I am interested in the ways in which people manage attention and self-presentation as they work, play and otherwise interact online. My goal is to impact both theoretical understandings of behavior, and our ability to better support day-to-day activities. I explore these issues in a series of projects that draw on both field studies of interaction in situ and laboratory studies that explore particular phenomena in more detail. I have also been involved in the development and evaluation of novel technologies, some of which have led to methodological advances in field data-gathering and experimental research.
I direct the Social Media Lab, where my students and I are involved in a range of projects that reflect the following themes.
Theme 1: Managing and Negotiating Attention
This work began with an interview and observational study of how people pay attention to others and manage privacy in open-plan offices. The motivating goal in this study was to understand why people felt comfortable working in open offices where their activities were constantly visible to others, but often balked at the possibility of sharing similar information with geographically distributed colleagues online (e.g., via always-on video cameras or screen sharing tools). We learned from this study that interpersonal attention in these environments is often a public act. To see what others are doing, one must move physically closer to them. This act of moving closer renders the act of paying attention visible, which both allows the object of attention to respond (e.g., by looking away or hiding sensitive information) and normatively constrains behavior.
A series of studies has led us to the premise that that negotiating states of mutual attention is usefully considered as what Clark refers to as joint activity. Joint activity occurs when two or more people act individually toward achievement of a shared goal, and individual acts occur in response to those by others. In the my framework for attention management, gathering refers to one person getting information about what others are attending to, such as tasks, other people or the gatherer herself. Display refers to information about one’s own attention that is available for others to gather.
Suppose Alex and Bill work across the room from each other. If Alex walks closer to Bill to gather information about whether or not Bill is available to talk, Bill may notice Alex’s presence and glance up at him. Alex may then notice Bill’s glance and return it. In this way, Alex’s approach simultaneously serves to gather and display. Closer proximity to Bill means Alex can gather more detailed information, and also makes Alex’s presence more noticeable to Bill. This triggers Bill’s glance, which allows Bill to see that Alex is approaching, and display via a glance at Alex that Bill has noticed Alex’s approach.
Most existing interaction technologies and frameworks treat conversation as the only truly interactive element of the attention negotiation and communication process. In considering the attentional processes leading up to conversation, others have focused either on gathering (e.g., interruption timing) or display (e.g., notification systems) behaviors. This separation ignores the interplay of these activities and makes it difficult to understand and support the ways in which people actually manage their attention.
Over the past five years we have developed several versions of the OpenMessenger system, which is a prototype messaging system developed to allow for laboratory experiments on joint attention in collaboration. Initial experiments with this system suggest that there is utility in this approach and that people’s information gathering behavior differs in an environment where their behavior is visible to others.
Birnholtz, J., Bi, N., & Fussell, S. Do you see that I see? Effects of perceived visibility on awareness checking behavior. Proceedings of the ACM Conference on Human Factors in Computing Systems (CHI 2012). [PDF]
Birnholtz, J., Schultz, J., Lepage, M., & Gutwin, C. (2011). A framework for supporting joint interpersonal attention management in distributed groups. Proceedings of INTERACT, Lisbon, Portugal, September 5-9, 293-312 [PDF]
Birnholtz, J., Gutwin, C., Hawkey, K. (2007) Privacy in the Open: How Attention Mediates Awareness and Privacy in Open-Plan Offices, Proceedings of the ACM Conference on Supporting Group Work (GROUP 07), November 4 – 7, Sanibel Island, FL, pp. 51-60.
Theme 2: Butler Lies: Social Inattention Management
A significant success of today’s communication environment is that the ability to communicate with other people is virtually constant, which enables frequent and novel modes of interaction and information sharing. At the same time, however, people increasingly report being overwhelmed or distracted by interaction opportunities. This can hurt productivity, cause stress, and complicate social relationships.
One cause of these problems is that technology designers have focused on support for coordinating co-presence and availability. Historically, availability management was largely a function of coordinating co-presence: establishing a time and place, and/or initiating interaction once co-presence was achieved. In an always-on world, however, people are technically always co-present. This foregrounds the problem of managing when one is not available, which can require some explanation for why – despite co-presence and seeming availability – interaction cannot take place. Social inattention refers to the deliberate social processes involved in avoiding or curtailing interactions, such as ignoring unwanted interruptions or excusing oneself from a conversation. Social inattention has significant relational consequences if not managed effectively, as being perceived to ignore others may offend or undermine relationships.
We have conducted a series of studies on butler lies, so called because they are the sort of lies that butlers would have told when answering the door and saying the master is unavailable. Butler lies are a common linguistic strategy for managing inattention. By understanding how people use and craft butler lies in various media and settings, our goal is to help better understand and support social inattention management.
