right part of the visualization explaining that larger circles“resolved more” and smaller circles “resolved less”, while thinner lines depicted“less cooperation” and thicker lines represented “more cooperation”.Figure 5.7: Community-centric visualization showing circles with lines depictingresolved tasks drawn between them. Boxes show details of pending tickets in pink,tickets resolved in grey and the input form.5.4 Methodology5.4.1 Recruitment and Participant OverviewThe kiosk was deployed over 8 weeks at C.A.I.S. Thirty three professionalswere selected from the children and young adults wing, which specializes in youngpatients, from newborns to 17 years of age. This decision was made because theprofessionals in this particular wing do not rotate as often as in other wings. We re-cruited 10 of these professionals for a post-deployment questionnaire, all of which1035.4. Methodologywere female, ranging from 24 to 51 years old. The reason for an all-female groupis that the wing where the prototype was deployed had no male staff assigned towork during our experiment. Within this group there were 2 nurses, 6 auxiliarynurses, 1 social therapist and 1 wing director. These 10 post-deployment partic-ipants were recruited when we visited the centre at the end of our deployment.Table 5.2 presents an overview of participants.N Male Female Age range Age Mean Age SD33 5 (15%) 28 (85%) 21-60 42.37 9.17Table 5.2: Participant overview.Figure 5.8: Only one professional (on the left) was trained on how to use the tech-nology. In this picture, she explains the kiosk purpose and functioning to a fellowworker.At the start of the experiment, we trained only one participant who subse-quently showed her peers how to use the ticketing system (Figure 5.8). Trainingthis participant took approximately 30 minutes and consisted of a researcher ex-plaining the purpose of the prototype as a way to submit “tasks” and to mark themas “done” and describing all parts of the prototype including the visualization area,input form, list of unresolved tickets and list of resolved tickets. We guided the1045.4. Methodologyparticipant through creating an example ticket and marking it as “done”. She wasthen asked to show the others how to use the system. This approach for trainingour participants was used because we wanted the adoption of the prototype to beas natural as possible, mitigating any bias that we may have introduced by trainingall the staff ourselves. Finally, it was made clear to all participants of our studythat the usage of the deployed technology was not a requirement. This allowed usto document the natural adoption, or rejection, of the introduced technology.5.4.2 Experimental DesignTo measure the effects of our ticketing system on the community and previ-ously documented work flows we performed interviews, applied questionnaires,measured group dynamics and evaluated the semantic content of the tickets sub-mitted. This allowed us to measure perception and adoption of the technology,usage patterns and meaning of information communicated. Furthermore, ticketssubmitted to the system during both visualizations were collected and analyzed us-ing network graph analysis and semantic text analysis. This allowed us to measurechanges in social capital, especially reciprocity. At the end of our 8-week study weconducted a post-deployment questionnaire to gain insights on how the communityperceived the introduction of the technology into their workplace. This was com-plemented with qualitative data gathered by interviewing some of the participants.After finalizing the study, the prototype was left as it was, without any type of in-tervention from us. Table 5.3 summarizes the instruments and measures adoptedfor the study. For the interviews and questionnaires, we report the anecdotal andsubjective results. Table 5.4 sumarizes the deployment schedule and treatmentsapplied. Figure 5.9 illustrates the prototype installed and being used in the nurses’room.Instrument MeasurementPost-Deployment Informal Interview Anecdotal perception of the systemPost-Deployment Questionnaire Likert scale of subjective impressionsTask issuing/resolving Reciprocity and graph densityTask text Exclusivity and awarenessTable 5.3: Instruments used for measuring work flow effect.1055.4. MethodologyWeek 1 to 4 Week 5 - 8All Participants Individual CommunalTable 5.4: Deployment schedule and treatments applied5.4.3 MeasuresNetwork Analysis of TasksThe network analysis of tasks is used to understand how people interacted witheach other during each visualization condition. In particular, how many connec-tions were created (graph density coefficient), to what degree people reciprocatedactions (graph reciprocity coefficient), and how much information people in thegroup had control over (centrality). In our prototype, a task is considered resolvedwhen a task that has been issued by one person is completed and marked as re-solved by another person. The collection of resolved tasks then forms a directedgraph where each person in the community is considered a node, and where re-solved tasks are considered directed links from the resolver of the task to the per-son that issued the task. This graph can then be analyzed for density (how manyconnections exist), reciprocity (to what degree people reciprocate actions) and cen-trality (how much information people have control over).Semantic Analysis of ContentWe used the Interactive Places Framework proposed by Memarovic et al. [186]to categorize content. The Interactive Places Framework models public displaycontent as information created and consumed by either individuals or groups ofpeople (i.e. with different levels of exclusivity) that when made available throughthe public display can be either explicit or implicit to a particular group of peopleor an individual (i.e. promoting different types of awareness).Awareness. Explicit communication portrays information that is valuable to par-ticular individuals within a community. Implicit communication is generalized in-formation not pertaining to a particular individual, but that it can apply to anybodyin the community. When analyzing the tasks, we considered explicit messages to1065.4. MethodologyFigure 5.9: The kiosk installed in the nurses’ room and being used by one nurse,while others perform daily activities.be exemplified by calls for action from either particular individuals or the commu-nity. For example “[John], water the plant”, “[MARTHA]: please fix the lamp inthe bathroom” or “TEAM don’t forget to stamp your reports”. By contrast, tasksconveying generalized states of the workplace or general messages of convivialitywere considered as implicit, for example “Have a good weekend” or “Emergencyroom light stopped working”.Exclusivity. When content is directed towards anybody that comes in contactwith the public display, said content can be understood as inclusive. When it islimited to a particular individual or group, it can be considered as exclusive. Whenanalyzing content we characterized exclusive content as being directed towards aparticular group, for example “[MARTHA]: please fix the lamp on the bathroom”or “[John], please take care of the garden”. We characterized inclusive content asgeneralized messages not directed towards a particular person or a group of people,but to the whole community, for example “Remember to close the door!”, “Patient[X] will not be receiving visitors this weekend.”).1075.5. Results5.5 ResultsOf the 33 people that registered to use the ticketing system, 21 actively used thesystem. There were a total of 159 messages produced in 8 weeks, with 109 of themsent during the individual-centric condition and 50 during the community-centriccondition. Most of these messages were related to important or day-scoped tasks,which explains the small amount of total messages in the two months the prototypewas deployed. During the individual-centric condition, 45/159 tasks (41%) weremarked as resolved by a person other than the one that created them (i.e, non-self-looping). During the community-centric condition, 37/50 tasks (74%) weremarked as resolved by someone other than themselves (Figure 5.10).5.5.1 Network AnalysisWe conducted social network analysis in order to measure how each visualiza-tion affected task reciprocity, density and centrality. When using the individual-centric visualization condition, participants would often resolve their own submit-ted tasks (51 self resolved, 45 non-self resolved), using the system as a logbook totrack their own personal work. We observed that in the individual-centric visualiza-tion the network of links resulting from task resolution had a higher density coeffi-cient of a¯ = 0.042 (when compared to a¯ = 0.035 in the community-centric condi-tion), as shown in Table 5.5.1. Similarly, the network of resulting tasks showed ahigher reciprocity coefficient (p = 0.210) during the individual-centric visualiza-tion than did the network of resulting tasks for the community-centric visualization(p = 0.131). This means that, on average, participants were more likely to interactwith a select number of individuals repeatedly during the individual-centric visu-alization than they were during the community-centric visualization.With the community-centric visualization condition the number of tasks com-pleted by their own issuer, that is using the system as a logbook for personal use,appears to diminish (11 self resolved vs 37 non-self resolved), as shown in Table5.5.1. As mentioned above, during this visualization the number of links result-ing from people resolving other people’s tasks was slightly less dense a¯ = 0.035and had a lower reciprocity coefficient p = 0.131 than the individual-centric vi-sualization. This means that people focused on interacting with a higher number1085.5. Results 0 10 20 30 40 50 60Week 1 Week 2 Week 3 Week 4Number of TicketsSubmittedResolvedResolved by othersResolved by selfA. Individual-centric Visualization 0 5 10 15 20 25 30Week 1 Week 2 Week 3 Week 4Number of TicketsSubmittedResolvedResolved by othersResolved by selfB. Community-centric VisualizationFigure 5.10: Total tasks submitted and resolved during 8 weeks of the proto-type. Figure (A) shows individual-centric visualization deployed on Week 1, for 4weeks. Figure (B) shows community-centric visualization deployed on week 4, for4 weeks.1095.5. Resultsof different people than during the individual-centric visualization, resulting in asparser network of task links exhibiting a lower reciprocity coefficient.During use of the community-centric visualization condition, the circle radiuswould only significantly change if a particular person solved more issues than theirpeers. This visualization, one participant noted, created confusion among the staffbecause when people submitted tasks “the circles [representing people] didn’t seemto change”. This, she argued, was discouraging for people that had previously fo-cused on “making their circles bigger” during the first visualization (individual-centric). Moreover, as one participant commented, professionals got “offended”with the legend added on the community-centric visualization. She recalled thatsome people stopped using the system because the circles representing them hadthin lines to other people, depicting “less cooperation”. When we looked at thenumber of people that used the system we found that 20% of the professionalsstopped using the system just after the community-centric visualization was in-stalled.Finally, we compared our reciprocity coefficients (individual p = 0.21, com-munal p = 0.131), by using a Fisher transform and using the z-score to comparefor significance. We found that differences in reciprocity observed were not statis-tically significant (z = 0.35 and significance p = 0.725).Visualization a¯ p σp pminPerfect reciprocal ... 1 ... − a¯1−a¯Individual-centric 0.042 0.210 0.122 -0.044Community-centric 0.035 0.131 0.118 -0.036Areciprocal ... 0 ... − a¯1−a¯Perfect areciprocal ... −1 ... −1Table 5.5: Analysis of density and reciprocity coefficients of the tasks issued andresolved during each one of the visualizations. For comparison, values are: fullyreciprocal (p = 1), areciprocal (p = 0) and fully antireciprocal (p = −1). Weprovide the standard deviation (σp) when any link from the network is removed,and minimum expected reciprocity coefficient (pmin) if there existed no reciprocallinks.We compare actor degree and betweenness centrality of all nodes (N = 33).To analyze for gregariousness, the graph used was analyzed using outdegree links.1105.5. ResultsIn our case an outlinkA→ B was created when a node “A” resolved a ticket issuedby “B”.Individual-Centric Community-CentricMeasure µ σ µ σ Sig. pDegree 0.0472 0.0506 0.0303 0.0549 0.0455Betweenness 0.0114 0.0261 0.0013 0.0053 0.0128Table 5.6: Analysis of centrality of the tasks issued and resolved during each oneof the visualizations. We compare the mean of all people’s actor centrality valuesusing a t-testWhen using the individual-centric visualization, the community seemed tohave an average higher degree centrality µ = 0.0472 that depicted a networkwhere several nodes resolved issues from a wide variety of people. We observedthat this effect was reduced on the community-centric visualization, where de-gree centrality was lower µ = 0.0303. This difference was statistically significant(t(32) = 2.08, p = 0.0455). During the individual-centric visualization a selectgroup of people seemed to have resolved tasks with as many people as possible inorder to increase the size of their circle. This may have been the reason that reci-procity and density were slightly higher during the individual-centric visualization.However, during the community-centric visualization, this select group of peoplewho had high centrality ratios became smaller (Table 5.5.1)We observed that during the individual-centric, visualization betweenness cen-trality µ = 0.0114 was higher than when the community-centric visualizationwas deployed µ = 0.0013 (Table 5.5.1). This change was statistically signifi-cant (t(32) = 2.63, p = 0.0128). That is, there is convincing evidence that duringthe individual-centric visualization a few individuals were connected to many peo-ple, that is holding positions “in-between” people by having exclusive control overtask information and completion. However, during the community-centric visu-alization, this group of people lost their “in between” position as people beganinteracting with each other directly. This resulted in a less dense network.1115.5. Results5.5.2 Semantic AnalysisA semantic analysis of tasks was carried out in order to measure changes inaccessibility and reach of information contained in the tasks (Table 5.5.2). Threepercent (3%) of tasks couldn’t be categorized as they were empty, possibly due toinput errors, like pressing the submit button before filling out the form. We ob-served that during the individual-centric visualization, the majority of tasks wereexclusive (57% of tasks), addressing particular professionals or groups of people,and often contained the capitalized name of the addressee in the body of the task,for example, “MARTHA please take care of my baby”. When the community-centric visualization was deployed, the number of inclusive tasks, i.e. tasks notaddressed to a particular individual and concerning anybody that came into contactwith the display, increased by 2%, while tasks addressed to a particular individualdecreased by 4%. Statements such as, “Have a good day everyone” were catego-rized as inclusive. We found that during the individual-centric visualization condi-tion, 52% of the submitted tasks were specific to a particular goal, person, or groupof people (e.g. “Please stamp the examination request from JACK.” and “ALEXplease install two dispensers of antiseptic gel in the corridor.”). Tasks submittedduring this visualization often called for specific actions from staff. By contrast,during the deployment of the community-centric visualization the number of im-plicit tasks depicting a status of the workplace (e.g. “Remember that today is astatutory day and we will have a party for the kids”) increased 7% and accountedfor 55% of messages submitted during this visualization.Ticket Content TypeExclusivity AwarenessVisualization Exclusive Inclusive Explicit ImplicitIndividual-centric 57% 43% 52% 48%Community-centric 53% 45% 42% 55%Table 5.7: Summary of text analysis of tickets.1125.5. Results5.5.3 Questionnaire and InterviewsOur post-deployment questionnaire was applied to 10 staff members (Table5.11). 