Download Postscript file for printing (125 KB)

Abstract

The content of the evolving design depends heavily on negotiation strategies, among other more subtle and ubiquitous social processes. Team members use utterances and gestures to introduce issues of design content and process into conversation and to demonstrate their commitment to partial solutions. Depending on the positioning and level of commitment of their team mates, they adopt appropriate strategies of persuasion. They carefully moderate their commitment to ideas to remain amenable to negotiation. They appeal to common sense, design theories, standard practices, expert practices, user preference and demonstrations with physical hardware in order to persuade. The emerging content of the design is inextricably linked to these collaborative processes.

This paper describes how the design evolves through designers' negotiation strategies and though topic shifts prompted by the design activity itself. This exploration suggests that videotape studies can provide useful process examples for engineering design education.

Modern interdisciplinary design demands that engineers learn to work well in teams. Teamwork requires individuals to express ideas and misgivings, listen, negotiate etc., that is to collaborate. Engineers need to be aware of various characteristics of collaboration so that they can identify successful and poor strategies within their own work practice.

Engineering design students at Stanford University, commenting on the usefulness of design process models and prescriptive design methods in their seven-month design project, noted that while the models and methods told them what to do, they provided little insight into how to do it. In particular, students wanted more help with group dynamics, solution development and project management.

The Delft Protocol Analysis Workshop presented us the opportunity to examine a two hour long videotape of a team of practising designers developing a preliminary solution for a product that mounts luggage on to a mountain bike. Videotape allows us to take a close, careful look at how the moment to moment activity steers the course of the design solution. Through repeated observation, analysis and discussion we can develop a better understanding of how professional designers do such things as collaborate, develop design solutions and manage their work.

1. Video Analysis Method

The analysis team consisted of four researchers with engineering design and design research backgrounds. Watching the tape for the first time to make general observations, all team members found it difficult not to engage in designing the product! We were drawn into judging the content of each proposal. With each viewing, more facets of the activity appeared. In particular, it helped us to watch the tape with an interdisciplinary team of social scientists, anthropologists, computer scientists and engineers. The tape served as a catalyst, provoking recall of hypotheses based on experience from outside the tape as well as from within. Repeated video watching determined which hypotheses were validated by several occurrences of supporting events in the tape.

The time limit and restricted setting of the protocol for the purpose of video taping emphasized certain aspects of design such as time pressure and decision making, whereas other aspects such as information gathering, organizational context, the ability to mull over ideas or engage in opportunistic solutions were restricted or removed from context. All participants felt out boundaries of what was within the "rules" to a greater or lesser extent. Collaboration was chosen for analysis both because of our interest in this area and because it was considered relatively insensitive to the protocol design. The analysis explores how the group works together under the given conditions, whatever they might be. Also, it became evident that social interaction strongly influenced the content of the emerging product.

2. Analysis Overview

Section 4 describes how the design discussion transitions from one topic to another. It describes the designers' negotiation strategies. It describes how they express their commitment to proposed solutions.

Section 5 summarizes our analysis of the collaboration.

3. Roles and Solution Preferences

Throughout the early stages of the session Kerry is engaged by the existing hardware and concepts. She seeks to ground the problem by examining the existing hardware and concepts in detail, (Transcript 2) (comments 1,3,6,8 ). John observes this process and then bids to work back at a higher level to get a broad view of the process they will follow before diving in (comment 9).

Transcript 3 illustrates how John begins to seek out ambiguity in the problem statement (comments 5,9). He negotiates for flexibility in wanting to consider internal and external frame backpack designs as well as all types of mountain bikes (comment 22). Kerry seeks to ground the problem, countering that the problem statement suggests they only consider the Hi-star external frame backpack (comments 4,12,14,16,18). Ivan's role is emerging as one of arbitrator. He considers both positions and falls to either side. As we enter Transcript 3 they are discussing the backpack...

Kerry has managed temporarily to gain an agreement from John and Ivan that they design for the HiStar backpack. But in the last statement of transcript 3, John, in agreeing to designing for the Hi-Star opens up a bid to design for most touring bikes (comment 22). This is counter to an earlier suggestion in 00:08:00 by Kerry to "make it a special mountain bike so it could have the stuff required attached [something] to it," tacitly agreed to by the group.

