Empirical Study of Information Design: Four Experiments

Current design theory sets out many rules and guidelines for designers, but good design is still difficult to replicate. Often the design principles found in the manuals are misapplied, resulting in designs that (1) do not fulfill their purpose and (2) disrupt the clarity of information. This thesis...

Full description

Bibliographic Details
Main Author: Alton, Noel Teresa
Format: Others
Published: BYU ScholarsArchive 2010
Subjects:
Online Access:https://scholarsarchive.byu.edu/etd/2118
https://scholarsarchive.byu.edu/cgi/viewcontent.cgi?article=3117&context=etd
Description
Summary:Current design theory sets out many rules and guidelines for designers, but good design is still difficult to replicate. Often the design principles found in the manuals are misapplied, resulting in designs that (1) do not fulfill their purpose and (2) disrupt the clarity of information. This thesis will review and provide experimental data supporting a model of visual form/visual purpose connections based on the semiotic of C.S. Peirce. This model was first used by Amare and Manning (2005, 2006, 2007, 2008, 2009) to evaluate and explain both effective and ineffective visual information design. This thesis will extend their approach, reporting on the results of four experiments to test the aesthetic appeal and information retention from various visual designs. The four experiments presented in this thesis show that viewer's ability to recall information does not coincide with designs that they find the most visually stimulating or visually pleasing. High indicative contrasts allow for higher retention rate, but those contrasts do not necessarily conform to viewer's aesthetic preferences. Low indicative contrast options have a lower retention rate, but are preferred aesthetically by viewers. Peircean analysis accounts for this disconnect between usability and preference and can help designers find the balance that is needed between these competing purposes in visual information design.