80% of participants responded positively that the system encouraged themto talk to other professionals to resolve issues. About half of the participants agreedthat the technology made them help other people that they were not previously inthe habit of helping. Moreover, most participants agreed that the system helpedthem resolve conflicts with other professionals. However, there were mixed feel-ings about whether respondents felt more connected to other professionals as aresult of using the system.10%20%30%40%50%60%70%80%90%100%Q1 Q2 Q3 Q4 Q5 Q6Percent of ResponsesQ1 - My routine changed after using the system.Q2 - System encouraged me to talk to other professionals to resolve issues.Q3 - Technology made me help other people I previously didn't help.Q4 - The system helped me resolve previous conflicts.Q5 - The system created new conflicts between people.Q6 - I feel more connected to other professionals after using the system.Strongly AgreeAgreeAgree somewhatNeutralDisagreeDisagree somewhatStrongly DisagreeFigure 5.11: Post-deployment questionnaire results.We interviewed 7 professionals after the system had been deployed for 8 weeks.1135.6. DiscussionInterviewees had divided opinions about their experience using the kiosk. Three(3) participants commented that the system aided them in managing daily tasks andcommunicating these tasks to other health professionals. One participant indicatedthat “[the system] let [her] write messages that [she] knew people were going toread”, while another said “I really like it, I read it every morning to know whatpeople are up to and what needs to be done.” We also encountered a sense of dis-trust towards the technology. One participant commented that “[she didn’t] thinkeveryone liked the idea of the circles. They didn’t like their names on the display”.When asked why, she argued that “they [were] afraid of making [spelling] mis-takes or breaking the computer”. One participant confided to us that “people [inC.A.I.S.] don’t like technology. Especially [in this wing]. They don’t trust thingsthat come from the management.” There were two (2) people, out of seven (7),that were indifferent or uninterested in using the kiosk. One professional, whichwe had previously identified as having a lot of difficulty using the keyboard com-mented that “[I] didn’t like it, it was hard to use”. When asked why she added“I don’t like computers”. Another professional simply commented “I didn’t use it[shrug]” and gave no further explanation.5.6 DiscussionIn this study we investigated the impact of information visualization on a singleaspect of Third-placeness, namely interpersonal socialization. Through an ethno-graphic study, which took the form of shadowing, we documented that aspectsof Third-placeness, like promoting a low profile and neutral ground, regularity,inclusiveness and conversation were promoted by workplace dynamics like socia-bility, spontaneity and story telling. We discovered that, in this community, Third-placeness already plays a key role in supporting efficient work flows and informa-tion sharing. To investigate how the introduction of technology may be able to en-hance Third-placeness, we introduced a task ticketing system designed to supportfavours and implemented two different visualizations in order to measure how reci-procity, density and centrality were impacted by each one. An individual-centricvisualization presented people as isolated individuals, while a community-centricvisualization showed connections between individuals.1145.6. Discussion5.6.1 SocialityFor Sociality, we measured observed differences in the level of reciprocity withthe use of the two different visualization methods. We found that each of our vi-sualizations led different types of social interactions and usage of the ticketingsystem. While an individual-centric visualization seemed to promote interactionwith direct peers, a community-centric visualization led to a desire to belong tothe community, or to avoid being perceived as not participating. The individual-centric visualization offered a clear representation of the quantity of resolved tasks.This seemed to have encouraged a sense of competition as people saw the circlesrepresenting them grow bigger as they used the system more often. As a result, pro-fessionals often interacted with the same group of people in order to increase theirvisual representation on the screen more easily. Tasks submitted during this visual-ization were more commonly addressed to particular individuals and were explicitin their content, often demanding particular actions. In contrast, the community-centric visualization reduced the ratio of reciprocal task resolutions and people in-teracted with each other more directly, reducing the number of people controllingtask resolution and information, possibly making task-resolution more efficient.We theorize, based on our qualitative observations, that the community-centric vi-sualization made people feel compelled to interact with different individuals in or-der to create new visual links to other people by resolving tasks issued by differentprofessionals.In our analysis of graph reciprocity and density, we found task-resolving ac-tions to be less reciprocal in the community-centric condition, but better spreadacross more people. When measuring the resulting network centrality, it was ob-served that both conditions resulted in similar networks with fewer people at thecentre of task resolution and issuing. However, when looking at betweenness cen-trality we observed that during a community-centric visualization the number ofpeople that controlled the flow of information within the community was reducedwhen compared to the individual-centric visualization. That is, a community-centric visualization encouraged people to interact with each other directly, pos-sibly to create new visualized connections where they did not previously exist.1155.6. Discussion5.6.2 PublicityRegarding the Publicity aspect measured – inclusiveness – our semantic anal-ysis of messages suggested that tasks submitted during the community-centric vi-sualization were often open-ended and implicit and directed towards the wholecommunity. We found that while an individual-centric visualization promoted thecreation of content that was mostly exclusive and explicit, i.e. directed towardsparticular individuals or groups, a community-centric visualization resulted in anincrease of inclusive and implicit messages that could be used and acted upon byanybody that came in contact with the display.Although the system was originally designed as a ticketing system to publicizetasks regarding the workplace, the professionals used it to post non-task messagesto other professionals (e.g. “[Doctor X] welcome back from your vacation”), en-couraging messages (e.g. “May we find strength to do our jobs today”), and openended social phrases (e.g. “Good day” or “Good weekend everyone”). In somecases we found traces of conversational posts (e.g. “I thank everyone for the sup-port and help. May god bless you. Kisses.” to which somebody replied “We arewith you [Martha]”). In total, 14% of all tasks during both deployments were non-task messages to other professionals.5.6.3 PhysicalityIn our initial ethnographic study [11] we noted how people gathered every dayin a room to document their daily routines in pen and paper log books while en-gaging in work-related and social conversation, which encouraged storytelling andsociability. We placed our prototype in this room and chose a situated display, thescale of which allowed visibility of its content throughout this room. We corrob-orated, in our informal conversations with participants, that our choice of displayand its placement in fact allowed people to be aware of the content of the displaywithout having to approach it. While this increased the accessibility to informationit also hindered participation from those people that felt their names were asso-ciated with a lack of participation, especially since a legend in the display readthat thinner lines between circles representing people meant “less cooperation”.We noted by speaking to participants after the study that many participants were1165.6. Discussionfamiliar with the technology, its purpose, and the information being displayed. Al-though the design of our study aimed to promote ease of use through the physicalscale and placement of the display we noted that this familiarity might have beenthe result of repetitive exposure to the technology, due to the fact that people hadto use the room on a daily basis, along with the pervasive use of storytelling inthe community, which would likely have facilitated inter-group teaching about thetechnology.5.6.4 ConclusionIn a previous study (Chapter 4) we learned that promoting certain behaviours,like trust, community attachment and community spirit, could help support theexperience of Third-placeness. However, understanding the connection betweencharacteristics of technology and these behaviours was a challenging endeavor ina Third Place, where external factors and constant change are almost inevitable.In the study presented above, we were able to control and document many of thefactors that introduced complexity in our previous study. Thus, we focused onmeasuring how one aspect of technology, namely how information is visualized,could affect a single aspect of Third-placeness, namely reciprocity. We conductedan ethnographic study and discovered that the study community leveraged Third-placeness to break through layers of bureaucracy and hierarchies to better conducttheir jobs. We learned that favours, or reciprocal actions, were key to this and setabout to introduce a technology to support this practice while allowing us to un-derstand how changes in said technology would affect people’s behaviours. Weintroduced a task ticketing system designed to support favours, consisting of adigital kiosk. We learned that differences in how such technology was designed,particularly how communication between participants was visualized, encourageddifferent ways of communicating that affected how often its members communi-cated with each other, how information was shared and who was included in theconversation. These differences had a measurable effect on the Social Capital typeof reciprocity within the community and in turn affected how people did favoursfor each other and the content and types of favours they requested. That is, thesechanges in the technology affected aspects of Sociality and Publicity within the1175.6. Discussioncommunity.118Chapter 6ConclusionThe goals of this work are to develop an updated concept of Third Place,“Third-placeness”, that reflects contemporary Third Places and technologies, andto document what characteristics of interactive public displays could support orencourage aspects of “Third-placeness” in real world scenarios. It is driven by thethesis that information technologies could support communication and socializa-tion practices that, in turn, support social, public and physical aspects of ThirdPlaces – collectively referred to as “Third-placeness’.We first distilled current knowledge of Third Place to develop the concept of“Third-placeness” and gathered knowledge about a particular type of informationand communication technology, interactive public displays, to outline characteris-tics of these technologies that could support aspects of Third-placeness. We out-lined a conceptual framework, the Sociality, Publicity and Physicality Framework,that gathers key characteristics of Third-placeness and characteristics of interac-tive public displays that could support them. We used this framework to guide thedesign and evaluation of two different interactive public display prototypes, intro-duced into two different communities. A first study looked at documenting if saidtechnologies could support behaviours that encourage aspects of Third-placenessin a traditional Third Place. A second study looked at piecing together how oneaspect of interactive public displays, information visualization, would affect inter-personal interaction between people, specifically reciprocity.The research yielded an updated view of third places called “Third-placeness”,a conceptual framework, the Sociality, Publicity and Physicality Framework, andlessons learned regarding characteristics of interactive public displays that couldsupport “Third-placeness”. This chapter summarizes our contributions and dis-cusses future research directions.1196.1. Contributions6.1 ContributionsThe primary contributions of this research are (1) an updated view of “ThirdPlace” relevant to contemporary places and technologies called third-placeness,(2) observational studies of the impact of interactive public displays on differentaspects of third-placeness, and (3) the articulation of key characteristics of interac-tive public displays to support third-placeness resulting in the Sociality, Publicity,Physicality Framework.6.1.1 Updated view of “Third Place” relevant to contemporaryplaces and technologies called third-placenessThird Places are defined by Oldenburg as physical locations between the home(first) and the workplace (second) place. These places provide a space for play-ful conversation between strangers and regulars that allow people to experiencea public sphere that is essential to social, cultural and political well-being. Theconcept of Third Place was coined at a time when information and communica-tion technologies had not yet had a strong influence in society. Today, however,technologies play many roles in the functioning of these places. We argue that theconcept of Third Place can exist independently of the architectural and geographi-cal constraints originally proposed by Oldenburg. In order to articulate the conceptof “Third-placeness”, we analyzed characteristics of contemporary Third Placesand some technologies commonly found in them. We defined Third-placeness asa state of socialization that allows for communication and interpersonal interac-tions that support the creation and maintenance of social capital. We organizeaspects that support Third-placeness in three domains: “sociality”, “publicity” and“physicality”. The sociality dimension gathers aspects of interpersonal socializa-tion through the concept of social capital, particularly reciprocity. The publicitydimension gathers aspects of governance, particularly information ownership, ac-cessibility and privacy. The physicality dimension gathers aspects that relate tophysical aspects of technology and places. Because Third-placeness arises as partof socialization, we hypothesize that it can arise in places and contexts not previ-ously thought of by Oldenburg.1206.1. Contributions6.1.2 Observational studies of the impact of interactive publicdisplays on different aspects of third-placenessWe document two observational studies of introducing interactive public dis-plays into two different communities to measure how technology affects differentaspects of Third-placeness. For sociality aspects we measure interpersonal social-ization, particularly reciprocity, and use social capital and social network analysisas methodologies. For publicity aspects we measure governance aspects of equalityand accessibility and use questionnaires and ethnographic observations. Regardingphysicality, we document how perceptual qualities of technology affect adoption –like regularity and ligheartedness – through ethnographic observations.Supporting aspects of third-placeness with a table-top garden and private so-cial network at a student-run coffee shop. We introduced a table-top gardenand private social network into a student-run coffee shop. The study investigatedthe feasibility of inducing behaviours that could lead to aspects of Third-placenesswith the use of interactive public displays. The technology was designed to supportthe creation of social capital by encouraging people to interact reciprocally to takecare of a table-top garden. We measured social network structure, social capital,aspects of Third-placeness, and adoption patterns. We learned that differences ingameplay (individual versus communal) had no significant effect on how peopleinteracted with each other. However, we learned that people who used the tech-nology experienced aspects of Third-placeness. We also observed that people whoused the technology had higher self-reported social capital than those who did notparticipate in our trial. Finally, we learned that the technology provided newcomerswith a tool through which to participate in the community more easily.Measuring the impact of information visualization on community reciprocitywith a task-ticketing system at a workplace. We introduced a task ticketingsystem into a workplace. The study investigated the impact of information visual-ization on interpersonal socialization, particularly reciprocity. Two visualizationswere tested. An individual-centric visualization presented people as isolated in-dividuals, while a community-centric visualization showed frequency of commu-1216.1. Contributionsnication between members of the group. We measured reciprocity, meaning oftasks, adoption patterns and aspects of Third-placeness. We found that each vi-sualization promoted different ways of using the ticketing system. We also foundthat while an individual-centric visualization promoted competition