Proceedings of HF 2002, Nov. 25-27, 2002, Melbourne, Australia

Managing Innovation in Scenario-Based Design

Steve Howard¹, Jennie Carroll¹, John Murphy² and Jane Peck²

Keywords:  Scenario-based design, contextual scenarios, innovation

Abstract

Scenarios are often proposed as a means of managing innovation in use-centred design.  Here we discuss ‘how’.  Scenarios allow design spaces to be mapped, both marking and pushing the boundaries of those spaces, and charting paths through the conceptual terrain therein.  They achieve this through facilitating two modes of design discourse (backward and forward chaining), and mapping the granularity of the design conversation.  

1.      Introduction

In HCI we are beset with contradictions and paradoxes to such an extent one wonders how we get anything done!  Let’s consider just two.

Our past informs our futures.  We face an apparent dilemma captured in the phrase ‘use-centred innovation’.  Bødker and Christiansen nicely restate this in terms of ‘border control’.  Designers “find themselves caught in a dilemma between awareness of tradition and orientation towards transcendence: on the one hand starting out from the praxis and history of users in question, on the other hand making sure that something qualitatively new gets shaped in the process” (Bødker and Christiansen, 2000).  It is this dilemma that we address below, and the role that scenarios can play in pulling us towards the future, in ‘forward chaining’.

Our futures become our past.  Paradoxically, we design for a situation that, in part, our innovations and their use will change in ways that are not altogether predictable.  Dahlbom and Ljungberg are pointed in their challenge, “Why, unless you are a historian, describe in detail a work practice that will soon be replaced due to new technology?” (Dahlbom and Ljungberg, 1998).  Use clearly is not stable; once we have moved from history to innovation we cannot long be sure that the world we have designed for will remain the world that the user experiences.  This is only partly captured by the ‘task artefact cycle’ (Carroll, Kellogg and Rosson (1991), Howard et al, (2001)), as the issue is not simply that the artefact and task exist in a reciprocal relationship, but that the user is constantly and actively, if not always deliberately, reformulating the nature of use.  It would be more accurate to think more broadly in terms of deliberate and non-deliberate changes in the interrelationships between the user and their work practice, evolving over time.  Our previous work has grappled with the processual nature of use (Carroll et al, 2002) and in this paper we address the role that scenarios can play in imposing the future on the present, in understanding the impacts through ‘backward chaining’.

In choosing to work within a scenario-based tradition we are, like Bødker and others, elevating the importance of the designer’s experience of the user, and their world, above that of technical descriptions of systems.  We wish that designer’s experience were rich and multilayered, and included the subtle nuances of use difficult to predict and record with technical languages. We wish that designers’ experience of the present helped them better understand the relationships between the past and the various possible futures.  We are seeking a practical aggregation of narrative based design, participatory design and theatre techniques such that designers can be immersed in rich but malleable depictions of future use (Howard et al, 2002a and 2002b).

The next section introduces our variant of scenario-based design, Contextual Scenario-Based Design.  In section 3 we consider Contextual Scenario-Based Design from three angles: riding the boundary between experience and expectation, stretching and manipulating that boundary, and monitoring the coverage of design options contained within the boundary.  Section 4 brings the paper to closure by revisiting the role of scenarios in use-centred envisionment.  

2.      Contextual Scenarios

Contextual scenarios reflect fragments of use, but fragments that describe more the context within which practice takes place rather than the goals or tasks of those actors involved in the practice.  Contextual Scenario-Based Design is the use of those contextual scenarios in seeding theatrical performances that ‘act-out’ the relationship between technology and future practice, and provide designers with a way of extracting meaning from those performances.  We have discussed contextual scenarios elsewhere (Howard et al, 2002a).  To illustrate the concept, figure 1, a traditional scenario, is recast as a contextual scenario in table 1.

Figure 1: Example Problem Scenario

Contextual scenarios can be summarised as:

·         Possessing a general theme. In our work contextual scenarios are written on the basis of previous ethnographic work.  In the problem scenario depicted above, Francesca experiences her activities as being fragmented and broken across a range of situations and technologies.  We wished to support Francesca in integrating her diverse activities through tools for social management and organization.

·         Identifying the primary actor (Francesca in this case) in each scenario, segmented according to variables of interest, e.g. gender, age, culture, experience with technology, knowledge of the situation, interpersonal knowledge etc.

·         Loosely describing the purposes and activities brought to and discovered in the situation by the primary and secondary actors. 

·         Identifying who else is involved in the events depicted in the scenario.  We should stress that the roles identified as primary and secondary are under constant and reciprocal redefinition during both design and use. 

·         Capturing the overall purpose that colours the events in the scenario and drives the actor’s individual purposes.  Such overall purposes may include shopping, leisure, rest and sleep, work, education, transit etc.

·         Describing the location or situation in which the events unfold.  We favour descriptions that are rich in such factors and include information on the physical, technological and social context.

·         Suggesting the general form factors relevant to the design discourse. 

Table 1: Problem Scenario recast as a ‘Contextual Scenario’

Not all contextual scenarios will include all aspects of the above classification but it is a useful orienting checklist.  Depending on the type of envisioning being conducted, scenarios may load more heavily on some categories than others. 

