Reflecting on Method: A ‘new media’ arts experience to probe technology set-ups
The healthcare domain has become a popular area of research in Information Systems but the complex and critical nature of many medical environments raises many practical questions of method. In the reflective spirit of this meeting, I present for debate an unusual method for probing technological alternatives using a ‘new media’ arts experience when neither in situ nor experimental means are appropriate.
The research problem explored in this paper is inspired by a previous ethnographic study comparing ward round interaction around paper versus electronic patient records in the intensive care unit (ICU) of a specialized cardiothoracic hospital. The patient record is the central focus of the ward round, which brings together 8-10 staff from a range of disciplines and domains of expertise each day to assess jointly the patient’s condition and required future care. In order to provide good medical care, the team must collaborate, or to be more specific, carefully negotiate the interaction, so that all appropriate information and expertise are shared, but no time is wasted. The aim of the previous study was to understand how the negotiation of the interaction differed between the two records.
Although many studies of technology describe interaction using Conversation Analysis, (e.g. Heath and Luff 2000), we chose to describe the non-verbal aspects of the interaction negotiation based on the arguments of Koschman and LeBaron (Koschman and LeBaron 2004). We drew upon the work of Kendon and his theory of the F-formation system (Kendon 1990), which describes how groups form, adjust their formation, and disband, when interacting. His numerous examples clearly demonstrate that formation adjustments reflect the negotiation process of interaction in semi-mobile, co-located teams. Form-factor of the technology used, however, can also influence formation. Figure 1, for example, contrasts the formation of a ward round team using a paper patient record (left) and an electronic patient record (right). There is tension between formation changes that occur as a means of interaction negotiation and ones that occur because of technology form-factor. It is this conflict and how teams solve it that this study aims to probe.
In the hospital context of the example pictured in Figure 1, the form-factor of the electronic patient record, a 19” display at the end of the patient bed, restricted the formation of the team and consequently the participants’ ability to feel part of the team and to utilize non-verbal behaviours in negotiating the interaction. The solution to this problem, caused in part by static display devices, might be the use of mobile devices that allow people to configure themselves as necessary. However, it is possible that using mobile devices would create new and equally problematic difficulties. To begin this investigation on the effect of display type on the negotiation of interaction, we look at what people do at the two extremes, posing the following research question:
How does a team negotiate interaction differently when using a large display as opposed to individual, small displays?
Hospitals, particularly intensive care units, are difficult environments in which to do pro-active (as opposed to observational) research, such as a technology probe or prototype, as any upset can be life-threatening to a patient. Unable to explore the above question in the hospital environment, we looked for ways to do so in the laboratory. Assessing collaboration with standard experimental methods, however, carries a substantial risk of not being externally valid (Hawkins 2004). Exploratory experiments (e.g. Rogers and Lindley 2004) avoid this problem to some extent, but still have limitations. After more specifically outlining the problems with experiments, we propose a ‘new media’ arts experience as an exploratory method that addresses some of these limitations and takes us a step further towards the successful assessment of such complex interactions as collaboration in the laboratory.
Standard experimental methods when used to assess co-located collaboration have two significant problems. First, choice of measures necessarily imposes judgement on what constitutes good collaboration. Many collaborative applications are critiqued on the equality of participation (e.g. Rogers et al. 2004; Morris et al. 2004). However, “equality of participation” may or may not be a measure appropriate to an environment. Second, even if the choice of measures can be justified, there remains the problem of interpretation. For example, the common quantitative measure of speech equality may say more about the process of the group than the effectiveness of the collaboration. As Rogers et al. (Rogers et al. 2004) point out, some group members might chose to present alternatives on the display without saying much about them, making them strong participators but ‘under-speaker’.
Using Rogers and Lindley’s (Rogers and Lindley 2004) excellent study comparing collaboration around vertical and horizontal displays as an example, we can see how exploratory experiments offset some aspects of the above problems. This study uses a mixture of quantitative and qualitative measures to compose a picture of how collaboration occurs in three conditions. Quantitative measures, such as number of suggestions per condition, are contextualized and often explained with qualitative measures such as role-switching. This approach addresses the problem of interpretation by making the criteria explicit, but only partially addresses the problem of judgement. The aim of this experiment is to test a specific assumption - that shared surfaces facilitate collaboration - and produce a “yes” or “no” answer. Although the analysis presents a much more complex description that may generalize more easily to other situations, the result is limited to a certain, unspecified, environment.
A further problem with both standard and exploratory laboratory experiments is that the tasks are contrived. The hospital environment has a large number of social variables which affect interaction. On the one hand, there is a strong hierarchy that can make doctor-nurse communication a touchy issue. On the other hand, the medical staff care deeply about doing their best for each patient, giving them strong motivation to solve conflicts of communication if they arise. Although re-creating the complexity of medical interaction in the laboratory would be unachievable, having the participants passionate about the end product of a task is important to ensure that there is a strong motivation to solve interaction conflicts.
‘New Media’ Arts Experience
Lucy Suchman (Suchman 2007) argues in Human-Machine Reconfigurations that ‘new media’ arts is a fertile ground for exploring complex human-computer interaction issues, such as the one central to this paper. Her discussions lead the reader away from the prevalent computer science occupation with determining and debating the agency of the ‘smart’ machine towards a depiction of agency as a construction made by humans using machines. She suggests that while AI as a field is occupied with the former debate, many ‘new media’ artists are on the forefront of exploring how humans give machines agency. As artists, their approach is not reasoned possibility, but one to provoke an experience in viewers so that they explore the boundary for themselves.
As researchers, we can take advantage of the personal exploration of an ‘interaction problem’ that we pose during a ‘new media’ arts experience, to gain as our ‘result’ a picture of the solution space – an approach depicted in Figure 2. This alternative attitude both to gaining ‘data’ by asking participants to solve the problem and to producing a ‘result’ by collating those solutions addresses the issues with laboratory experiments concerning ‘contrived tasks’ and ‘judgement of collaboration.’ The study described in the following section is a practical realization of this theoretical idea. Because the artistic experience is housed in a laboratory, the study may resemble a conventional experiment, but the differences will be clearly pointed out.
(1) K. Hawkey, Workshop Notes: Methodologies for Evaluating Collaboration in Co-located Environments. http://www.edgelab.ca/CSCW/Workshop2004/WorkshopNotes.pdf
(2) C. Heath, P. Luff, Technology in Action, Cambridge University Press, Cambridge, UK, 2000.
(3) A. Kendon, Conducting Interaction: Patterns of Behavior in Focused Encounters, Cambridge University Press, Cambridge, UK, 1990.
(4) T. Koschman, C. LeBaron, Reconsidering Common Ground: Examining Clark’s Contribution Theory, Proc. of ECSCW 2003, 81-98.
(5) M.R. Morris, D. Morris, T. Winograd, Individual Audio Channels with Single Display Groupware: Effects on Communication and Task Strategy, Proc. of CSCW 2004, 242 – 251.
(6) Y. Rogers, W. Hazlewood, E. Blevis, Y.K. Lim, FingerTalk: Collaborative Decision-Making Using Talk and Fingertip Interaction Around a Tabletop Display. Proc. CHI 2004, p. (TO DO).
(7) Y. Rogers, S. Lindley, Collaborating around Vertical and Horizontal Displays: which way is best?, Interacting With Computers, 16, (2004) 1133-1152.
(8) L. Suchman, Human-Machine Reconfigurations, Cambridge University Press, Cambridge, UK, 2007.