with peers, acommunity-centric visualization promoted the desire to belong to the group, possi-bly to avoid being perceived as not participating. Specifically, an individual-centricvisualization supported a network characterized by few people controlling task res-olution and task information, while a community-centric visualization resulted infewer reciprocated interactions but more connections across the whole community.Regarding ticket content, an individual-centric visualization promoted messagestargeted towards particular individuals or groups and demanding specific actions,while a community-centric visualization supported more open-ended and inclusivecontent. Finally, through ethnographic observations, we discovered that aspects ofThird-placeness – particularly promoting a low profile and neutral ground, regular-ity, inclusiveness and conversation – were promoted by workplace dynamics likesociability, spontaneity and story telling.6.1.3 Articulation of key characteristics of interactive public displaysto support third-placeness resulting in the Sociality, Publicity,Physicality FrameworkWe reviewed characteristics of interactive public displays that we thought hadthe potential to support different aspects of Third-placeness. Specifically, we re-viewed geometric and perceptual characteristics of displays, characteristics of dis-plays that support publicity, and characteristics of displays that are considered in-teractive. We organized these characteristics in terms of how they could potentiallysupport different aspects of Third-placeness, to develop the Sociality, Publicity andPhysicality Framework. From the starting point of aspects of Third-placeness, wedeveloped a set of three design dimensions, each one relating to one of the threeaspects of Third-placeness previously identified. Each dimension, or key prob-lem area, outlines characteristics of interactive public displays that could supportrelated aspects of Third-placeness. These dimensions are: Organizational, Situ-ational and Physical. The “organizational” dimension collects characteristics of1226.2. Initial Design Guidelinestechnology and aspects of Third-placeness that relate to social interaction, specif-ically how people create social capital and how technology enables this throughinteractivity. The “situational” dimension collects characteristics of technologyand aspects of Third-placeness that relate to the experience of a public sphere char-acterized by familiarity, playfulness and conversation, specifically how technologycan support it through rules of ownership, accessibility and privacy. The “phys-ical” dimension collects characteristics of interactive public displays and aspectsof Third-placeness that relate to the physical experience of socialization when inpublic, particularly how technology is used and adopted in the experience of ThirdPlace. Finally, we presented a diagram, the “Sociality, Publicity and PhysicalityDesign Circle”, that can be used as a visual representation of the information pre-sented through the framework, in order aid with the design of interactive publicdisplays for Third-placeness.6.2 Initial Design GuidelinesFollowing is a proposed initial design guidelines derived from lessons learnedand discussed earlier in this dissertation. Designers can use these initial recom-mendations to inform their approach to designing technology for Third-placeness.The guidelines fall into three categories, informed by the Sociality, Publicity andPhysicality Framework: those that relate to the structural and normative aspects ofsocial links that enable inter-personal socialization (Sociality); those that relate toaspects of governance, which enforce the type of social interaction that character-ize the experience of Third-placeness (Publicity); and those that relate to physicalaspects of technology and places that support the experience of Third-placeness.6.2.1 SocialityUse Iconic information representation to minimize misinterpretation. Abstract(symbolic and indexical) content can result in varying interpretations of informa-tion that could hinder interpersonal interaction between members of a community.Leverage visualizations of information that resemble what they stand for (iconic)to avoid misinterpretation. For example, in our workplace study (Chapter 5) people1236.2. Initial Design Guidelinesgave varying meanings to lines and circles representing people and task resolution,this could have been possibly avoided by representing tasks more transparently.Information should reflect or complement existing practices of a community.By allowing information to embody activities of a community that support theexperience of Third-placeness, designers can provide tools to support such expe-rience. For example, designing the ticketing system introduced in our workplacestudy (Chapter 5) which followed the existing practices of favours allowed us tobetter communicate to the community the purpose and usage of the technology.Encourage group collaboration to promote interpersonal interactions. Encour-aging group collaboration can promote interpersonal interactions between mem-bers of a group that could support reciprocity and trust. Promoting individualactions, for example by promoting competition, could encourage individuals toact on community goals (social agency) out of self interest. In both our coffeeshop (Chapter 4) and workplace (Chapter 5) studies we documented how design-ing information and interaction as part of a collective effort could be beneficial tocreating a sense of community. In both cases we also documented that rewardingindividual actions encouraged people to act on these action out of self interest.6.2.2 PublicityUse filtered accessibility to support sense of belonging. Use open accessibilityto support newcomers. By restricting access to information to members of a com-munity, designers can promote a sense of exclusivity and belonging. When peoplesee themselves as part of a group they are more likely to have increased trust inthat group. However, unrestricted access to information can help newcomers to acommunity feel more welcomed to interact with the group. For example, in ourcoffee shop study (Chapter 4) the need to use mobile devices to use the technol-ogy supported a sense of exclusivity that seemed to have increased trust within thatgroup; by contrast the open accessibility of information on the large display al-lowed most members of the community to be aware of some public aspects of thisprivate community that we hypothesize, in some cases, encouraged newcomers toparticipate.Incorporate daily routines and existing practices of a community into your de-1246.2. Initial Design Guidelinessign to promote familiarity. Daily routines and activities of a community evidencethe unique characteristics of Third-placeness that members of such groups seek.Documenting and designing for these practices can help introduce technologiesthat are familiar, accessible and useful to a particular community. For example,in our workplace study (Chapter 5) an initial ethnographic study of the commu-nity allowed us to document daily routines and practices that helped us design aprototype that proved helpful for the majority of members of the community.Promote Third-placeness governance through ownership and accessibility ofinformation. Support a sense of community (inclusiveness) by supporting collec-tive ownership and open accessibility of information. Encourage rootedness byallowing collective ownership of information. Promote neutrality by leveraginganonymous ownership of information. For example, in the case of our workplacestudy (Chapter 5) we observed that by using a visualization that encouraged infor-mation to be perceived as part of a collective effort encouraged a greater sense ofrootedness and inclusiveness in content that was directed towards the community,rather than individuals.6.2.3 PhysicalityTake into account the physical aspects of technology (geometry, reach, per-ceptual qualities) when designing for Third-placeness. You can leverage physicalaspects of technology, for example the shape and orientation of displays, to supportdifferent types of information ownership and accessibility. Chapter 2 provides aninsight on several key physical characteristics of interactive public displays. Chap-ter 3 provides a mapping between key aspects of interactive public displays andaspects of Third-placeness.Your design should function both in the background and foreground of theThird-placeness experience. Regularity, one key aspect of Third-placeness, canresult in technology becoming part of the background of sociability. Design forthis background state, but allow your design to be able to become part of the fore-ground of Third-placeness when needed. For example, in both our coffee shopstudy (Chapter 4) and workplace study (Chapter 5) we observed that technologyfluctuated between being in the foreground and the background of social activ-1256.3. Limitationsities. In both studies the repetitive exposure to the technology accentuated thisbackground state.6.3 LimitationsThe studies presented in this dissertation are not intended to represent all situa-tions. We conducted these studies as a way to document whether or not behavioursof Third-placeness could be promoted through the use of interactive public dis-plays and to understand how one particular aspect of Third-placeness, namely reci-procity, is affected by changes in information representation enabled by said tech-nologies. Further research on other behaviours and aspects of Third-placeness arerequired.Furthermore, low sample sizes (29 for the coffee shop and 33 for the work-place) made statistical significance measurements challenging, while short deploy-ment time spans made it difficult to capture data and information on long-termadoption. This could potentially undermine the scientific validation of our find-ings. However, our research is not meant to have universal validity, but rather toprovide a focused understanding of the communities observed. Further researchwith larger communities and over longer periods of time are undoubtedly required.In our studies, we found it challenging to document the connections betweenthe introduced technologies and the observed behaviours of people that used them.In the coffee shop study (Chapter 4) the complex nature of a Third Place – like un-expected changes in community members and large amount of concurrent activi-ties – introduced a high level of complexity and external factors that made drawinga connection between treatments and measurements difficult. While our secondstudy attempted to control for many variables – like changes in community andactivities performed by the community – it still fell short in controlling for exter-nal factors between technology characteristics and aspects of Third-placeness. Abetter experimental design, for example with counterbalanced deployments, couldhelp tackle this limitation.Finally, this research was not intended to validate the secondary contributionof the Sociality, Publicity, Physicality Framework. A more thorough evaluation ofthe framework with design professionals is required.1266.4. Directions for Future Work6.4 Directions for Future WorkWe have presented an approach to support the architectural and geographi-cally independent experience of “Third-placeness” with technology through theconcept of “Third-placenes”. We have conducted two studies; one that documentsthe promotion of behaviours that could support aspects of Third-placeness throughthe use of technology; one to measure how information visualization affects theThird-placeness aspect of reciprocity. This is only an initial step towards betterunderstanding the roles that technology can play in supporting the experience ofThird Place. In this section we present directions for future work.Measurement of other aspects of Third-placeness. We have measured howrepresenting and visualizing interpersonal interaction using interactive public dis-plays affects reciprocity. The exploration of other aspects of Third-placeness areneeded. A good start would be to look at other aspects of Third-placeness pre-sented in chapter 3, for example how differences in information synchronicity im-pact bridging and bonding connections in a community, or how information own-ership affects conversation.Case studies with other communities. This research studied a Canadian com-munity of University students and a Brazilian team of professionals. Studies withother types of communities, cultures and settings are required. Studies of how dif-ferent communities shape and experience Third-placeness can have the potentialof a great contribution to the understanding Third-placeness.Case studies with larger communities and longer time spans. Our approachneeds to be applied to larger communities making it easier to compare statisticaldifferences between test conditions. Similarly, time spans of deployment greaterthan four months are needed to better document how such communities adoptand adapt the technologies over long periods of time. Understanding how Third-placeness evolves over time in a community and how technology can support thisevolution is important.1276.5. Lessons LearnedValidation of the Sociality, Publicity and Physicality Framework. We devel-oped the Sociality, Publicity, Physicality Framework as a tool to design and com-municate characteristics of technology in relation to aspects of Third-placeness.Future work is needed to evaluate whether or not the aspects proposed apply toother communities than those discussed here.Improvement of methodologies used to capture changes in social capital andcommunication. We believe that the methodologies used in our studies could beimproved. We found the ethnography and social network analysis methodologiesused here initially informative. However, we felt that our observations could havebenefited from a finer-grained understanding of social capital changes and more de-tailed measures capturing changes in communication patterns. A good start wouldbe to test new social network analysis measures like clusterability, clique analysis,cohesiveness and link reachability.6.5 Lessons LearnedWe observed that differences in the design of technology can impact how com-munities experience Third-placeness. In our studies we observed that even smallchanges in the design of interactive public displays can have implications on howpeople use these technologies. For example, variations in the way in which inter-personal interaction is visualized can result in people communicating differently.These changes in usage can have an impact on how people experience Third-placeness or build social capital.We also learned that the experience of Third-placeness is not one that can begeneralized across communities. Technology can provide tools that communitiescan use in order to experience Third-placeness. Meeting the expectations of allmembers of a community is difficult, because communities are formed of varyingindividuals and in constant change. However, technology can provide tools thatindividuals or sub-groups of a community can leverage to experience aspects ofThird-placeness that are important to them. For example lowering barriers to in-formation for newcomers or providing asynchronous communication channels fordistributed groups.1286.5. Lessons LearnedIn the following subsections we use the dimensions of “Third-placeness” pro-posed in Chapter 3 to summarize our findings in more detail. The section Social-ity Aspects gathers lessons learned related to how socialization occurs, and reportsmeasurements of social capital gathered through questionnaires and social networkanalysis. The section Publicity Aspects compiles lessons learned regarding therules that promote a type of socialization often found in Third Places, and reportsmeasurements of governance obtained using questionnaires and ethnographic ob-servations. Finally, the Physicality Aspects section presents lessons learned regard-ing physical aspects of technology that can promote the experience of Third-Place,and reports ethnographic observations.Sociality Aspects. In the coffee shop study (Chapter 4) we learned that technol-ogy could promote behaviours that can lead to the creation and maintenance ofsocial capital. We observed that people who used the technology had a strongerself-reported sense of social capital (agency, trust, community spirit, place attach-ment and reciprocity) than patrons who did not use the technology. Social net-work analysis of density, reciprocity and centrality showed that game design dif-ferences had no effect on how people used the technology. However, qualitativedata suggests that an individual game design could support social agency out ofself-interest, while a communal game design could lead to collaboration and com-munity awareness. The workplace study (Chapter 5) looked at how one aspect ofsociality (reciprocity) is affected by differences in information