3.      Contextual Scenario-Based Design

Contextual scenarios intentionally lack detail- their aim is, in part, to seed a discussion between the actors and the designers that ‘discovers’ the detail.  They are also rich in context and global purpose- this is no accident as we wish to counterbalance the overriding influence of the artefact and task, which is found in many traditional applications of scenario-based design, with a rich and malleable depiction of use-in-context.

In typical usage, scenarios are walked through by users and designers (Carroll, 2000).  Contextual scenarios are intended to support acting-out and hence play a role more similar to stage directions than detailed scripts.  The use of contextual scenarios is described in Howard et al (2002a).  Briefly, actors draw a performance from the seeds provided in both contextual scenarios and ‘props’ (described as ‘With what’ in the contextual scenario of Table 1).  Props can take many forms but are typically blank and malleable form factors that, during the performance, are endowed with characteristics and capabilities (see Howard et al 2002b).

The performance portrays imaginary situations of use but does not proceed uninterrupted- the design team is able to interject with ‘constraints’ that are drawn from the scenarios.  Constraints typically play two roles.  They increase the stakes (the importance of achieving the goal) or introduce hurdles for the actor to overcome (‘do it faster’, ‘juggle more conflicting goals’ etc).  Opportunities for innovation are identified in the interplay between the performance and the design discussion.

From the perspective of innovation the roles of contextual scenarios, and the associated acting-out, include:

·         Riding the boundary between ‘experience and expectation’: A contextual scenario captures, albeit in fragmented form, both current or intended use and the agreed upon or emerging meanings that a participatory design team hold to at a given point in time.  Bødker and Christiansen refer to this role as the ‘boundary object”.  We like to think about moving from experience to expectation as ‘forward chaining’ the design discourse, and understanding the implications that expectations have for current practice as ‘backward chaining’.  This issue is considered in section 3.1.

·         Allowing the boundary to be moved in deliberated ways:  Once established, a boundary can be manipulated.  One’s focus may narrow or broaden, for example moving from individual user-technology experience to more broadly examining the entire group.  The focus may also shift, for example moving from one user cohort to another.  Elsewhere (Howard et al 2002a) we have discussed the use of scenarios in two ways: top down (where the actor has freedom in the situation and that freedom is systematically withdrawn) and bottom up (where the actor is heavily constrained initially, but given increasing freedom during successive performances as the devices become more capable).  This issue is considered in section 3.2.

·         Monitoring the coverage of the space therein: Boundaries contain spaces of design possibilities.  Scenarios, like design rationale (Moran and Carroll, 1996), allow a record to be kept of the coverage achieved in any design session.  This issue is considered in section 3.3. 

In addition to such ‘instrumental’ purposes, scenarios provide a shared semantics that is under constant renegotiation within the participatory design team.  For those concerned with innovation, scenarios should be vignettes located in space and time but stripped of any detail that might firmly direct the discourse.  In reflecting typical work practice (used broadly to include practices related to social activities, leisure etc) scenarios should hold that work practice up to question.

3.1.     Boundary Riding

Figure 2 captures the relationship between experience of current practice and expectations of future practice, in terms of two processes.  In backward chaining the impact on current practice is examined from the perspective of some idealised contextual scenario.  Constraints may be provided by understanding current practice but the design discourse proceeds so as to impose the design on the current situation of use, thereby exploring the design’s impact.   

Figure 2 Backward and Forward Chaining

Forward chaining moves from an understanding of current practice to imagined future use, grounding the emerging design in the current usage situation.  Here constraints are provided directly by current practice, and the design discourse proceeds so as to remove those constraints through design innovation.  This is illustrated in the data shown in Episode 1 below where in response to a current problem faced by the actor (purchasing music) a solution was proposed by a member of the design team (“would it be useful to have audio?”).

In both modes a ‘chain’ of constraints (resulting in a series of modified contextual scenarios) connects current and imagined work practice, each link in the chain reflecting either increased knowledge about the design or its likely impact.  In practice, backward and forward chaining are tightly interleaved resulting in design discourse that attempts the co-evolution of an understanding of both current and future use.

What is missing from this account of scenario-based design as chaining is the flavour of estimation and guesswork that characterises the process.  Bødker and Christiansen prefer to think of this as abduction.  Our own view is that it may productively be seen as creativity and idea generation and this is discussed briefly in section 4. 

3.2.     Moving, stretching and shrinking the boundary

Contextual scenarios contain the raw ingredients of a tension between technology and practice.  They allow us to manipulate, through their enactment, that tension and thereby stretch the boundary of the emerging design space.  There are decisions to be made in how such tensions should be structured.  We have given actors props that can do ‘anything you want it to’ (similar to Iacucci et al’s (2000) ‘magical thing’) and then, through iterations of the performance, we have restricted the features and functions that the props possess.  In examining the effects of these restrictions on successive performances we establish the extent of the creative boundary.  We have also started the performance with a restricted prop, where capabilities are carefully and explicitly limited, only later to be relaxed as the performance evolves.  This is reflected in figure 3 (adapted from Howard et al, 2002a) where over the course of the design discourse, the prop is moved from science fiction (where the prop is endowed with many highly capable features) to plausible fiction (where the features of the prop might be so mundane as to be currently available).