Kerry's preference to try to pin down part of the solution is repeated throughout the tape, as is John's preference for preserving ambiguity. Table 1 illustrates their design solution preferences for key issues.

The groups roles are summarized in Table 2. Clips throughout the paper will illustrate the development of these roles. Ivan gradually emerges as an organizer, time keeper and as an occasional arbitrator between John and Kerry. He manages the white board lists and generally keeps the group on track. Are these roles predetermined or adopted? From the tape we cannot know whether the characters play the same roles in other design situations. Nor can we tell how the emergence of one role depends on that of another. However it is likely that a certain amount of compensating occurs, each designer seeking to fill in gaps they perceive in the group approach. Kerry exhibits behavior noted by [Guindon, 90] to be consistent with that of domain experts, quickly pruning the search space to promising solutions. In section 4 we characterize how the individuals reconcile their different positions to engage in effective collaboration .

4. Making Progress - Foci and Transitions Mediated through Social Interaction

Many researchers have characterized design as being opportunistic or chaotic, moving fluidly between requirements and details in response to information or ideas uncovered that are worthy of immediate exploration, [Guindon, 90]. The session is replete with designers abstracting from the context to gain perspective and seeking context to ground the design with affordable, manufacturable alternatives. They develop requirements by considering candidate solutions, user scenarios and tinkering with hardware. For example, requirements for easy attachment, low center of gravity and strap containment all emerge from the context of the problem, by working with the backpack around the bike or examining the user specs. (The structured design methods do not address how to seek out alternatives and generate solutions from the context of the problem.)

4.2 A Scheme to Describe Design Progress Mediated through Social Interaction

As the design team negotiates the problem space, each designer makes bids to have issues they think important discussed by the team. Having called focus on an issue, the other designers might engage in the focus adding ideas toward a partial solution. The content of design evolves then through discussions adding incremental solution additions, use scenarios, justifications and information seeking questions. The designers acknowledge other contributions with nods and short phrases or they call into question an aspect of a proposed solution. They align themselves with various aspects of the evolving solutions and approach and distance themselves from others. We looked for evidence presented in the videotape that designers were happy with the alignment of the team and if not, how they sought to change it.

We characterize the design discussion as focusing in on issues and then transitioning topics. However, there is evidence that the designers are continuously engaged in multiple activities at different levels. Although they focus in on issues, they continuously monitor the progress of the solution from the point of view of various requirements and solution alternatives. They reflect on their course of action, monitoring and modifying their process. And there is evidence that each designer they monitors their teammates utterances and actions on a moment to moment basis and moderates their talk accordingly.

It is difficult to represent such a rich process in a scheme of categories. However we have chosen to do so to try to illustrate how the evolution of the design content is governed by the social interaction in the team.

A scheme to describe how the design discussion focuses and transitions mediated through social interaction is proposed below. Before offering an example of the use of the categories in this scheme, a few notes regarding their status and purpose are appropriate. The categories draw upon those used by other researchers, such as Guindon's `partial solution' term but are not an exact copy of any one set, [Baya et al, 92], [Stauffer and Ullman, 87], [Christiaans and Dorst, 92], [Tang and Leifer, 89], [Minneman, 91]. Many different coding schemes are possible, each appropriate when taking a particular point of view on the subject material. We make no claim that these categories are entirely mutually exclusive; many of the phrases in the transcript carry some force in more than one of the categories. However, the ambiguities are few enough that classification and interpretation provide useful insight into the dynamics of the group and problem.

Our presentation focuses on a five minute piece of transcript seventy minutes into the design session. The team members are attempting to persuade each other of the merits of their suggestions on rack attachment. We begin by looking at how the content of the discussion evolves. Then we examine the social interaction at work.

4.3 Focus: Building a Partial Solution

The segment begins with several calls-for-focus on the issue of where and how to attach the rack to the bike (Transcript 4a) (comments 1,3). Kerry then outlines a basic proposal for a partial solution (comment 4), with a use scenario to justify it; this is in part a restatement of one of the options that's already been identified. J and I offer acknowledgments (comments 5,6) indicating that they now share her focus on this partial solution. Then, in a series of statements (comments 7,9), K adds details and justifications to this proposed solution, which I and J acknowledge. J then calls a piece of the proposed partial solution into question (comment 10), offering some alternatives to a detail that's being considered; K and I bolster the initial proposal (comment 11,12) with several references to both scenarios and to broad principles such as strength and reliance on standards. Acknowledgments (comments 13,14) indicate that each continues to be engaged by the focus.