visualization in in-teractive public displays. We observed that representing people as individuals pro-moted competition, prompting people to interact repetitively with a small group ofkey people. In this case, those that reciprocated actions the most gained positionsof information brokers within the group. By contrast, visualizing the connectionsbetween people led to participants seeking to interact with as many people as pos-sible to create new links. This reduced overall reciprocity in the community, butpromoted a more widespread flow of information across the community. Thesefindings corroborated that changes to the way in which information is representedcan have an effect on how people socialize and create social capital, thus affectinghow people experience Third-placeness in their community.1296.5. Lessons LearnedPublicity Aspects. The design of the technology introduced in the coffee shopstudy (Chapter 4) proved to have increased aspects of Third-placeness (being com-munal, filled with regulars, welcoming, and a leveller) in the perception of thepeople that used it. Moreover, people felt a sense of belonging to the virtual com-munity, and a will to help other users of the game if needed. In the workplace study(Chapter 5) we observed that how interpersonal socialization is visualized also hasthe potential to affect aspects of accessibility and equality. We documented thatvisualizing people as individuals promoted the creation of messages directed to-wards individuals or particular groups, likely due to people aiming to increase theirindividual status. By contrast, a visualization of participants in the context of inter-personal connections supported content that was more general and could be actedupon by anyone, likely due to people looking to interact with as many people aspossible to increase their social capital in this way. Additional to these findings wedocumented that workplace dynamics like sociability, spontaneity and storytellingpromoted moments of Third-placeness characterized by a low profile, neutrality,regularity, inclusiveness and conversation. While the setting of this study was spe-cial in many ways, imbued with a cultural sociability common in Brazilian culture,we hypothesize that other settings where social interaction occurs could supportaspects of Third-placeness in a similar manner.Physicality Aspects. In the coffee shop study (Chapter 4) we observed that de-signing technology as furniture that is routinely found in a coffee shop promotedinitial adoption. However, we learned that Third Place “regulars” quickly becamefamiliar with the introduced technologies and stopped using it. This suggests thattechnology should play a supporting role in , rather than central to, the experi-ence of Third-placeness, rather than attempting to be the main provider of it. Inthe workplace study (Chapter 5) we observed that publicity aspects of informationare affected by the inherent physical characteristics of interactive public displays.These can affect how people share information and change how people experienceThird-placeness. For example, we documented that the public display promotedpublic awareness of people’s actions, which supported people interacting with oth-ers they would not have otherwise interacted with, as well as resolving conflictswith their peers. However, this public awareness also promoted fear of embarrass-1306.6. Concluding Commentsment and anxiety of being perceived as not collaborating enough in the community.6.6 Concluding CommentsToday, information and communication technologies have a growing numberof roles in determining the ways in which we socialize. Our research has takeninitial steps towards building a deeper understanding of ways in which to under-stand and support a type of socialization that Oldenburg once observed in “greatgood places”. In this dissertation we took a closer look at a particular type ofsocialization, of which Third Place is a physical embodiment, and how technolo-gies can support it. The research contributes (1) an updated view of “Third Place”relevant to contemporary places and technologies called third-placeness, (2) obser-vational studies of the impact of interactive public displays on different aspects ofthird-placeness, and (3) the articulation of key characteristics of interactive publicdisplays to support third-placeness resulting in the Sociality, Publicity, PhysicalityFramework. We hope our work serves as a beacon and roadmap to help future re-search better design technologies to support social experiences that that ultimatelylead to stronger communities.131Bibliography[1] R. Calderon, S. Fels, and J. Anacleto, “Workshop on human computer in-teraction in third places,” in CHI ’13 Extended abstracts on Human factorsin computing systems, CHI EA ’13, (New York, NY, USA), pp. 3167–3170,ACM, 2013.[2] R. Calderon, S. Fels, J. Anacleto, N. Memarovic, and T. Thompson, “Hack-ing HCI3P: second workshop on human computer interaction in thirdplaces,” in Proceedings of the 2014 companion publication on Designinginteractive systems, DIS Companion ’14, (New York, NY, USA), pp. 195–198, ACM, 2014.[3] N. Memarovic, S. Fels, J. Anacleto, R. Calderon, F. Gobbo, and J. M. Car-roll, “Rethinking third places: contemporary design with technology,” Thejournal of community informatics, vol. 10, no. 3, 2014.[4] IWCSC ’10: first interdisciplinary workshop on Communication for sustain-able communities, SIGDOC 2010, (New York, NY, USA), ACM, 2010.[5] A. Bueno, J. C. Anacleto, R. Calderon, S. Fels, and R. Lea, “ICT to supportcommunity gardening: a system to help people to connect to each other inreal life,” in Proceedings of the 2014 companion publication on Designinginteractive systems, DIS Companion ’14, (New York, NY, USA), pp. 133–136, ACM, 2014.[6] R. A. Perez de Almeida, M. Blackstock, R. Lea, R. Calderon, A. F. do Prado,and H. C. Guardia, “Thing Broker: a Twitter for things,” in Proceedingsof the 2013 ACM conference on Pervasive and ubiquitous computing, Ubi-Comp ’13, (New York, NY, USA), pp. 1545–1554, ACM, 2013.132Bibliography[7] R. Calderon, M. Blackstock, R. Lea, S. Fels, A. Bueno, and J. Anacleto,“RED: A framework for prototyping multi-display applications using webtechnologies,” in Proceedings of the international symposium on Pervasivedisplays, PerDis ’14, (New York, NY, USA), pp. 148–153, ACM, 2014.[8] R. Calderon, S. Fels, R. Lea, M. Alexandre, and J. Anacleto, “Selfish VS al-truistic co-located play and their effect on Third Place trust,” in Proceedingsof the GRAND AGM, In Press, 2012.[9] R. Calderon, S. Fels, R. Lea, and O. Neumann, “Harvesting communal inter-action with public displays through place-dependent community displays,”in Workshop on Large displays in urban life. CHI2011 Workshop, 2011.[10] R. Calderon, S. Fels, and J. Anacleto, “Towards understanding the effectsof public displays on the functioning of a Brazilian health-care centre,” inProceedings of INTERACT 2013, 2013.[11] R. Calderon, S. Fels, J. L. de Oliveira, and J. Anacleto, “UnderstandingNUI-supported nomadic social places in a Brazilian health care facility,” inProceedings of the 11th Brazilian symposium on Human factors in comput-ing systems, IHC ’12, pp. 76–84, Porto Alegre, Brazil: Brazilian ComputerSociety, 2012.[12] R. Calderon, S. Fels, J. Anacleto, and J. Leite de Oliveira, “Towards support-ing informal information and communication practices within a Brazilianhealthcare environment,” in CHI ’13 Extended abstracts on Human factorsin computing systems, CHI EA ’13, (New York, NY, USA), pp. 517–522,ACM, 2013.[13] R. Calderon, M. Blackstock, R. Lea, S. Fels, A. Bueno, and J. Anacleto,“Supporting conversation and community interaction with a table-top com-munity garden application,” in Proceedings of the international symposiumon Pervasive displays, PerDis ’14, (New York, NY, USA), pp. 190–191,ACM, 2014.[14] R. Calderon, R. Lea, M. Blackstock, and S. Fels, “Developing cross-displayapplications using the really easy displays (RED) framework,” in 2013 In-133Bibliographyternational symposium on Pervasive displays, (Mountain View, California),2013.[15] R. Oldenburg, The great good place: cafes, coffee shops, community cen-ters, beauty parlors, general stores, bars, hangouts, and how they get youthrough the day. Paragon House, New York, 1989.[16] R. D. Putnam, Bowling alone : the collapse and revival of American com-munity. Simon & Schuster, New York ; London :, 2000.[17] J. Habermas, Structural transformation of the public sphere: an inquiry intoa category of bourgeois society. MIT Press, 1989.[18] M. Weiser, “The computer for the 21st century,” Scientific American,vol. 253, no. 3, 1991.[19] I. Guy, M. Jacovi, A. Perer, I. Ronen, and E. Uziel, “Same places, samethings, same people? Mining user similarity on social media,” in CSCW’10: Proceedings of the 2010 ACM conference on Computer supported co-operative work, (New York, NY, USA), pp. 41–50, ACM, 2010.[20] M. Karam, “The coffee lab: developing a public usability space,” in CHIEA ’10: Proceedings of the 28th of the international conference extendedabstracts on Human factors in computing systems, (New York, NY, USA),pp. 2671–2680, ACM, 2010.[21] E. F. Churchill, L. Nelson, and L. Denoue, “Multimedia flyers: informationsharing with digital community bulletin boards,” pp. 97–117, 2003.[22] E. F. Churchill, L. Nelson, and G. Hsieh, “Cafe life in the digital age: aug-menting information flow in a cafe-work-entertainment space,” in CHI ’06extended abstracts on Human factors in computing systems, CHI EA ’06,(New York, NY, USA), pp. 123–128, ACM, 2006.[23] J. Yoon, J. Oishi, J. Nawyn, K. Kobayashi, and N. Gupta, “FishPong: en-couraging human-to-human interaction in informal social environments,” inProceedings of the 2004 ACM conference on Computer supported coopera-tive work, CSCW ’04, (New York, NY, USA), pp. 374–377, ACM, 2004.134Bibliography[24] S. Izadi, G. Fitzpatrick, T. Rodden, H. Brignull, Y. Rogers, and S. Lindley,“The iterative design and study of a large display for shared and sociablespaces,” in DUX ’05, (New York, NY, USA), p. 59, AIGA: American Insti-tute of Graphic Arts, 2005.[25] S. Hosio, H. Kukka, and J. Riekki, “Leveraging social networking servicesto encourage interaction in public spaces,” in Proceedings of the 7th inter-national conference on Mobile and ubiquitous multimedia, MUM ’08, (NewYork, NY, USA), pp. 2–7, ACM, 2008.[26] J. F. McCarthy, S. D. Farnham, Y. Patel, S. Ahuja, D. Norman, W. R. Ha-zlewood, and J. Lind, “Supporting community in third places with situatedsocial software,” in C&T ’09, (New York, NY, USA), pp. 225–234, ACM,2009.[27] N. Taylor and K. Cheverst, ““This might be stupid, but...”: participatorydesign with community displays and postcards,” in OZCHI ’08, (New York,NY, USA), pp. 41–48, ACM, 2008.[28] R. Oldenburg, Celebrating the third place: inspiring stories about the greatgood places at the heart of our communities. Marlowe & Co., New York,2001.[29] E. Relph, Place and Placelessness. Pion, 1976.[30] F. Lukermann, “Geography as a formal intellectual discipline and the wayin which it contributes to human knowledge,” in The Canadian geographer,vol. 8, pp. 167–172, 1964.[31] D. Seamon, “Place, place identity, and phenomenology: a triadic interpreta-tion based on J.G. Bennett’s systematics,” in The Role of Place Identity n thePerception, Understanding and Design of the Built Environemnt (H. Casakinand S. Porta, eds.), Betham Science Publishers, 2011.[32] D. Seamon and J. Sowers, “Place and placelessness,” in Human geography(P. Hubbard, R. Kitchen, and G. Vallentine, eds.), Sage, 2008.135Bibliography[33] D. Seamon, “Social capital in the creation of human capital,” Aether: thejournal of media geography, vol. 3, pp. 1–19, 2008.[34] J. Rutherford, Identity: community, culture, difference, ch. The third space:interview with Homi Bhabha, pp. 207–221. Lawrence and Wishart, 1990.[35] D. Sanghamitra, “Book reviews: communicating in the third space,”Transnational literature, vol. 3, no. 1, 2010.[36] J. Scheible and T. Ojala, “MobiLenin combining a multi-track music video,personal mobile phones and a public display into multi-user interactive en-tertainment,” in MULTIMEDIA ’05, (New York, NY, USA), pp. 199–208,ACM, 2005.[37] K. Chadios, “The urban coffee shop,” Master’s thesis, MIT, Department ofArchitecture, 2005.[38] M. Ellis, The coffee house: a cultural history. George Weidenfeld & Nichol-son, 2004.[39] A. Ellis, The penny universities: a history of the coffee-houses. Secker &Warburg, 1956.[40] E. Ward, The London spy. Folio Society, London, 1955.[41] R. Sennett, The fall of public man: on the social psychology of capitalism.Cambridge University Press, 1976.[42] M. Pendergrast, Uncommon grounds: the history of coffee and how it trans-formed our world. Basic Books, 2010.[43] H. Schultz, Pour your heart into it: how Starbucks built a company one cupat a time. Hyperion, 1997.[44] R. Gaudio, “Coffetalk: Starbucks and the commercialization of casual con-versation,” pp. 659–691, 2009.[45] M. Carmona, T. Heath, T. Oc, and S. Tiesdell, Public places: urban spaces.Architectural Press, 2003.136Bibliography[46] J. Fulcher, Capitalism. a very short introduction. Oxford University Press,2004.[47] L. J. Hanifan, “The rural school community center,” Annals of the Americanacademy of political and social science, vol. 67, 1916.[48] J. S. Coleman, “Social capital in the creation of human capital,” The Ameri-can journal of sociology, vol. 94, pp. 95–120, 1988.[49] “Oxford English Dictionary.” http://oed.com, 2011 (last visited 2016).[50] C. Cartiere and S. Willis, eds., The practice of public art. Routledge Re-search in Cultural and Media Studies, 2008.[51] E. Chalmin, M. Menu, and C. Vignaud, “Analysis of rock art painting andtechnology of Palaeolithic painters,” Measurement science and technology,vol. 14, no. 9, p. 1590, 2003.[52] H. Sampson, A history of advertising from the earliest times: illustrated byanecdotes, curius specimens and biographical notes. Chatto and Windus,1857.[53] F. E. Basten, Great American billboards: 100 years of history by the side ofthe road. Ten Speed Press, 2007.[54] W. H. Wilson, “The billboard: bane of the city beautiful,” August 1987.[55] M. Crowe, “Neon signs: their origin, use, and maintenance,” in Preservingwhat’s new, vol. 23, International association for preservation technology,1991.[56] A. McCarthy, Ambient television: visual culture and public space. DukeUniversity Press, 2001.[57] B. Bratton and N. Jeremijenko, “Suspicious images, latent interfaces,” inSituated technologies pamphlets: situated advocacy (T. S. Omar Khan andM. Shepard, eds.), Architectural league of New York, 2008.137Bibliography[58] H. Benko, A. D. Wilson, and R. Balakrishnan, “Sphere: multi-touch inter-actions on a spherical display,” in Proceedings of the 21st annual ACM sym-posium on User interface software and technology, UIST ’08, (New York,NY, USA), pp. 77–86, ACM, 2008.[59] G. Beyer, F. Alt, J. Mu¨ller, A. Schmidt, K. Isakovic, S. Klose, M. Schiewe,and I. Haulsen, “Audience behavior around large interactive cylindricalscreens,” in Proceedings of the 2011 annual conference on Human factors incomputing systems, CHI ’11, (New York, NY, USA), pp. 1021–1030, ACM,2011.[60] K. Swaminathan and S. Sato, “Interaction design for large displays,” Inter-actions, vol. 4, no. 1, pp. 15–24, 1997.[61] E. Huang, A. Koster, and J. Borchers, “Overcoming assumptions and un-covering practices: when does the public really look at public displays?,”in Pervasive computing (J. Indulska, D. Patterson, T. Rodden, and M. Ott,eds.), vol. 5013 of Lecture notes in computer science, pp. 228–243, SpringerBerlin / Heidelberg, 2008.[62] J. Borchers, O. Deussen, and C. Kno¨rzer, “Getting it across: layout issuesfor kiosk systems,” SIGCHI Bulletin, vol. 27, pp. 68–74, October 1995.[63] D. Vogel and R. Balakrishnan, “Interactive public ambient displays: transi-tioning from implicit to explicit, public to personal, interaction with multipleusers,” in Proceedings of the 17th annual ACM symposium on User interfacesoftware and technology, UIST ’04, (New York, NY, USA), pp. 137–146,ACM, 2004.[64] M. Finke, A. Tang, R. Leung, and M. Blackstock, “Lessons learned: gamedesign for large public displays,” in DIMEA ’08, (New York, NY, USA),pp. 26–33, ACM, 2008.[65] H. Brignull and Y. Rogers, “Enticing people to interact with large publicdisplays in public spaces,” in Proceedings of INTERACT’03, 2003.138Bibliography[66] M. Ziefle, “Effects of display resolution on visual performance,” Humanfactors, vol. 40, no. 4, pp. 554–568, 1998.[67] B. Shneiderman, “Response time and display rate in human performancewith computers,” ACM Computing surveys, vol. 16, no. 3, pp. 265–285,1984.[68] P. N. Day, J. Colville, and C. Rohan, “An evaluation of sunlight-viewabledisplays,” in Proceedings of the 24th BCS Interaction specialist group con-ference, BCS ’10, (Swinton, UK, UK), pp. 384–392, British Computer So-ciety, 2010.[69] H. Arendt, The human condition. University of Chicago Press, 1958.