Critique of the Role of Sketch Finish and Style in User Responses to Early-Stage Design Concepts

Does sketch quality impact how well consumers view a potential product?

Macomber and Yang both theorized that the quality of design sketches impacted consumer response [1]. Concept sketches and computer-aided drafting models of several principal shapes were created to test this theory. The study's authors theorized that consumers prefer more finished sketches and models rather than rough concept sketches.

As new products are composed of a combination of novel or new combinations of existing solutions, consideration must be made to ensure that the end-state product meets consumer demand [2]. This may be accomplished through focus groups and direct client feedback. One of the most expedient methods of immediate client feedback is the presentation of concept models.

Concept models may run the gamut from rough sketches completed in seconds to highly elaborate computer-aided drafting models that require hours to create. Since design processes are iterative, minimizing concept iterations allows for a faster overall solution to be developed. As a result, designs may be continually reworked until a solution that meets engineering intent and consumer demand is achieved [3]. Rework costs time and money, which may result in design cost overruns.

Five levels of quality can define sketches, according to Rodgers et al. The first level is indicated by a simple monochromatic series of lines that do not include shading or engineering annotations. There is only minimal effort towards attempting to show 3d form. The second level uses minimal annotations and increases the detail of the drawing. These sketches do not use shading; however, line thicknesses may be varied to understand contours and dimensions better. The third level adds limited shading and actual numerical dimensions. Usage instructions, functionality descriptions, and general descriptions may be added. The fourth level of sketching includes detailed shading and contour lines. Color may also be used to illustrate aspects of the design. The fifth level involves extensive shading and annotations and may be considered a complete sketch. These sketches are almost always colored. Fifth-level sketches involve the extensive use of lines and will appear extraordinarily busy [4].

Alternatively, sketches may be categorized based on functionality. Some have defined sketches as capturing "fleeting ideas" on paper. Rodgers describes sketches that capture ideas in terms of a tiered system. The first tier is the thinking sketch tier. These sketches are used to guide the design team toward a solution. The second tier is the prescriptive sketch tier. These sketches direct a drafter toward creating a final drawing. The third tier is the talking sketch tier. This living document evolves into a finished product through collaborative exchange [5]. Originally these sketches would be performed in person; however, modern technology now allows for Internet-based collaboration.

Rodgers further defines sketches based on the metaphysical nature of the physical expression of the product. Sketches may communicate the physical nature of a concept from the designer's thoughts. Sketches may facilitate the visualization of familiar objects to expedite the design process. Finally, sketches may record observed objects for design modification [5]. The synergy between these three sketch functions allows the design engineer to conceptualize and communicate design intent to the client.

This study carries merit as significant resources may be expended to create a concept model. Finding the equilibrium point where effort contrasted against consumer preference allows for economizing engineering resources. Dym et al. wrote that design problems are initiated by a client seeking to establish potential consumer benefits [6]. By improving the resource efficiency of the designer-client interaction, higher profit margins can be seen by engineering and manufacturing organizations while passing along cost savings to the downstream consumer groups.

The first shape tested was a generic cube. Two additional shapes were used: a 1970s amoeba-style chair and a Motorola RAZR cellphone. These three basic shapes were sketched to varying degrees of finish. Additionally, generic CAD models were created for these shapes. The same sketch artist was used for all sketches and models. The same basic size constraints were used in all regards to minimize bias.

Macomber and Yang utilized a paid survey service to obtain a representative sample of individuals from the United States. This survey service does not ensure normalization across demographic lines. As this type of survey is highly subjective on user preference with a skew towards technologically inclined individuals willing to complete surveys as a means of supplemental income, it may not be considered representative of the majority of consumers. The authors did not recognize this potential pitfall. This problem could be corrected by surveying various individuals using multiple survey styles.

Generally, this study does not provide any unexpected insights. Finished, high-quality sketches using level 4 sketch criteria are typically more aesthetically pleasing than rough, unfinished sketches using level 1 and 2 sketch criteria. This drives more individuals to prefer finished sketches over unfinished sketches. The selection of sketches over CAD models is less expected; however, unsurprising. This preference may reflect choosing human-created art over machine-defined polygons and splines.
Macomber and Yang also collected responses from the survey respondents regarding how each sketch and model made them feel. While this is highly subjective, this data does allow for limited trend analysis. This was accomplished by limiting responses to single words and categorizing these responses accordingly.

Hand sketches may be assumed to carry a distinct humanizing flavor based on the one-word survey responses. Respondents described hand sketches as “clear” and “detailed.” While this may not evoke much in the way of art critique, it does speak to the mindset of the individuals rating the sketches. Negatively rated hand sketches received words such as "sloppy" and "messy."

CAD models were frequently described as "boring" and "bland" in the study [1]. These ratings were given despite the general assumption that a CAD model is just as clear and detailed as a finished sketch. Based on these facts, it may be inferred that the impersonal nature of CAD models is distasteful to end consumers seeking to understand the product. Design engineers seeking client satisfaction may find it more beneficial to focus early concept development efforts on finished concept sketches over CAD modeling.
Overall, the conclusions reached by the authors held statistical significance. While the research methodology could have been improved somewhat, this does not detract from the results' validity. The conclusions reached also match common sense: individuals prefer finished sketches over unfinished ones.

References:

[1] Macomber, B and M. Yang (2011). The Role of Sketch Finish and Style in User Responses to Early Stage Design Concepts, ASME International Design Engineering and Technical Conferences, August 28-31, 2011, Washington DC USA.

[2] Pahl, G., Beitz, W., Feldhusen, J., Grote, K., 2007, Engineering Design: A Systematic Approach, 3rd ed., Springer Science & Business Media, London.

[3] Jensen, J., 2006, A User's Guide to Engineering, Pearson, Saddle River, NJ.

[4] Rodgers, P., Green, G., McGown, A., 2000, Using Concept Sketches to Track Design Progress, Design Studies, vol no .21 (5), pp 451-464.

[5] McGown, A., Green, G., Rodgers, P., 1998, "Visible Ideas: Information Patterns of Conceptual Sketch Activity," Design Studies, vol no. 19 (4), pp 431-453.

[6] Dym, C., Agogino, A, Eris O., Frey D., and L. Leifer. (2005) "Engineering Design Thinking, Teaching, and Learning," Journal of Engineering Education, vol no. 94 (1).