We present the results of an exploratory study investigating the feasibility of using multimedia software to teach life skills to adults with learning difficulties. As a precursor to determining whether the clients would benefit from the software, we needed to address the issue of realism in visual displays, to discover if photorealistic images of a familiar kitchen and utensils were essential, or if the clients would be able to abstract and apply information from generic cartoon-like representations. The level of realism was varied in two sets of tasks: object recognition exercises and problem-solving scenarios. Realistic versions of each task contained photorealistic images, and the problem-solving scenarios used images and speech of a support worker known to the participants to supply feedback and prompts. Unrealistic versions used clip art images and a cartoon-style character instead of the support worker. Contrary to expectations, measurements of errors and reaction times revealed the level of realism to have a negligible effect upon user performance in both sets of tasks. What has emerged is the overwhelming effect of individual differences on the design and evaluation of learning software. © Springer-Verlag 2004.
CITATION STYLE
Hetherington, R., Crerar, A., & Turner, P. (2004). A question of realism. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 3196, 68–76. https://doi.org/10.1007/978-3-540-30111-0_6
Mendeley helps you to discover research relevant for your work.