[70] P. Virilio, The vision machine. BFI, 1994.[71] L. S. Heon, “Putting the show on the road,” in Billboard art on the road(L. S.-H. P. Diggs and J. Thompson, eds.), MASS MoCA, 1999.[72] S. McQuire, The media city: media, architecture and urban space. Sage,2008.[73] “Television history: the first 75 years.” http://www.tvhistory.tv/tv-prices.htm, 2011 (last visited 2016).[74] “Barrooms with a view,” Time, March 1947.[75] “TV: a market in New York now,” Sponsor, vol. 49, 1947.[76] L. Spiegel, Make room for TV: television and the family ideal in postwarAmerica. University of Chicago Press, 1992.[77] A. Morrison, A. Oulasvirta, P. Peltonen, S. Lemmela, G. Jacucci, G. Reit-mayr, J. Na¨sa¨nen, and A. Juustila, “Like bees around the hive: a compara-tive study of a mobile augmented reality map,” in CHI ’09: Proceedings ofthe 27th international conference on Human factors in computing systems,(New York, NY, USA), pp. 1889–1898, ACM, 2009.139Bibliography[78] K. O’Hara, M. Glancy, and S. Robertshaw, “Understanding collective playin an urban screen game,” in Proceedings of the 2008 ACM conference onComputer supported cooperative work, CSCW ’08, (New York, NY, USA),pp. 67–76, ACM, 2008.[79] E. Eriksson, T. R. Hansen, and A. Lykke-Olesen, “Reclaiming public space:designing for public interaction with private devices,” in Proceedings of the1st international conference on Tangible and embedded interaction, TEI’07, (New York, NY, USA), pp. 31–38, ACM, 2007.[80] L. Terrenghi, A. Quigley, and A. Dix, “A taxonomy for and analysisof multi-person-display ecosystems,” Personal and ubiquitous computing,vol. 13, pp. 583–598, November 2009.[81] S. Harrison and P. Dourish, “Re-place-ing space: the roles of place and spacein collaborative systems,” in Proceedings of the 1996 ACM conference onComputer supported cooperative work, CSCW ’96, (New York, NY, USA),pp. 67–76, ACM, 1996.[82] J. Scheible, V. H. Tuulos, and T. Ojala, “Story Mashup: design and evalu-ation of novel interactive storytelling game for mobile and web users,” inProceedings of the 6th international conference on Mobile and ubiquitousmultimedia, MUM ’07, (New York, NY, USA), pp. 139–148, ACM, 2007.[83] T. Olsson, “Understanding collective content: purposes, characteristics andcollaborative practices,” in Proceedings of the fourth international confer-ence on Communities and technologies, C&T ’09, (New York, NY, USA),pp. 21–30, ACM, 2009.[84] J. Lehikoinen, A. Aaltonen, P. Huuskonen, and I. Salminen, Personal con-tent experience. John Wiley and Sons Ltd., 2000.[85] S. D. Farnham, J. F. McCarthy, Y. Patel, S. Ahuja, D. Norman, W. R. Ha-zlewood, and J. Lind, “Measuring the impact of third place attachment onthe adoption of a place-based community technology,” in Proceedings of the27th international conference on Human factors in computing systems, CHI’09, (New York, NY, USA), pp. 2153–2156, ACM, 2009.140Bibliography[86] D. Frohlich, A. Kuchinsky, C. Pering, A. Don, and S. Ariss, “Requirementsfor photoware,” in Proceedings of the 2002 ACM conference on Computersupported cooperative work, CSCW ’02, (New York, NY, USA), pp. 166–175, ACM, 2002.[87] G. Jacucci, A. Oulasvirta, A. Salovaara, and R. Sarvas, “Supporting theshared experience of spectators through mobile group media,” in Proceed-ings of the 2005 international ACM conference on Supporting group work,GROUP ’05, (New York, NY, USA), pp. 207–216, ACM, 2005.[88] A. Salovaara, G. Jacucci, A. Oulasvirta, T. Saari, P. Kanerva, E. Kurvinen,and S. Tiitta, “Collective creation and sense-making of mobile media,” inProceedings of the SIGCHI conference on Human Factors in computing sys-tems, CHI ’06, (New York, NY, USA), pp. 1211–1220, ACM, 2006.[89] N. Kaviani, M. Finke, S. Fels, R. Lea, and H. Wang, “What goes where?:designing interactive large public display applications for mobile device in-teraction,” in Proceedings of the first international conference on Internetmultimedia computing and service, ICIMCS ’09, (New York, NY, USA),pp. 129–138, ACM, 2009.[90] S. Hosio, H. Kukka, M. Jurmu, T. Ojala, and J. Riekki, “Enhancing interac-tive public displays with social networking services,” in Proceedings of the9th international conference on Mobile and ubiquitous multimedia, MUM’10, (New York, NY, USA), pp. 23:1–23:9, ACM, 2010.[91] L. Palen and P. Dourish, “Unpacking “privacy” for a networked world,” inProceedings of the SIGCHI conference on Human factors in computing sys-tems, CHI ’03, (New York, NY, USA), pp. 129–136, ACM, 2003.[92] H. Nissenbaum, “Protecting privacy in an information age: the problem ofprivacy in public,” Law and philosophy, vol. 17, pp. 559–596, 1998.[93] M. M. Mantei, R. M. Baecker, A. J. Sellen, W. A. S. Buxton, T. Milligan, andB. Wellman, “Experiences in the use of a media space,” in Proceedings ofthe SIGCHI conference on Human factors in computing systems: reaching141Bibliographythrough technology, CHI ’91, (New York, NY, USA), pp. 203–208, ACM,1991.[94] E. F. Churchill, L. Nelson, L. Denoue, J. Helfman, and P. Murphy, “Sharingmultimedia content with interactive public displays: a case study,” in Pro-ceedings of the 5th conference on Designing interactive systems: processes,practices, methods, and techniques, DIS ’04, (New York, NY, USA), pp. 7–16, ACM, 2004.[95] S. Hosio, H. Kukka, and J. Riekki, “Social surroundings: bridging the virtualand physical divide,” Multimedia, IEEE, vol. 17, no. 2, pp. 26 –33, 2010.[96] R. Jose´, H. Pinto, B. Silva, A. Melro, and H. Rodrigues, “Beyond interac-tion: tools and practices for situated publication in display networks,” inProceedings of the 2012 international symposium on Pervasive displays,PerDis ’12, (New York, NY, USA), pp. 8:1–8:6, ACM, 2012.[97] R. Xiong and J. Donath, “PeopleGarden: creating data portraits for users,” inProceedings of the 12th annual ACM symposium on User interface softwareand technology, UIST ’99, (New York, NY, USA), pp. 37–44, ACM, 1999.[98] L. A. Suchman, Plans and situated actions: the problem of human-machinecommunication. Cambridge University Press, 1987.[99] S. Rafaeli, “Interactivity: from new media to communication,” in Advanc-ing communication science: merging mass and interpersonal process (J. M.Wiemann and S. Pingree, eds.), pp. 110–134, Sage, 1988.[100] W. Holfelder and D. Hehmann, “A networked multimedia retrieval manage-ment system for distributed kiosk applications,” in Proceedings of the Inter-national conference on Multimedia computing and systems, pp. 342 –351,1994.[101] L. Webber, “Plasma displays,” in IEEE International conference on PlasmaScience, p. 140, 1996.[102] E. Johnson, “Touch displays, a novel input/output device for computers,”Electronics Letters, vol. 1, no. 8, pp. 219 –220, 1965.142Bibliography[103] S. Lee, W. Buxton, and K. C. Smith, “A multi-touch three dimensionaltouch-sensitive tablet,” SIGCHI Bulletin, vol. 16, pp. 21–25, April 1985.[104] P. Dietz and D. Leigh, “DiamondTouch: a multi-user touch technology,”in Proceedings of the 14th annual ACM symposium on User interface soft-ware and technology, UIST ’01, (New York, NY, USA), pp. 219–226, ACM,2001.[105] A. W. Burks, “Icon, index, and symbol,” Philosophy and phenomenologicalresearch, vol. 9, no. 4, pp. 673–689, 1949.[106] J. D. Johansen, Signs in use: an introduction to semiotics. Routledge, 2002.[107] S. A. Munson, E. Rosengren, and P. Resnick, “Thanks and tweets: compar-ing two public displays,” in Proceedings of the ACM 2011 conference onComputer supported cooperative work, CSCW ’11, (New York, NY, USA),pp. 331–340, ACM, 2011.[108] J. Holopainen, A. Lucero, H. Saarenpa¨a¨, T. Nummenmaa, A. El Ali, andT. Jokela, “Social and privacy aspects of a system for collaborative publicexpression,” in Proceedings of the 8th international conference on Advancesin computer entertainment technology, ACE ’11, (New York, NY, USA),pp. 23:1–23:8, ACM, 2011.[109] F. Cristian, “Synchronous and asynchronous,” ACM Communications,vol. 39, no. 4, pp. 88–97, 1996.[110] K. O’Hara, M. Lipson, M. Jansen, A. Unger, H. Jeffries, and P. Macer,“Jukola: democratic music choice in a public space,” in Proceedings ofthe 5th conference on Designing interactive systems: processes, practices,methods, and techniques, DIS ’04, (New York, NY, USA), pp. 145–154,ACM, 2004.[111] H. Brignull, S. Izadi, G. Fitzpatrick, Y. Rogers, and T. Rodden, “The intro-duction of a shared interactive surface into a communal space,” in Proceed-ings of the 2004 ACM conference on Computer supported cooperative work,CSCW ’04, (New York, NY, USA), pp. 49–58, ACM, 2004.143Bibliography[112] S. Reeves, S. Benford, C. O’Malley, and M. Fraser, “Designing the spectatorexperience,” in Proceedings of the SIGCHI conference on Human factors incomputing systems, CHI ’05, (New York, NY, USA), pp. 741–750, ACM,2005.[113] M. Ten Koppel, G. Bailly, J. Mu¨ller, and R. Walter, “Chained displays: con-figurations of public displays can be used to influence actor-, audience-, andpasser-by behavior,” in Proceedings of the 2012 ACM annual conferenceon Human factors in computing systems, CHI ’12, (New York, NY, USA),pp. 317–326, ACM, 2012.[114] M. Perry, S. Beckett, K. O’Hara, and S. Subramanian, “WaveWindow: pub-lic, performative gestural interaction,” in ACM international conference onInteractive tabletops and surfaces, ITS ’10, (New York, NY, USA), pp. 109–112, ACM, 2010.[115] J. Zlatev, “Embodiment, language, and mimesis,” in Body, language andmind: embodiment (T. Ziemke, J. Zlatev, and M. F. Roslyn, eds.), vol. 1,Mouton de Gruyter, 2007.[116] S. Fels, “Designing intimate experiences,” in IUI ’04, (New York, NY,USA), pp. 2–3, ACM, 2004.[117] W. Ryan, W. R. Hazlewood, and K. Makice, “Twitterspace: a co-developeddisplay using Twitter to enhance community awareness,” in Proceedings ofthe 10th anniversary conference on Participatory design, PDC ’08, (Indi-anapolis, IN, USA), pp. 230–233, Indiana University, 2008.[118] T. Varoudis, S. Dalton, K. Alexiou, and T. Zamenopoulos, “Ambient dis-plays: influencing movement patterns,” in Proceedings of the 2011 annualconference extended abstracts on Human factors in computing systems, CHIEA ’11, (New York, NY, USA), pp. 1225–1230, ACM, 2011.[119] Y. Rogers, W. R. Hazlewood, P. Marshall, N. Dalton, and S. Hertrich, “Am-bient influence: can twinkly lights lure and abstract representations trigger144Bibliographybehavioural change?,” in Proceedings of the 12th ACM international con-ference on Ubiquitous computing, Ubicomp ’10, (New York, NY, USA),pp. 261–270, ACM, 2010.[120] J. Mu¨ller, F. Alt, D. Michelis, and A. Schmidt, “Requirements and designspace for interactive public displays,” in Proceedings of the internationalconference on Multimedia, MM ’10, (New York, NY, USA), pp. 1285–1294,ACM, 2010.[121] N. Sambasivan, N. Rangaswamy, E. Cutrell, and B. Nardi, “Ubicomp4d: in-frastructure and interaction for international development: the case of urbanIndian slums,” in Ubicomp ’09, (New York, NY, USA), pp. 155–164, ACM,2009.[122] J. M. Carroll, “Community computing as human computer interaction,” Be-haviour & Information Technology, vol. 20, no. 10, pp. 307 – 314, 2001.[123] B. Kirman, S. Lawson, C. Linehan, F. Martino, L. Gamberini, andA. Gaggioli, “Improving social game engagement on Facebook through en-hanced socio-contextual information,” in CHI ’10, (New York, NY, USA),pp. 1753–1756, ACM, 2010.[124] L. Backstrom, E. Sun, and C. Marlow, “Find me if you can: improvinggeographical prediction with social and spatial proximity,” in WWW ’10:Proceedings of the 19th international conference on World wide web, (NewYork, NY, USA), pp. 61–70, ACM, 2010.[125] P. Peltonen, E. Kurvinen, A. Salovaara, G. Jacucci, T. Ilmonen, J. Evans,A. Oulasvirta, and P. Saarikko, “It’s mine, don’t touch! Interactions at alarge multi-touch display in a city centre,” in Proceedings of the twenty-sixth annual SIGCHI conference on Human factors in computing systems,CHI ’08, (New York, NY, USA), pp. 1285–1294, ACM, 2008.[126] J.-T. Kim, J.-H. Lee, H.-K. Lee, and E.-H. Paik, “Design and implementa-tion of the location-based personalized social media service,” in ICIW’2010,pp. 116 –121, 9-15 2010.145Bibliography[127] W. Griswold, P. Shanahan, S. Brown, R. Boyer, M. Ratto, R. Shapiro, andT. Truong, “ActiveCampus: experiments in community-oriented ubiquitouscomputing,” Computer, vol. 37, no. 10, pp. 73 – 81, 2004.[128] K. Lochner, I. Kawachi, and B. P. Kennedy, “Social capital: a guide to itsmeasurement,” Health and place, vol. 5, no. 4, pp. 259 – 270, 1999.[129] N. Lin, “Building a network theory of social capital,” Connections, vol. 22,no. 1, pp. 28–51, 1999.[130] D. Castiglione, J. W. van Deth, and G. Wolleb, eds., The handbook of socialcapital. Oxford University Press, USA, 2008.[131] T. Glynn, “Psychological sense of community: measurement and applica-tion,” Human relations, vol. 34, pp. 789 – 818, 1981.[132] W. Davidson and P. Cotter, “Measurement of sense of community within thesphere of the city,” Journal of applied social psychology, vol. 16, pp. 608 –619, 1986.[133] D. Perkins, P. Florin, R. R.C., A. Wandersman, and D. Chavis, “Participationand the social and physical environment of neighborhood blocks: crime andcommunity context,” American journal of community psychology, vol. 18,pp. 83 – 115, 1990.[134] D. Julian, T. Reisch, R. Carrick, and C. Katrencich, “Citizen participation-lessons from a local United Way planning process,” American journal of theAmerican planning association, vol. 63, pp. 345 – 355, 1997.[135] J. Western, R. Stimson, S. Baum, and Y. Van Gellecum, “Measuring commu-nity strength and social capital,” Regional studies, vol. 39, no. 8, pp. 1095–1109, 2005.[136] M. S. Smith, “Social capital in online communities,” in Proceedings of the2nd PhD workshop on Information and knowledge management, PIKM ’08,(New York, NY, USA), pp. 17–24, ACM, 2008.146Bibliography[137] S. Wasserman and K. Faust, Social network analysis: methods and applica-tions. Cambridge University Press, 1994.[138] M. Burke, R. Kraut, and C. Marlow, “Social capital on Facebook: differen-tiating uses and users,” in Proceedings of the SIGCHI conference on Humanfactors in computing systems, CHI ’11, (New York, NY, USA), pp. 571–580,ACM, 2011.[139] Y. Jung, R. Gray, C. Lampe, and N. Ellison, “Favors from Facebook friends:unpacking dimensions of social capital,” in Proceedings of the SIGCHI con-ference on Human factors in computing systems, CHI ’13, (New York, NY,USA), pp. 11–20, ACM, 2013.[140] Y. Liu, J. Venkatanathan, J. Goncalves, E. Karapanos, and V. Kostakos,“Modeling what friendship patterns on Facebook reveal about personal-ity and social capital,” ACM Transactions in computer-human interaction,vol. 21, no. 3, pp. 17:1–17:20, 2014.[141] D. R. Millen and J. F. Patterson, “Stimulating social engagement in a com-munity network,” in CSCW ’02, (New York, NY, USA), pp. 306–313, ACM,2002.[142] H. Jiang and J. M. Carroll, “Social capital, social network and identitybonds: a reconceptualization,” in Proceedings of the fourth internationalconference on Communities and technologies, C&T ’09, (New York, NY,USA), pp. 51–60, ACM, 2009.[143] J. Proctor, “Le marche´ St. George, outlaw-quiche and the battle forVancouver’s soul; Battle for beloved community cafe pits people who planLotusland against residents who live in it.”http://www.cbc.ca/news/canada/british-columbia/le-marche´-st-george-outlaw-quiche-and-the-battle-for-vancouver-s-soul-1.3357702, 2015 (lastvisited 2015).[144] K. N. Hampton and N. Gupta, “Community and social interaction in thewireless city: wi-fi use in public and semi-public spaces,” New media &society, vol. 10, no. 6, pp. 831–850, 2008.147Bibliography[145] G. Fleishman, “Some cafe owners pull the plug on lingering wi-fi users.”http://www.nytimes.com/2005/06/13/technology/13wifi.html, 2005 (lastvisited 2015).[146] J. Guynn, “Coffee shops are taking wi-fi off the menu.”http://articles.latimes.com/2010/aug/08/business/la-fi-cafe-wifi-20100808,2010 (last visited 2015).[147] O. Strand, “The new coffee bars: unplug, drink, go.”http://www.nytimes.com/2010/08/25/dining/25coffee.html? r=1, 2010 (lastvisited 2015).[148] P. G. Rowe, Design thinking. MIT Press, 1987.[149] O. Akin, “Variants in design cognition,” in Knowing and learning to design(C. Eastman, M. McCracken, and W. C. Newstetter, eds.), pp. 105–124,Elsevier, 2001.[150] M. E. Atwood, K. W. McCain, and J. C. Williams, “How does the designcommunity think about design?,” in Proceedings of the 4th conference onDesigning interactive systems: processes, practices, methods, and tech-niques, DIS ’02, (New York, NY, USA), pp. 125–132, ACM, 2002.[151] J. M. Greenbaum and M. Kyng, eds., Design at work: cooperative design ofcomputer systems. Hillsdale, NJ, USA: L. Erlbaum Associates Inc., 1991.[152] G. Bjerknes, P. Ehn, and M. Kyng, eds., Computers and democracy: a Scan-dinavian challenge. UK: Avebury: Aldershot, 1987.[153] K. Bodker, F. Kensing, and J. Simonsen, Participatory IT design: designingfor business and workplace realities. Cambridge, MA, USA: MIT Press,2004.