Reynolds, L., Smith, M., Birnholtz, J., Hancock, J. Butler Lies From Both Sides: Actions and Perceptions of Unavailability Management in Texting, To be presented at the Conference on Computer-Supported Cooperative Work (CSCW 2013), February 23-27, 2013. [PDF]
Birnholtz, J., Dixon, G., Hancock, J. (2012). Distance, ambiguity and appropriation: Structures affording impression management in a collocated organization. Computers in Human Behavior, 28, 1028-1035 [PDF]
Hancock, J., Birnholtz, J., Bazarova, N., Guillory, J., Perlin, J., Amos, B. (2009) Butler Lies: Awareness, Deception and Design, in the Proceedings of the ACM Conference on Human Factors in Computing Systems (CHI 09), Boston, MA, April 4-9, pp. 517-526
Theme 3: Understanding Collaboration and Coordination Problems
Studying real-world work, social and educational collaborations in the field is an invaluable way to identify problems in attention management and coordination that merit further study. My early work as a graduate student has impacted the design and study of systems for data sharing and remote experimentation, as well as discussions of of “hyperauthorship” in high energy physics.
The coordination problems I observed in these collaborations led me to my current interests in how people manage attention to each other. I have recently conducted two field studies of collaboration between Cornell’s Ithaca campus and the Weill Cornell Medical College in New York City. These studies have highlighted unique difficulties of single-institution collaboration and motivated bibliometric analyses of publication patterns at Cornell. This work has involved postdocs Laura Forlano and Saeko Nomura, graduate students Xuan Zhao and Shion Guha, and collaborators Connie Yuan, Geri Gay and Caren Heller. It is supported by a 2-year NSF award, on which I am the PI.
I have also studied collaboration and coordination in range of other settings. By studying collaborations in the real world, it is often possible to not just learn lessons, but also have a rewarding impact on people’s lives. In this vein, I have collaborated with Ron Baecker on the design and evaluation of a social gaming system for senior citizens, and with Ravin Balakrishnan on systems using mobile phone projectors to enhance education in rural India and to increase social engagement in public spaces.
Birnholtz, J., Forlano, L., Yuan, Y. C., Rizzo, J., Liao, K., Gay, G., & Heller, C. (2012). One University, Two Campuses: Initiating and Sustaining Research Collaborations Between Two Campuses of a Single Institution. Proceedings of iConference 2012, 33-40. [PDF}
Birnholtz, J.P. (2006) What Does It Mean To Be An Author? The Intersection of Credit, Contribution and Collaboration in Science, Journal of the American Society for Information Science and Technology 57(13), pp. 1758-1770.
Birnholtz, J., Bietz, M. (2003) Data at Work: Supporting sharing in science and engineering, Proceedings the ACM Conference on Supporting Group Work (GROUP ’03), Sanibel Island, FL, Nov 9 – 12
Theme 4: Architecture, Location, Attention and Self-Presentation
Given the importance of locations and their demarcations in systems like Foursquare, I have recently become interested in systems that use the GPS location features on mobile devices, but do not share this in the form of traditional locations (e.g., names of places). One popular example of such a system is Grindr, a location-aware mobile dating application for the gay community. A unique feature of Grindr in comparison to other location-based tools like Foursquare is that it shows how far away other users are (e.g., in feet or miles), but does not name their actual location.
The effect of this, we argue, is that Grindr “co-situates” a range of individuals in a single virtual location. By co-situation, I mean the assemblage of individuals with multiple goals (ranging from idle chatting to arranging anonymous sex) and from multiple physical locations. While all designed spaces facilitate co-situation in some sense, location-aware mobile apps like Grindr do so importantly differently.
Where physical spaces (at any scale: room, building, neighborhood, etc.) are defined by clear boundaries and necessarily contain all of the people within them, location-aware apps like Grindr are defined instead by the number of nearby users displayed and the density of the geographic distribution of these users. In other words, the users one sees when signing into Grindr from a location are not influenced by geographic boundaries (walls, borders, etc.) as much as by how many others are using Grindr nearby. If, for example, one is in a city and there are 100 users within a mile, one will see others only within a mile. In a more rural area, however, one sees people who are tens or even hundreds of miles away. Thus, the effect of co-situation is that other users visible on Grindr may or may not have similar goals, expectations or norms, and one may or may not be able to discern where they are. In this stream of work, I aim to explore the effects of co-situation in Grindr and other systems on people’s interactions and self-presentation behavior.
The goal in this research is use co-situation as a vehicle for understanding the interplay between location and architecture that is opened up by location-aware technologies, and the relationship between this interplay and people’s online interactions.