4.4 Strategies of Persuasion

Notice that John and Kerry demonstrate commitment to their ideas and adopt persuasive strategies. The designers masterfully invoke the support of neutral parties such as common sense, higher principles or theories, and expert or standard practices to support their opinions. These serve to depersonalize the debate, in addition to being means of persuasion and explanation of rationale.

Common sense:
Appealing to common sense is a tactic to build support for an idea from commonly held beliefs. Kerry's suggestion " it makes a lota sense" in the Transcript 4a (comment 7) prefaced with "I think", suggests she is embarking on a persuasive strategy but is open to negotiation. John counters with the same strategy in (comment 10) suggesting of his idea: "does it makes more sense?"

Higher principles/ requirements/ theories:
Appealing to higher principles, theories or requirements also appears frequently in designers' persuasion. In comment 9 Kerry suggests "I think we can do that reliably"; in comment 12 "this is strong, Allen Wrenches are standard " John counters with "do we really wanna use these lugs for speed of disassembly," in comment 10; and "it just doesn't seem real elegant to me" in comment 21. These higher principles or broad requirements serve both as explanation and means of persuasion.

Expert Witnesses/ neutral parties:
A third persuasive strategy is to appeal to established methods by established experts. Kerry in comment 9 suggests "I think we can do that reliably kind of a la Blackburn rack use their kind of [attachment]" An even stronger appeal is made outside this section of transcript in the 62nd minute "I mean if you really need a adjustment I think all these Blackburn racks would have adjustments."

In Transcript 5, J is seated on the bike and K is experimenting with backpack position, while I lists on the white board. I repeats sentences as he writes on the white board (comments 4,6) explicitly communicating his interpretation of what J and K are doing. K places the backpack in several positions suggested by J, offering advice like "see if you can steer" (comment 8). In transcript 5, the designers offer little commitment to each statement prefacing statements with "maybe" "what about", "would it be too funky to". They are proposing suggestions that don't necessarily constitute opinions and paying extensive attention to communicating their actions. Calls to question, accompanied by rationale, are quickly agreed to (comments 1,2,3). The designers complete each others sentences perhaps to communicate they understand the others concerns. There are no justifications or reiterations of position.

Once designers begin to preface statements with "I think" or "my opinion is", they are clearly in the realm of offering an opinion. However depending on the context or intonation, these can either be interpreted as "I think, but I don't know for sure" or "I think and I don't care what anybody else thinks." Thus there are several linguistic cues to determine the speakers level of commitment to a proposition, as described in [Schiffrin, 90]. As they begin to embark on strategies of persuasion, parties may lessen commitment in the interests of negotiation using phrases like "it seems to me" and "the way I see it." "It seems to me" rather than "I think" somewhat lessens the control of the speaker, serving to depersonalize the debate.

Skillful use of persuasive strategies, paying attention to communicating assumptions and appropriate moderation of commitment, maintains the negotiation process.

4.6 Opportunistic Strategies of Persuasion - Give and Take

We return to transcript 4, where the group is still engaged in focusing on rack attachment. In Transcript 4a, Kerry backed by Ivan has tried to persuade John that they should use the standard hex bolt attachment method. John remains unconvinced (Transcript 4b) (comments 18,20), using the higher principle of elegance as a reason for his doubt, and Ivan and Kerry again offer justifications for the initial proposal (comments 21,22). Each designer is maintaining their position. There is evidence of an impasse.

During this impasse, Kerry bids to shift the debate to more neutral territory (comment 23). She assumes tacit agreement on the issue of primary attachment with hex bolts (using Allen wrenches) and then seeks to address John's concern about elegance elsewhere in the design, offering several possible alternative solution details for secondary attachment. John joins the new focus with an alternative partial solution to secondary attachment (comment 24). Kerry has been successful in shifting the debate by using an opportunistic strategy of give and take. Seeing the discussion had reached an impasse she sought to move to another part of the design space, taking tacit agreement on primary attachment and giving ways to address the elegance concern through other means.

