The User-Centered Design Process: Novice Designers’ Use of Evidence in Designing from Data

Author: Erin Friess
Degree: Ph.D. in Rhetoric, Carnegie Mellon University, 2008

 This dissertation investigates the connections between rhetorical theory and usability practice within the decision-making meetings of novice designers. In particular, this study investigates the evidence used by the novice designers, specifically evidence borne out of user-centered research, to support or refute a claim. For one year, the decision-making meetings of a group of approximately 20 novice designers who were working on a real task for a real client were recorded. This group was dedicated to the principles of user-centered design and had been charged by their client to do more user-centered research than is typical for a design project.

The year-long descriptive case study reveals two major results. First, over 40% of the claims made by the group have no spoken support whatsoever. Second, when evidence is given to support a claim, about a quarter of that support given in a manner that references an authority, a written text, or a usability finding (a type of evidence I modify from Kuhn (1991) to call “fact-based evidence”). Instead, the majority of the time, these novice designers use hypothetical stories or designer opinion (what I call “pseudoevidence”) to support a claim. This reliance on pseudoevidence is surprising, given that a large portion of their time was dedicated to activities associated with user-centered research—activities that would ostensibly produce ready-made fact-based evidence. Further, though these designers claim to be advocates of user-centered design, their lack of fact-based evidence suggests that these novice designers may be doing something more akin to designer-centered design than user-centered design.

While there are many possible reasons why this group relied on pseudoevidence, here I analyze two possible reasons. First, I investigate this group’s use of high-order abstractions in a particular construction that McGee terms “ideographs.” I find that one specific community-established ideograph is used almost completely within pseudoevidence, draws discussions quickly to a close, and always enables the speaker to have his or her claim adopted by the group. Thus, it appears that this ideograph located within pseudoevidence is a powerful persuasive device. Second, I investigate pseudoevidence in terms of face (Goffman) and politeness (Brown and Levinson). While I anticipated that fact-based evidence would be more face-threatening than pseudoevidence (leading the group to use pseudoevidence more frequently), I find that fact-based evidence and pseudoevidence are equally highly face-threatening for this group. I suggest that for this group being inefficient is more egregious than threatening face. Therefore, pseudoevidence, which is perceived by this group to be more efficient than fact-based evidence, is used more readily.

This research is an early descriptive assessment of novice designers’ use of evidence. Clearly, more research is needed to determine if this group’s use of evidence is idiosyncratic, to develop other explanations for the predominance of pseudoevidence, and to identify the implications of these findings on user-centered design theory and pedagogy.