The essence of our process lies in being able to support use-centred design teams in exploring the reciprocity between the technology (with props acting as a general proxy for technology) and practice (modelled as constraints on the performance and captured initially in the contextual scenarios).  As shown in figure 3, top down contextual scenario-based refinement involves jointly manipulating the technology and the practice.  In figure 3, the device capabilities are rich early on in the performance, and the costs of failure (i.e. the stakes) are low.  As the performance proceeds and the designers gain insight into the questions at hand, contextual constraints are introduced, the actor is restricted, and the stakes are raised in ways that are interesting for design.  This is done to stress the practice/technology (i.e. the performance/prop) interaction, and forces the design team to innovate in order that the actor may overcome the hurdles that are being imposed.

For illustration, episode 2 illustrates an interaction that moves from science fiction (as suggested by the actor ‘scanning her reaction to the gift’) to currently available solutions (‘matching previous purchases with the options that are coming up’).

We find exploring current technology to be a useful way of learning about work practice, identifying its transcendent features and key issues.  Though it may not reliably yield innovative technology, the design team’s sensibility to the user in her context is greatly improved.  Episode 3 provides an example of the performance revealing an intangible technological impact (related to privacy and personalisation) and providing the design team with an opportunity to reason about contextual influence over user activity.

3.3.     Covering the terrain within the boundary

In addition to providing assistance in planning scenario-based design (by suggesting issues for consideration), contextual scenarios as illustrated in table 1, provide the basis of a simple record keeping mechanism for documenting the coverage of the space within the boundary (illustrated in table 2).  The checklist in table 2 is not intended to be complete, indeed we believe it will be varied in ways that reflect local design practices and concerns, however in concentrating on recording both the assumptions held and the discoveries made it aims to capture the push and pull forces as considered during Contextual Scenario-Based Design.

Table 2: Documenting the coverage with a Contextual Scenario

4.      Concluding Comments

Boundary riding, boundary stretching and tracking the terrain therein.  We may have pushed our metaphor too far but seeing use-centred innovation as the confluence of push and pull forces, of experience and expectation, of backward and forward chaining helps us understand the contradictions and dilemmas that beset us.  We are caught between honouring current practice and wishing to take flight with any number of imagined futures.  Scenarios help us negotiate that tension.  Their roles are numerous (see figure 4) but in part they gain their power from allowing us to reason about the preservation of current practice in the revised or imagined practice (A); in assisting in backward chaining the impact of imagined innovations on current practice (B); in forward chaining to imagined practice (C); and in tracking the changes that occur as, during the use-centred design process, current practice becomes future (D). 

Figure 4 Relationship Between Practice and Scenarios

5.      Acknowledgments

This project is being jointly funded by The University of Melbourne through a MRDG Collaborative Research Project, and Novell Pty Ltd through their Customers of the Future programme.  Thanks to the anonymous reviewers whose feedback has improved this paper.

References

·         Bødker, S and Christiansen, E. (2000), Scenarios as springboards in design of CSCW.  In G.Bowker, L.Glasser, S.L.Star and W.Turner (Eds) Social Science Research, Technical Systems and Cooperative Work.  Pub: Erlbaum, Mahwah NJ.  

·         Carroll, J.M. (2000), Making Use: scenario-based design of human-computer interactions.  MIT Press.

·         Moran, T. P. & Carroll, J. M. (Eds.) (1996). Design rationale: Concepts, techniques, and use. Mahwah, NJ: Lawrence Erlbaum Associates.

·         Carroll, J.M., Kellogg, W.A. and Rosson, M.B. 1991.  The Task Artifact Cycle.  In J.M.Carroll (ed) Designing Interaction: Psychology at the human computer interface.  Pub: CUP. 

·         Carroll, Jennie., Howard, S., Vetere, F., Peck, J. and Murphy, J.  (2002) Just what do the youth of today want?  Published in Proceedings of Hawaiian International Conference on Systems Sciences, 2002.

·         Dahlbom, B., Ljungberg, (1998) Mobile Informatics, Scandinavian Journal of Information Systems, 10 (1&2), 1998, 227-234

·         Howard, S et al (2002a) Provoking Innovation: Acting-out in Contextual Scenarios.  Accepted for publication in the British Computer Society Human Computer Interaction (BCSHCI) annual conference.

·         Howard, S et al (2002b) Using ‘Endowed Props’ In Scenario-Based Design.   Submitted to NordiCHI, October 2002.

·         Howard., S., Carroll, Jennie., Vetere, F., Peck, J. and Murphy, J.(2001) Young People, Mobile Technology and the Task Artefact Cycle.  Published in Proceedings of OZCHI2001, Fremantle, November 2001. Pp 63-69. 

·         Iacucci, G., Kuutti, K., & Ranta, M. 2000. On the move with a Magic Thing: Role Playing in Concept Design of Mobile Services and Devices. Paper presented at the Designing Interactive Systems: processes, practices, methods, and techniques. Brooklyn, NY.