Topic shifts to avoid or postpone conflict
In addition to an opportunistic means of persuasion through Give and Take above, Kerry's topic shift away from the primary attachment to secondary attachment serves to avoid or postpone conflict in the group, since the group is at an impasse. Several other types of topic transition are identified in this section.

Ideas gracefully lose steam
The group encourages John's partial solution for secondary attachment, but he then calls it into question and no-one protests. This appears to be mutual recognition that the idea is not worth further consideration at this time. Rather than state as much, the idea simply loses steam. (Processes such as listing also tend to lose steam as ideas become exhausted.) This opens an opportunity for a bid for focus which Ivan takes and so a transition of focus occurs. (Transcript 4c) Ivan calls for focus (comments 31,33) on a discussion of rack width, which has caught his attention as John gestures with the backpack frame. He develops a concern that the rack might be too wide for general riding, drawing upon the context of rack use.

Topic prompted transitions
John follows up with a question confirming Ivan's concern and Ivan bids to look for marketing information to answer their concern. He moves to the table to look for information on marketing research, and the group follows. As Ivan searches for the information, John also calls for focus on a current use scenario of bungee cords (comment 42) drawn from the comparison with the partial solution the group was working on; Ivan pursues his focus bid in parallel, rather than giving full attention to John's proposal. John laughs occasionally at his suggestions and prefaces them with "maybe" indicating he has a low level of commitment. Kerry offers acknowledgments and supporting comments but does not engage in developing the idea. There is some sense of exploration in the conversation but nobody bids to change the topic except for John himself, indicating the group is not unhappy with the process or that perhaps they are relaxing. One topic seems to prompt another. The current usage of bungee cords prompts the concern that their product must compete with that, leading to the notion that they could purchase rather than manufacture a good solution, leading to a request for information on the manufactured rather than sales cost, leading to an assumption about the manufactured cost as a percentage of the sales price.

Halting to seek information
Ivan does not find information in the marketing survey relating to rack use to resolve the issue of rack width, but this is never explicitly stated within the group which is now following along with John's comments. The group does not return to the issue of rack width for another 25 minutes.

It is worth noting that information seeking often halts a focus in the activity. Further when the designers find information they often do not use it in ways observable to us, and if they do not find it, still, they often return to a different topic. Information seeking serves to broaden the designers knowledge of options, yet it rarely adds to the knowledge space in a predictable way, which perhaps explains why transitions of focus often occur during information seeking.

Returning to Key Issues
The group eventually and repeatedly returns to issues such as positioning, attachment and materials, indicating that these are key issues to them in the design problem. This perhaps indicates their ability to shift topics effectively to stay productive. [Hales, 93] argues that design managers need to `window out' and `window in' `concentrating effectively on the detail, while at the same time keeping the wider context in mind, a crucial aspect of managing engineering design'. In transitioning from topic to topic and yet returning to key issues, we see evidence of how a group manages itself in monitoring the broader problem yet focusing to define details, through social interaction.

4.8 Process Prompted Transitions (Calls to order)
In the short phrases that follow (comments 62 - 74) (Transcript 4d), the group seems to be mentally relaxing, exploring the topic of acronyms for their own amusement rather than exerting themselves to resolve issues raised recently.

Interspersed in this, Ivan makes three repeated bids for a new focus (comments 75,83,85), indicating he is ready to move on and not happy with the current process of the group, one of relaxation. He calls to focus on an issue of process: what should we do next? The group continues to wander but he finally engages them by specifically proposing to strike issues he considers resolved from the white board list, bidding for focus with comment 96 (not shown). `Positioning, have we thrown out the folds in the middle?" The group responds by attending to the issues he raises from the white board list.

4.9 Breakdowns and Arbitration

In transcript 6 Ivan senses an impasse and actively jumps in as an arbitrator to negotiate the conflict between John and Kerry, who are committed to opposing positions. Following an extended discussion in which Kerry has proposed the bike dimensions will not vary enough to merit an adjustable solution, John makes a strong bid for an adjustable solution (comment 1). John supports his opinion with appeals to theory of good human factors and the absent users preference, indicating he is strongly committed to `it should be adjustable'. Kerry shrugs indicating she has stated her position and will not offer any more argument. The stand off is apparent to Ivan who takes upon the role of arbitrator and offers a compromise; `whatever idea we come up with, we can look at ways of making it adjustable'. Since they are not making progress they defer the discussion of adjustability until there is a specific solution to consider and more progress can be made.