[154] J. Grudin, “The case against user interface consistency,” Communications ofthe ACM, vol. 32, no. 10, pp. 1164–1173, 1989.148Bibliography[155] J. D. Gould and C. Lewis, “Designing for usability: key principles and whatdesigners think,” Communications of the ACM, vol. 28, no. 3, pp. 300–311,1985.[156] S. K. Card, T. P. Moran, and A. Newell, The psychology of human-computerinteraction. Lawrence Erlbaum Associates, Inc. Publishers, 1983.[157] D. Norman, “Cognitive engineering,” in User centered system design (D. A.Norman and S. W. Draper, eds.), pp. 31–62, Lawrence Erlbaum Associates,Inc., 1986.[158] D. Norman, The psychology of everyday things. Basic Books, 1988.[159] K. J. Vicente, Cognitive work analysis: towards safe, productive, andhealthy computer-based work. Hillsdale, NJ, USA: L. Erlbaum AssociatesInc., 1999.[160] G. Fischer and E. Scharff, “Meta-design: design for designers,” in Pro-ceedings of the 3rd conference on Designing interactive systems: pro-cesses, practices, methods, and techniques, DIS ’00, (New York, NY, USA),pp. 396–405, ACM, 2000.[161] W. Kunz and H. Rittel, “Issues as elements of information systems,” Centerfor planning and development research, 1970.[162] C. Alexander, S. Ishikawa, and M. Silverstein, A pattern language: towns,buildings, construction. New York: Oxford University Press, 1977.[163] H. Rittel, “Second-generation design methods,” in Developments in designmethodology (N. Cross, ed.), pp. 317–327, New York: John Wiley& Sons.,1984.[164] J. Jones, Design methods. Van Nostrand Reinhold, 1970.[165] N. Cross, Design thinking: understanding how designers think and work.Berg, 2011.[166] D. G. Ullman, “A taxonomy for mechanical design,” Research in Engineer-ing Design, vol. 3, pp. 179–189, 1992. 10.1007/BF01580519.149Bibliography[167] H. A. Simon, The sciences of the artificial. Cambridge, MA, USA: MITPress, 1996.[168] D. A. Schon, “Designing as reflective conversation with the materials of adesign situation,” Research in engineering design, vol. 3, no. 3, pp. 131–147,1992.[169] D. Schon, The reflective practitioner: how professionals think in action.New York NY: Basic Books, 1983.[170] B. Lawson, How designers think: the design process demystified. Architec-tural Press, 4 ed., 2005.[171] C. Alexander, Notes on the synthesis of form. Harvard University Press,1964.[172] R. Krishnamurti, “Explicit design space?,” Artificial intelligence engineer-ing design annals, vol. 20, no. 2, pp. 95–103, 2006.[173] W. J. Mitchell, The logic of architecture: design, computation, and cogni-tion. Cambridge, MA, USA: MIT Press, 1st ed., 1990.[174] T. R. Gruber, “A translation approach to portable ontology specifications,”Knowledge acquisition, vol. 5, no. 2, pp. 199–220, 1993.[175] J. Hois, M. Bhatt, and O. Kutz, “Modular ontologies for architectural de-sign,” in Proceedings of the 2009 conference on Formal ontologies meetindustry, (Amsterdam, The Netherlands, The Netherlands), pp. 66–77, IOSPress, 2009.[176] C. Parent and S. Spaccapietra, “An overview of modularity,” in Modularontologies: concepts, theories and techniques for knowlege modularization(H. Stuckenschmidt, C. Parent, and S. Spaccapietra, eds.), Springer, 2009.[177] O. Akin, Psychology of architectural design. Pion Ltd, London, 1986.[178] C.-S. Chan, “Cognitive processes in architectural design problem solving,”Design Studies, vol. 11, no. 2, pp. 60 – 80, 1990.150Bibliography[179] S. S. Emdanat and E.-G. Vakalo, “An ontology for conceptual design in ar-chitecture,” in Proceedings of the third conference on computer aided archi-tectural design research in Asia, (Osaka, Japan), pp. 425–434, CAADRIA,1998.[180] H. Hutchinson, W. Mackay, B. Westerlund, B. B. Bederson, A. Druin,C. Plaisant, M. Beaudouin-Lafon, S. Conversy, H. Evans, H. Hansen,N. Roussel, and B. Eiderba¨ck, “Technology probes: Inspiring design forand with families,” in Proceedings of the SIGCHI Conference on HumanFactors in Computing Systems, CHI ’03, pp. 17–24, ACM, 2003.[181] K. Huotari and J. Hamari, “Defining gamification: A service marketing per-spective,” in Proceeding of the 16th International Academic MindTrek Con-ference, pp. 17–22, ACM, 2012.[182] S. Deterding, D. Dixon, R. Khaled, and L. Nacke, “From game design ele-ments to gamefulness: Defining ”gamification”,” in Proceedings of the 15thInternational Academic MindTrek Conference: Envisioning Future MediaEnvironments, pp. 9–15, ACM, 2011.[183] J. Hamari, J. Koivisto, and H. Sarsa, “Does gamification work? – a literaturereview of empirical studies on gamification,” in Proceedings of the 201447th Hawaii International Conference on System Sciences, pp. 3025–3034,IEEE Computer Society, 2014.[184] J. Robertson, “Likert-type scales, statistical methods, and effect sizes,”Communications of the ACM, vol. 55, no. 5, pp. 6–7, 2012.[185] M. C. Kaptein, C. Nass, and P. Markopoulos, “Powerful and consistent anal-ysis of Likert-type rating scales,” in Proceedings of the SIGCHI conferenceon Human factors in computing systems, CHI ’10, (New York, NY, USA),pp. 2391–2394, ACM, 2010.[186] N. Memarovic, M. Langheinrich, and F. Alt, “Interacting places: a frame-work for promoting community interaction and place awareness throughpublic displays,” in IEEE International conference on Pervasive computingand communications workshops, pp. 327–430, March 2012.151Bibliography[187] P. Dourish, “Implications for design,” in Proceedings of the SIGCHI con-ference on Human factors in computing systems, CHI ’06, (New York, NY,USA), pp. 541–550, ACM, 2006.[188] J. Reid, R. Hull, B. Clayton, T. Melamed, and P. Stenton, “A researchmethodology for evaluating location aware experiences,” Personal and ubiq-uitous computing, vol. 15, no. 1, pp. 53–60, 2011.[189] M. Hammersley and P. Atkinson, Ethnography: principles in practice. Psy-chology Press, 1995.[190] J. Hughes, V. King, T. Rodden, and H. Andersen, “The role of ethnographyin interactive systems design,” Interactions, vol. 2, no. 2, pp. 56–65, 1995.[191] J. Van Maanen, Tales of the field : on writing ethnography. University ofChicago Press, Chicago :, 1988.[192] G. Bell, M. Blythe, and P. Sengers, “Making by making strange: defamil-iarization and the design of domestic technologies,” ACM Transactions inComputer-human interaction., vol. 12, no. 2, pp. 149–173, 2005.[193] D. E. Forsythe, “It’s just a matter of common sense: Ethnography as invis-ible work,” Computer supported cooperative work (CSCW), vol. 8, no. 1,pp. 127–145, 1999.[194] T. Salvador, G. Bell, and K. Anderson, “Design ethnography,” Design man-agement journal, vol. 10, no. 4, pp. 35–41, 1999.[195] S. Shahidi and Z. M. Kasirun, “Using ethnography techniques in developinga mobile tool for requirements elicitation,” in Proceedings of the 2009 Inter-national conference on Information management and engineering, ICIME’09, (Washington, DC, USA), pp. 510–513, IEEE Computer Society, 2009.[196] K. Ruhleder and B. Jordan, “Capturing complex, distributed activities:video-based interaction analysis as a component of workplace ethnography,”in Proceedings of the international conference on Information systems andqualitative research, (London, UK, UK), pp. 246–275, Chapman & Hall,Ltd., 1997.152[197] A. Crabtree, T. Rodden, P. Tolmie, and G. Button, “Ethnography consideredharmful,” in Proceedings of the SIGCHI conference on Human factors incomputing systems, CHI ’09, (New York, NY, USA), pp. 879–888, ACM,2009.[198] D. R. Millen, “Rapid ethnography: time deepening strategies for HCI fieldresearch,” in Proceedings of the 3rd conference on Designing interactivesystems: processes, practices, methods, and techniques, DIS ’00, (NewYork, NY, USA), pp. 280–286, ACM, 2000.[199] W. J. Winkelman and N. V. Davis Halifax, “Power is only skin deep: aninstitutional ethnography of nurse-driven outpatient psoriasis treatment inthe era of clinic web sites,” Journal of medical systems, vol. 31, pp. 131–139, Apr. 2007.[200] E. Laurier, A. Whyte, and K. Buckner, “An ethnography of a cafe,” Journalof mundane behaviour, 2001.[201] E. Laurier, “How breakfast happens in the cafe,” Time and society, vol. 17,no. 1, pp. 119–143, 2008.[202] E. Laurier, “Drinking up: conversational resources of the cafe,” Languageand communication, vol. 28, no. 2, pp. 165–181, 2008.[203] D. Garlaschelli and M. I. Loffredo, “Patterns of link reciprocity in directednetworks,” Physical review letters, vol. 93, 2004.[204] F. Roberts, Discrete mathematical models, with applications to social, bi-ological, and environmental problems. Prentice-Hall, Englewood Cliffs,1976.153Appendix A – BackgroundReview TaxonomiesDOMAIN CLASS SUBCLASSDisplaysGeometryShapeScaleOrientationReachOn bodyBody reachPlace reachResolutionSpatialTemporalPerceptual CapabilitiesViewing angleContrastBrightnessFigure 1: Taxonomy for Displays154Appendix A – Background Review TaxonomiesDOMAIN CLASS TY PEPublicityOwnershipIndividualCollectiveAccessibilityOpenFilteredPrivacyAnonymousAvatar/PersonaTruthfulFigure 2: Taxonomy for Public Displays155Appendix A – Background Review TaxonomiesDOMAIN CLASS TY PEInteractivityRepresentationIconicSymbolicIndexicalSynchronicitySynchronousAsynchronousAuthorshipHiddenRevealedAmplifiedEmbodimentResponseControlReflectionBelongingAwarenessPeripheralFocalDirectFigure 3: Taxonomy for Interactive Public Displays156Appendix A – Background Review TaxonomiesDOMAIN CLASS TY PEPlaceLocationInternalExternalEnsemblePhysical characteristicsHuman activitiesCharacterInsidenessExistentialBehaviouralEmpatheticVicariousFigure 4: Taxonomy for Place.DOMAIN CLASS TY PEThird PlaceGovernanceNeutralityInclusivenessRootednessLocation and AppearanceAccessibilityInconspicuousnessAmbienceFamiliarityPlayfulnessConversationFigure 5: Taxonomy for Third Place157Appendix A – Background Review TaxonomiesDOMAIN CLASS TY PESocial CapitalConnection TypeBridgingBondingCapital TypeTrustReciprocityCapital ReachParticularizedGeneralizedCapital GainPersonalCollectiveFigure 6: Taxonomy for Social CapitalDOMAIN CLASS SUBCLASSQuantitative definitionsPropertiesLimitsQualitative definitionsConnectionsDependenciesConceptual definitionsContextualAbstractPhysicalFigure 7: Ontologies of an architectural design.158Appendix A – Background Review TaxonomiesDOMAIN CLASS SUBCLASSDimensionsSituationalGeographicalAnthropologicalOrganizationalParti (concept)LinksPhysicalGeometryMaterialsProductionFigure 8: Design dimensions that provide focus during architectural design.159Appendix B – Surveys andInterviewsInitial Inquiry of Third PlacesFocus group interview script1. Tell us your name, your occupation, and what you think is important for youto feel engaged in a community.2. What do you look for in choosing a place to visit and belong to?3. What do you look for in choosing a community to belong to?4. What technologies (if any) do you use to strengthen your bonds with thecommunities you belong to?5. Do you believe that there exists a community between people that visit thisplace?6. Which technologies (phone, smart phone, laptop, tablet, screen, social net-work, none) do you use when visiting this place?7. Which technologies do you believe can help grow a community in this place?Interview Script1. What is a community to you?2. What is public and private in your community?3. What places does your community frequent the most and why?160Initial Inquiry of Third Places4. What do you look for in a place and a community?Technology Usage QuestionnairePlease mark any answers that apply to yourself in the following questions.1. How often do you come to this venue Rarely, Once a week, 2 to 3 times per week,More than 3 times perweek2. Do you most often come to this venue...Alone,With someone else,With 2-4 people,With 5 or more people3. How long you stay, in average, in this venue Less than 30 mins,  30 mins to 1 hour, More than an hour4. What do you come to this venue for Read,  Work,  Talk to friends,  Grab a snack or drink,  Watchsports,  People Watch,  Other5. Devices you use most when coming to this venue Books,  Cellphone,  Laptop,  Other6. What do you use your electronic device for Working,  Chatting or Talking,  Social Networking,  Browsing In-ternet7. What’s your relationship with this venue It’s just a place. Not attached.,  I like it. Somewhat attached.,  I loveit. Attached.,  I’m highly attached8. What’s your relationship to other customers They’re all strangers,  I know some of them,  They’re close friends9. Are you aware of the displays in such venue and content I’m not aware,  I know they’re there.,  I come here to watch them.161CAISCAISSocial Capital QuestionnaireMark your agreement to the following statements from 0 to 7, where 0 meanscompletely disagree and 7 means completely agree1. I often engage in conversation with other visitors of this place that are fromthe opposite sex. 1  3  3  4  5  6  72. I generally interact with visitors of this place from older/younger generationsthan mine. 1  3  3  4  5  6  73. Differences in religious or political views prevent me from interacting withother visitors. 1  3  3  4  5  6  74. I generally interact only with visitors from the same ethnic background asme. 1  3  3  4  5  6  75. I only interact with visitors with a similar degree of education as mine. 1  3  3  4  5  6  76. If being a regular visitor of this place since a while ago, I generally engagewith newcomers. If a newcomer, I generally socialize with long-term regu-lars. 1  3  3  4  5  6  77. I interact with visitors of this place that generally have the same income asme. 1  3  3  4  5  6  78. I often meet family members in this place. 1  3  3  4  5  6  7162CAIS9. If I had a dispute with another visitor of this place I would be willing to seekmediation. 1  3  3  4  5  6  710. I generally take the initiative to what needs to be done. 1  3  3  4  5  6  711. I frequently help people in this place, even if I have little time to spare or ifoutside my personal interests. 1  3  3  4  5  6  712. I often volunteer in a group that meets in this place. 1  3  3  4  5  6  713. I often meet new people in this place. 1  3  3  4  5  6  714. Number of community events organized by visitors of this place that I haveattended in the past 6 months. 1  3  3  4  5  6  715. Number of organizations or clubs that use this place as a meeting point ofwhich I’m an active member. 1  3  3  4  5  6  716. Number of organizations or clubs that use this place as a meeting point onwhich I’m on a management or organizing committee. 1  3  3  4  5  6  717. Number of times in the past 3 years in which I have joined a local communityproject or working bee that used this place as a meeting point. 1  3  3  4  5  6  718. Number of times I have been part of a project to organize a new service inthis place. 1  3  3  4  5  6  7163CAIS19. Personally now someone that works in this place. 1  3  3  4  5  6  720. Personally know a manager of this place. 1  3  3  4  5  6  721. Personally know the owner of this place. 1  3  3  4  5  6  722. Personally know a regular costumer of this place. 1  3  3  4  5  6  723. To what extent would you trust your close friends? 1  3  3  4  5  6  724. To what extent would you trust the management of this place? 1  3  3  4  5  6  725. To what extent would you trust other visitors of this place? 1  3  3  4  5  6  726. To what extent would you trust the people that work in this place? 1  3  3  4  5  6  727. How often do you and your friends exchange practical help or advice? 1  3  3  4  5  6  728. How often do you and visitors of this place exchange practical help or ad-vice? 1  3  3  4  5  6  729. How often do you and workers of this place exchange practical help or ad-vice? 1  3  3  4  5  6  730. I feel safe coming here alone. 1  3  3  4  5  6  7164CAIS31. Most visitors of this place can be trusted. 1  3  3  4  5  6  732. If another visitor of this place needs one dollar I would lend it to him/her. 1  3  3  4  5  6  733. This place has the reputation for being a safe place. 1  3  3  4  5  6  734. I feel safe coming with my friends to this place. 1  3  3  4  5  6  735. I have confidence in this particular establishment. 1  3  3  4  5  6  736. If applicable, I have confidence in this particular chain or brand. 1  3  3  4  5  6  737. I have confidence in the people that work in this place. 1  3  3  4  5  6  738. I have confidence in the people that manage this place. 1  3  3  4  5  6  739. I have confidence in other visitors of this place. 1  3  3  4  5  6  740. Visitors of this place are really willing to help each other out. 1  3  3  4  5  6  741. Visitors of this place share the same values. 1  3  3  4  5  6  742. I feel a strong sense of identity with a local community when at this place. 1  3  3  4  5  6  743. I am well informed about this establishment’s a airs and plans. 1  3  3  4  5  6  7165CAIS44. Feel emotionally attached to this place. 1  3  3  4  5  6  745. Feel that belong to this place, specially its community. 1  3  3  4  5  6  746. I would like to be able to be engaged with the community that meets in thisplace in 3 years time. 1  3  3  4  5  6  747. I am proud to be a visitor of this place and member of this community. 