Having made a strong statement to which Kerry offers no response, John lessens his commitment with `em, that's my opinion, opinion, not fact,' in apparent effort to lower any tension that may have arisen. John then shifts the topic.

4.10 Contextual Strategies of Persuasion: Stories

There are few stories in this session, perhaps because of the timeframe, but the designers use some techniques of storytelling.

[Schiffrin, 90] offers a framework to describe how stories (or reporting) are used in persuasion noting that story tellers use:

• Selective Interpretation - recounting aspects of the event preferred by the narrator;
• Evaluative devices - highlighting parts of the experience from the narrators perspective, to show the narrator taking an orientation to what is being talked about;
• Deictic shifts - shifting time, place and participants from the conversational world (storytelling world) to the story world;
• Contextualization - framing an event within the story world.

Without even telling a complete story, selective interpretation and reframing of past design experiences or even of the problem statement, (see Transcript 3), and creation of scenarios play an important role in developing, representing and communicating partial design alternatives.

5. Conclusion

Many solution proposals and interpretations of requirements clearly arise from designers interacting with available hardware. They also emerge as part of the ongoing activity. We have focussed on the designers as actors that interpret the hardware, examining how their utterances steer the activity. A compelling analysis would also result from examining how hardware acts as a negotiator to steer the activity.

The design progresses as the group focuses and transitions from topic to topic. Still there is evidence that team members are continuously engaged in monitoring multiple issues at multiple levels of attention. Transitions occur when:

team members seek to shift the debate to another topic,

team members seek to change the process,

prompted by related topics,

topics lose steam,

processes lose steam,

team members stop to seek information.

This is story of one group of designers that we have used to illustrate strategies of design collaboration. There are surely many other methods and interpretations. However the tape provides a valuable means of introspection and reflection for the design student. Watching, discussing and reflecting upon such tapes provides a means for design students to become aware of the variety of productive and counterproductive strategies and processes available to them. The tapes makes these processes available and identifiable. With this awareness it becomes easier to identify when oneself or members of one's own team are following counterproductive strategies. Video tapes offer the opportunity of process examples with the context necessary for the student to gain a fuller appreciation of strategies that work well in design.

Acknowledgments

We also wish to thank the University of Delft Faculty of Industrial Design Engineering and the participating designers for making this protocol analysis possible.

References

Christiaans, Henri H. C. M. and K. Dorst, "Cognitive models in industrial design engineering: A protocol study." 4th International Conference on Design Theory and Methodology. Held: Scottsdale, AZ, USA, Sept. 13-16, 1992. Design Theory and Methodology - DTM '92 American Society of Mechanical Engineers, Design Engineering Division (Publication) DE v 42. Publ by ASME, New York, NY, USA. p 131-140.

Guindon, R., Designing the Design Process: Exploiting Opportunistic Thoughts, Human -Computer Interaction, vol. 5, pp. 305-344 (1990)

Hales, Crispin, Managing Engineering Design, p17, Longman, UK (1993)

Jordan, B. and Henderson, A. Interaction Analysis: Foundations and Practice. The Journal of the Learning Sciences. 1995 Vol 4 No.1.

Minneman, Scott, The Social Construction of a Technical Reality, Ph.D. Thesis, Stanford University (1991)

Schriffin, Deborah, The management of a co-operative self: the role of opinions and stories, in Conflict talk: Sociolinguistic investigations of arguments in conversations, Ed. Allen D.Grimshaw, Cambridge University Press (1990)

Stauffer, L.A. and D.G. Ullman, Protocol Analysis of Mechanical Engineering Design. In Proceedings of the 1987 International Conference on Engineering Design, Boston, Mass, August (1987)

Tang, John C. and L.J. Leifer, Observations from an Empirical Study of the Workspace Activity of Design Teams, in Proc.of the 1st Int.Conf.on Design Theory and Methodology, ASME Des.Tech.Conf. Held: Montreal, Que, Can, Sept. 17-21 (1989)