1  3  3  4  5  6  748. I visit this place because I enjoy the place. 1  3  3  4  5  6  749. I visit this place because I wanted to become part of a group/community thatmeets here. 1  3  3  4  5  6  750. I visit this place because I wanted to give something back to a group orcommunity that meets here. 1  3  3  4  5  6  751. I visit this place to meet people and make friends. 1  3  3  4  5  6  752. I visit this place due to a series of coincidences and unexpected connections. 1  3  3  4  5  6  753. I visit this place because I’m forced to do it. 1  3  3  4  5  6  754. I visit this place because my friends visit it. 1  3  3  4  5  6  755. I believe that employees of this place take visitor’s interests into account. 1  3  3  4  5  6  7166CAIS56. I believe that managers of this place take visitor’s interests and views intoaccount. 1  3  3  4  5  6  757. I believe that owners of this place take visitor’s interests and views into ac-count. 1  3  3  4  5  6  758. If applicable, I believe that this particular chain or brand takes visitor’s in-terests and views into account. 1  3  3  4  5  6  7Post-Deployment Interview Script1. What aspects of the prototype did you find useful and not useful?2. What aspects of the prototype did you find aesthetically pleasant and un-pleasant?3. What aspects currently existing in the prototype would you modify and how?4. What functionalities, currently not existing in the prototype, would you liketo see implemented?5. Did you feel that this technology brought you closer to the community thatgathers in this place?6. Do you have any other comments?Post-Deployment QuestionnaireMark your agreement to the following statements from 0 to 7, where 0 meanscompletely disagree and 7 means completely agree1. I changed my daily routines since the system was installed. 1  3  3  4  5  6  72. The system made me talk to other professionals to solve issues. 1  3  3  4  5  6  7167MySeedlings3. The system made me help professionals that I didn’t help before. 1  3  3  4  5  6  74. The system helped me resolve conflicts with other professionals. 1  3  3  4  5  6  75. The system created conflicts with other professionals. 1  3  3  4  5  6  76. I feel more connected to other professionals after using the system. 1  3  3  4  5  6  77. I know what the visualization represents (balls and lines on a rotating circle). 1  3  3  4  5  6  78. The visualization changed how I perceived the previous system. 1  3  3  4  5  6  7MySeedlingsThird Place SurveyMark your agreement to the following statements from 0 to 5, where 0 meanscompletely disagree and 5 means completely agree. (We have specified here theconcept measured in between parentheses).1. (Community) Taking care of the garden felt like a communal effort withother members of MySeedlings. 1  3  3  4  52. (Regulars) I feel like a regular after playing MySeedlings. 1  3  3  4  53. (Welcoming) MySeedlings felt welcoming when I first started using it. 1  3  3  4  5168MySeedlings4. (Leveller) MySeedlings allowed me to interact with people I would have nototherwise interacted with. 1  3  3  4  55. (Playful) Using MySeedlings was a playful experience. 1  3  3  4  5Survey to Capture Perception (Enjoyment and Intrusiveness) by Usersand Non-UsersMark your agreement to the following statements from 0 to 5, where 0 meanscompletely disagree and 5 means completely agree. (We have specified here theconcept measured in between parentheses).1. (Enjoyable) The MySeedlings system (Game and Garden) makes the expe-rience of visiting “Seedlings Cafe” more enjoyable. 1  3  3  4  52. (Intrusive) The MySeedlings system (Game and Garden) is distracting orintrusive to my regular experience of ”Seedlings Cafe”. 1  3  3  4  5Survey to Capture Sense of Belonging, Exclusiveness and SocialAgency.Mark your agreement to the following statements from 0 to 5, where 0 meanscompletely disagree and 5 means completely agree. (We have specified here theconcept measured in between parentheses).1. (Belonging to Game) After playing MySeedlings I feel a sense of belongingto MySeedlings. 1  3  3  4  52. (Belonging to Cafe) I feel a sense of belonging to ”Seedlings Cafe” afterplaying MySeedlings. 1  3  3  4  5169MySeedlings3. (Feeling of Exclusiveness) MySeedlings is not for everyone, it’s an exclusiveexperience. 1  3  3  4  54. (Social Agency) I would like to meet people that I interacted with in My-Seedlings, someday in the future. 1  3  3  4  5Informal Interview ScriptFamiliarity1. Do you know what MySeedlings is?2. Do you know about the MySeedlings’s display and garden?Perception3. What are your thoughts about MySeedlings?4. Would you like to keep MySeedlings permanently at the coffee shop?Usage5. Have you used the seating areas (couch, cushions) next to the garden ordisplay in the last 4 weeks?6. Have you visited the URL or scanned the QR Code on the MySeedlingsdisplay?7. Have you created an account or signed into MySeedlings in the last 4 weeks?Informal InterviewThe following two questions were used as an informal interview script appliedto MySeedlings users after their participation.1. Do you have any thoughts about MySeedlings?2. Which game play (“Water” points or “Harvest” points) did you prefer andwhy?170MySeedlingsSocial Capital SurveyMark your agreement to the following statements from 0 to 5, where 0 meanscompletely disagree and 5 means completely agree. (We have specified here theconcept measured in between parentheses).Informal Structures1. (Exclusion) In general, I have had more contact with other people in My-Seedlings / Seedlings-Cafe that are similar to me. 1  3  3  4  52. (Particularized Agency) I have helped others in MySeedlings / Seedlings-Cafe, outside of my personal interests. 1  3  3  4  5Formal Structures3. (Participation in Community) In a year from now I would (play MySeedlingsif there was no prizes for participation/volunteer at Seedlings-Cafe). 1  3  3  4  5Informal Norms4. (Particularized Trust) If another MySeedlings user / Seedlings-Cafe visitorthat I have never met before needed one dollar I would lend it to him/her. 1  3  3  4  5Formal Norms5. (Community Spirit) People in MySeedlings / Seedlings-Cafe are often will-ing to help each other out. 1  3  3  4  56. (Place Attachment) I feel a sense of identity with other people here in My-Seedlings / Seedlings-Cafe. 1  3  3  4  5171MySeedlings7. (Feelings of Safety and Generalized Trust) In general I trust other users of(MySeedlings/Seedlings-Cafe). 1  3  3  4  58. (Formal Reciprocity) I played MySeedlings / visit Seedlings-Cafe because Iwanted to give something back to the place or other people in there. 1  3  3  4  5172Appendix C – ResearchMethodologiesThis section presents the methodologies to capture and evaluate the effects ofinteractive public displays on communities used in this research. In particular wediscuss ethnographic tools that can be useful in capturing practices of a commu-nity, survey instruments used to capture a community’s overall social capital, anduseful network analysis tools to gain detailed insights on social capital aspects of acommunity.Ethnographic Documentation of Practices and AdoptionWhen designing interactive public displays for socialization it can be difficultto capture subtle changes a community goes through when technology is deployed[27] using only quantitative methodologies. This is because technology can takesocial meanings that escape quantitative measurements when embedded into thefabric of communities. Ethnographic methodologies, Dourish et al. argue [187],can help mitigate this burden by helping researchers capture participants not aspassive “users”, but as actors with cultural and social circumstances that affectinghow they use and adopt new technologies. Ethnographic methods are helpful infactoring out novelty effects and reducing the observer bias in observational stud-ies [188] . The main purpose of ethnography is to record people’s actions in anevery day context, that is in the field, and focuses on user observations and infor-mal conversations [189]. Ethnography involves researchers immersing themselvesin the group that they wish to study [187], and by doing so gaining a deeper un-derstanding of the hidden workings of that particular group, the social links thattie the group together, and how every day goals are achieved. Ethnography gives173Ethnographic Documentation of Practices and Adoptiondesigners the ability to understand work places as socially organized settings, anduncover the hidden structures that allow for such sociability [190] and can serve asa powerful tool to understand context and cultural factors that affect a community[191].The value of ethnography for understanding how people appropriate technol-ogy has previously been discussed in the field Human Human Computer Interaction(HCI), particularly by Bell et al. [192], Forsythe [193] and Salvador et al. [194].Researchers that have experimented supporting ethnographic methods in HCI in-clude Shahidi et al. [195] who have worked on developing mobile tools for re-quirement elicitation using ethnographic methods, particularly taking and organiz-ing field notes, and Ruhleder et al. [196] who explore the use of video to capturedistributed and complex collaboration in communities. Some researchers, partic-ularly Dourish [187] and Crabtree et al. [197], caution that such methods must beapplied with care because some of their practices, if not understood, can be harm-ful to research. The present research uses the field methods design proposed byMillen et al. [198] for studies in the field of HCI based on key informants, collab-orative data analysis and interactive observations. In particular, we focus on twomethodologies: shadowing and field notes (Figure 6.6).Technique PurposeShadowing Understand existing practices by maintaining adistance from community.Field Notes Gain an in-depth understanding change in prac-tices and thoughts by becoming part of commu-nity.Table 1: Ethnographic techniques used in this research.During the “shadowing” technique researchers need to follow people that be-long to a community as they go about their daily routines. They must gather in-formation that could help them characterize the current adopted practices at un-derstanding the daily flow of information, existent social ties and the functioningof the studied place. This methodology has been previously used in the field ofHuman Computer Interaction. Shadowing a community to understand the role oftechnology in a community has been previously used by Winkelman et al. [199],who study the introduction of clinic web sites to a community of outpatient nurses.174Ethnographic Documentation of Practices and AdoptionIn a third place setting, Laurier et al. [200] has previously used ethnographic shad-owing to observe the functioning of a coffee shop by recording daily activities andconversations. In our research shadowing observations were tabulated accordingto characteristics of third places cited by Oldenburg [15]. Table 2 presents theconcepts of third place used in this research and their definition, used as points ofdeparture to classify our shadowing observations.Characteristic Observations GuidelinesWelcoming Events demonstrating that newcomers are well received by acommunity. In terms of a place, characteristics that make itreachable to a wide type of peoples.Accommodating When a community is open to its members’ needs and wishes,or when a place listens to the wishes of the people that visit it.Low-profile Characteristics of a place that make it economically and cul-turally accessible, that is unpretentious. Events and interactionthat bear commonality.Neutral Ground Events demonstrating that peoples from different economic,cultural, racial, religious and political groups are not only wel-comed by the community or place.Inclusive When minorities are welcomed by a community or place.Events or stories that show that minorities are part of a com-munity.Regulars The existence of people that are “always there” or regularly visita place or regularly found when the community gathers.New friendships Events demonstrating that newcomers are assimilated by exist-ing members of the community, or that new friendships betweenstrangers arise.Playful Events demonstrating a playful ambience in a place, or the al-lowance of games, amusement and lightheartedness. When se-riousness is not the predominant characteristic of a communityor place.Conversation Events and stories where conversation occurs.Table 2: Third Place characteristics and guidelines used to classify shadowing ob-servations.Field notes, or fieldwork journals, can be an effective tool to document eventsbeing studied while in the field. This tool provides a chronological account ofthe research and contains not only observations, but a researcher’s feelings andinvolvement that have strong analytic significance to understanding a community[189]. A journal consists of writing observations, conversations and stories experi-175Social Capital Surveysenced while immersing oneself in a community of people [189]. Observations arecomplemented with analytic notes and memos that document a researcher’s reflec-tion on the recorded events. The recorded observations and analytic observationsgo hand in hand as they allow the researcher to reflect on his or her observations,provide a preliminary analysis of the data and guidance for future analysis. Fieldnotes in Human computer Interaction research are often used to capture the peopleand events that are the focus of a research [198]. They can help detail peoplesconcerns as they approach new technologies, provide access to idiosyncrasies androutines, and help researchers document adoption patterns that other methods can-not capture. Observational field notes have been previously used in third places todocument routines [201] and conversations [202] to document the conversationalfunctioning of third places. In the present research we report field notes in the formof captured stories recorded by a researcher visiting the studied place as a regularof such place. Notes were typed on a laptop computer and included time and date ofthe observation, any important contextual information that (e.g. “the place is busy,with loud chatter that makes it hard to think”), any conversations that took placeduring this event, and notes regarding the researcher’s interpretation of the event.It should be noted that the complexity of real world deployments, in particularlythird places where often many events occur simultaneously, may have forced us toapply our observations in a selective manner. This selective approach is inherent tothe field notes tool [189] as it is impossible for researchers to document all eventsobserved. However, we aim at being impartial when faced with a high bandwidthof events and record such conflicts.Social Capital SurveysWestern [135] proposes that social connections can be understood as havingtwo dimensions: structural and normative, across formal and informal links. Struc-tural measurements focus on understanding size, capacity, openness, homogeneityand density of the community’s network, while normative measurements focus ontrust, unity and reciprocity experienced by members of such group. Both measure-ments have two dimensions; formal, that is institutions, and informal links, thatis friends and relatives. Measuring social capital in small co-located communities176Social Capital SurveysStructuresInformal FormalNorms Informal Informal Networks Mixed NetworksFormal Mixed Networks Formal NetworksTable 3: A model to measure Social Capital based on Structures and Norms(adapted from Western et al. [135]varies from this approach. The interwoven nature of small communities and thestructural and functional nature of the spaces where such communities meet, shapeand modify how such communities form. The socially dense nature of socializationplaces, the multitude of communities that overlap in both space and time, and thestrong retail nature of such spaces make the measurement of social capital in suchspaces difficult to perform. We propose an adapted measurement of social capi-tal that includes the daily functioning of coffee shops as starting point to developtechnologies for third-placeness.A version of Western’s measurements, adapted to the functioning of small thirdplaces can be used to measure the quality and strength of the social links that existin smaller place bound communities. On-site observations of third place function-ing and discussions management of several third places led us to the measurementspresented in Table 3. As such measurements illustrate, social interactions in thirdplaces are strongly linked to food consumption, customer to customer interactionand customer to owner interactions.Measurement Definition Sample QuestionInformal StructuresCommunity divisions How important are differencesin age, or gender, ethnicity, ed-ucation, religious and politicalbeliefs.In general I have had more con-tact with other people in thisplace that are similar to me.Particularized agency Capacity or will to plan and ini-tiate action to help others in agroup, for example seeking me-diation when conflict arises.I have helped others in thisplace, outside of my personalinterests.Formal Structures177Social Capital SurveysMeasurement Definition Sample QuestionParticipation in localcommunityHow often people participatein the community, for exam-ple volunteering, participatingin community projects, etc.In a year from now I would vol-unteer in this place or partici-pate if there was no remunera-tion.Generalized agency Capacity or will to initiate ac-tion through institutional meansto help a community, for exam-ple signing a petition.In the last year I have joinedefforts with other people to re-solve a problem of this place.Friends in institutionalnetworksHow friends people have in in-stitutional parts of the com-munity, for example being ac-quainted to a place’s ownersand employees, or group lead-ers.I personally know the owner oran employee of this place.Informal NormsParticularized trust Trust in other individuals in thecommunity.If another person that I havenever met before needed onedollar I would lend it tohim/her.Informal reciprocity andexchangeExchange of practical help oradvice with other individuals inthe community.I often exchange practical helpor advice with other people inthis place.Formal NormsOpenness and toleranceto diversityWill to welcome new culturesand lifestyles.I enjoy cultural and lifestyledifferences that exist in thisplace.Feelings of safety andgeneralized trustFeelings of safety and trustfrom the general community.I trust others in this place.Confidence in links Confidence in institutional linksof policing and management.I have confidence in the man-agement of this place or com-munity.Community spirit Sense of identity and will tohelp each other out.People here are often willing tohelp each other out.Place attachment Emotional attachment, or senseof belonging, to a place.I feel a sense of identity withother people in this place.178Network Analysis Tools for Social CapitalMeasurement Definition Sample QuestionFormal reciprocity Will to give back to the commu-nity in general.I visit this place because Iwanted to give something backto the place or people in here.Trust of links Trust that institutional links ofpolicing and management takethe views of people into ac-count.The management of this place,or leader of this group, take intoaccount the views of people andtheir needs.Table 4: Measurements of Social Capital (adapted from Westernet al. [135]). We provide prototypical questions used throughoutthis research for each measure.Network Analysis Tools for Social CapitalNetwork analysis consists of describing social interactions as networked graphsconsisting of nodes and arcs between such nodes. Often, people are modelled asnodes and inter-personal interactions and social links (e.g. friendships) are mod-elled as arcs between such nodes. This approach provides researchers with pre-cise ways to define and capture social concepts and frameworks to evaluate the-ories about social interactions between people [137]. Furthermore, it provides awide selection of well-tested tools and approaches to concepts that risk being onlydefined in a metaphorical manner. Understanding concepts like favours, socialstrength and social importance can be tackled through well studied methodologieslike reciprocity, density and centrality. Network analysis has been previously usedby researchers in the field of Human Computer Interaction as a tool to understandsocial interaction and social capital [138–140] (Chapter 2). We focus on three mea-surements: Reciprocity, which models how much people are willing to reciprocateactions in a community; Density, modelling how many of these actions are ty-ing up a community; and Centrality, or how such actions structure the communityand control the flow of information, that is people’s popularity or gregariousness(Degree Centrality), speed at which actions and information travel within the com-munity (Closeness Centrality), and how many people are in-between other people179Network Analysis Tools for Social Capitalserving as connectors of the group or holders of information (Betweenness Cen-trality).For the measures of Reciprocity and Density we followed the methodologyproposed by Garlaschelli et al.[203] to compute graph reciprocity by computing ameasure of link reciprocity that includes the actual degree of correlation betweenmutual links in a directed non-self-looping network. For the measures of Centrality(Degree, Closeness and Betweenness) we followed the methodologies proposed byWasserman [137].DensityGraph density depicts how many task links exist between all people, comparedto how many task links could have been created between all people in the commu-nity. This coefficient is computed by the equation (1) where the maximum densityis defined as dmax = N(N−1). Graph density can help us understand how peopleusing the system interacted with the community as a group. A graph with highdensity would mean that almost every person interacted with almost everyone elsein the group. A graph with low density would mean that people created links withonly a few of all possible people.a¯ =LN(N − 1) (1)ReciprocityWe followed the methodology proposed by Garlaschelli et al. [203] to com-pare graph reciprocity by computing a measure of link reciprocity that includes theactual degree of correlation between mutual links in a directed non-self-loopingnetwork defined by the equation (2). Where r is the definition of reciprocity as theratio of number of links pointing in both directions L ↔ to the total number oflinks L in a graph, r = L ↔ /L. And where a¯ measures the ratio of observed topossible directed links, or network density where self-loops are excluded.This approach allows us to distinguish between reciprocal (p > 0) and anti-reciprocal (p < 0) networks, which represents a Pearson Correlation of reciprocity.180Network Analysis Tools for Social CapitalNetworks with a reciprocity coefficient p = 0 are considered a-reciprocal net-works, while if all links occur in reciprocal pairs the coefficient p = 1. Onemay also compute the minimum reciprocity for given graph density as: pmin ≡−a¯/(1− a¯). In this case there are no reciprocal links in the network, i.e., L↔= 0.One can then calculate the standard deviation in terms of the sum of the varianceswhen computing r for all nodes when a node aij is removed computing the stan-dard deviation in terms of the sum of the variances when computing r for all nodeswhen a node aij is removed (σ2p =∑i<j(p− pij)2).A graph with a high reciprocity coefficient would imply that people often in-teract with people that have previously interacted with them, that is interacting inpairs or small clusters. In contrast, a graph with a similar number of links but alower reciprocity coefficient implies people do not reciprocate when another peo-ple reciprocate interact with them, or that people interact with a larger group ofpeople indiscriminately.p =r − a¯1− a¯ (2)CentralityCentrality measures can be used to determine the importance of a person in asocial network [136, 137] . Common centrality measures include degree, closenessand betweenness. Degree centrality is defined by the of the in-degree or out-degreeratio of a particular node calculating its importance in a network. Closeness cen-trality calculates how long information is spread from a node to all other nodessequentially. Betweenness centrality calculates the shortest path between everynode within the network, determining which nodes are most central.Degree Centrality is defined as the number of links that enter (inlinks) a node,or number of social ties a person has (outlinks). A person with a high degree cen-trality maintains many links with other people, as a result a person with highercentrality can gain access to and/or influence other persons in his or her commu-nity. This measure is effective at finding actors that occupy a structural positionthat serve as a source for larger volumes of information exchange and social capi-tal transactions with other people. By contrast, a person with a low degree central-181Network Analysis Tools for Social Capitality maintains few or no relations and often marginalized when it comes to socialagency. Actor-level Degree Centrality is computed [137] as each node’s numberof links in a non-directed graph CD(ni) = d(ni) =∑j xij =∑j xji. In directedgraphs the Degree Centrality index is the sum of outlinks from a node to all ad-jacent nodes (CD(ni) = d(ni) =∑j xij). This definition, however is dependenton the size of the group. To standardize or normalize this index, so that networksof different sizes (g) can be compared, we divide it by the maximum possible in-degree links, which would be g − 1 if every node is connected to a node i. Sothe normalized Degree Centrality for each node is given by Equation 3. DegreeCentrality can be applied to directed graphs if all nodes in the graph are allowed tocreate links with everyone in the network being used to compute the index. In thecase of Centrality, in a directed graph inlinks (links coming into a indegree node)are interpreted as popularity or prestige, while outlinks (those going out of an out-degree node) are often associated with gregariousness. Moreover, outlinks reflecteach node’s choices made on creating such links. Because of this we computeDegree Centrality with outdegree nodes (outlinks) in our directed graphs.CD(ni) = d(ni)/(g − 1) (3)Closeness Centrality of a node is defined as the sum of its shortest paths to allother nodes in the graph. In connected graphs the shortest distance or geodesicdistance between two nodes is the number of links that connects them. This mea-sure is often used to measure how long information is spread from a node to allother nodes sequentially. An actor with a high Closeness Centrality has minimumpath distances from all other members of the graph and will approach g − 1. Thisperson will be close to many others in the community and can quickly interactand communicate with them without going through many intermediaries. Actorcloseness centrality is computed as the inverse of the sum of geodesic distancesfrom an actor i to g − 1 other actors: (i.e., the reciprocal of its farness score)[137]: Cc(ni) = [∑gj=1 d(ni, nj)]−1. To Standardize this measure, we divide it by themaximum possible distance g−1, giving Equation 4. However, there is a drawbackto this measure. Closeness can only be computed for a connected graph becauseif two nodes are not mutually reachable (if no paths exist between i and j) their182Network Analysis Tools for Social Capitalgeodesic distance becomes ‘infinite’. Furthermore, in directed graphs the geodesicdistance between two nodes differs with the nodal order, that is d(ni, nj) may notbe equal to d(nj , ni), and thus closeness will only be defined if the graph is stronglyconnected (directed paths in both directions). In scenarios where some members ofa community are new and have not yet created links or in trial scenarios where par-ticipants register, but choose not to participate this condition might be difficult toapply. An improved actor-level closeness index can consider how proximate ni isto the actors in its influence range. This index is called Influence Range ClosenessCentrality and computes the standardized inverse average distance between a nodeni to every other node reachable from it. This index can then be used for directedgraphs that are not strongly connected. Opposed to the commonly used ClosenessCentrality (Equation 4) Influence Range Closeness Centrality considers only dis-tances from a node ni to nodes in its influence range J , or nodes reachable fromni. We compute this index as the ratio of the fraction of nodes reachable by a node(|J |/(n − 1))to the average distance of these nodes from ni (∑(d(ni, nj))/|J |),giving Equation 5C ′c(ni) =g − 1[∑gj=1 d(ni, nj)]= (g − 1)Cc(ni) (4)C ′c(niJ) =|J |/(n− 1)∑(d(ni, nj))/|J | (5)Betweenness Centrality measures the number of times a node acts as a bridgealong the shortest path between two other nodes. A high Betweenness Centralityindex represents an actor occupying a “between” position on the geodesics con-necting many pairs of other actors in the networks. In a social network, this personhas an increased control over the flow of information or exchange or resources,controlling social agency and information, and in turn social capital. It is often as-sumed that if more than one geodesic links a pair of nodes, we assume that each ofthese shortest paths has an equal probability of being used. Computing Between-ness Centrality for an actor i is the sum of the proportions, for all pairs of actors jand k, in which node ni is involved in the jk geodesics: Cb(ni) =∑j<kgjk(ni)gjk.As with other centrality ways of normalization we divide this measure by the max-imum possible betweenness, giving Equation 6. Although Betweenness Centrality183Network Analysis Tools for Social Capitalwas originally developed for non-directed graphs it can be used in directed graphs.However, the normalized value C ′b(ni) should be multiplied by a factor of 2, toadjust for maximum undirected ties. It is important to notice that this measure as-sumes that interaction between people occurs along the shortest possible paths andneglects the possibility of communication along other means (non-geodesic paths).C ′b(ni) =(g−1) C2∑j<kgjk(ni)gjk (6)Weighted GraphsIn social network analysis it is common to have situations where pairs ofnodes have several arcs between them. For example in cases where repetitiveinteraction between people takes place. This is a special case of graph that isknown as a valued graph where arcs between nodes are valued, that is, associ-ated with each arc is a value. This value, in such case represents the number ofdirected or undirected interaction between nodes. If such arcs are undirected andnode i and j are connected the relationships between such nodes are symmetrical(lk = (ni, nj) = (nj , ni) = lm). In a valued directed graph the arc from nodeni to nj is not the same arc from nj to ni ((lk = (ni, nj) 6= (nj , ni) = lm)).Valued graphs are also known as weighted graphs [204]. Many of the studies inour work use weighted graphs to represent the interaction between people. For ex-ample, when a person A sends three messages to person B a link will exist definedby lk(i→j) = 3, if person B sends two messages to person A a second link willexist defined by lk(j→i) = 2. In this special case a few distinctions in the abovemethodology must be made.Density is defined as the ration of the number of arcs present to the maximumpossible that could arise. That is, each pair of nodes is analyzed and a value of 1 isgiven if a link exists or a value of 0 if a link is absent, these values are then summedand divided by all the links measured. This allows us to generalize the concept ofdensity by averaging the values attached to all links across all measured arcs. Thismeasure then represents the average weight of the arcs in the graph [137]. Equation184Network Analysis Tools for Social Capital7 illustrates this, where the sum is taken over all possible arcs k.a¯ =∑Wkg(g − 1) (7)Reciprocity using formula 2 is optimized for non-weighed networks. It as-sumes that links between nodes have equal weight Wk = 1, allowing to sim-plify∑i 6=j a2ij =∑i 6=j aij = L. However, in weighted networks∑i 6=j a2ij >∑i 6=j aij . We follow Garlaschelli et al.’s approach to computing reciprocity asa correlation coefficient r given by equation 8. This change retains computingpmin ≡ −a¯/(1 − a¯) and obtaining the standard deviation in terms of the sumof the variances when calculating r for all nodes when a node aij is removed(σ2p =∑i<j(p− pij)2).r ≡∑i 6=j(aij − a¯)(aji − a¯)∑i 6=j(aij − a¯)2(8)Centrality (Degree Centrality index) in directed graphs is the sum of outlinksfrom a node a to all adjacent nodes. If the network being analyzed is weightedthe Degree Centrality index is the sum of weights of outlink edges from a node toall adjacent nodes (CD(ni) = d(ni) =∑jWxij), one can then apply equation3 as previously stated. We use weights to represent the number of interactionsbetween people and consider each interaction of equal value (1). In this approach,weights represent a quality of the network that does not affect the length of thearc. Thus, when analyzing Closeness Centrality (the sum of all shortest paths to allother nodes in the graph) we can use equation 4 without modification. In the caseof computing Betweenness Centrality (number of times a node acts as a bridge) theweight of an arc, in our case representing multiple arcs between two nodes has noeffect on equation 6, as the definition already assumes that multiple paths betweennodes have equal probability of being used.185


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items