Master of Arts (MA)
Domangue, Mathews, Sun, Roussel, and Guidry (2004) trained participants to generate valid exemplars from an artificial grammar using either memory-based or model-based processing. Their results showed that learning by memory-based processing resulted in fast but inaccurate performance, while model-based learning resulted in slow but accurate performance. Attempts to integrate both types of training did not result in fast and accurate string generation. Fast and accurate performance was achieved by Sun and Mathews (2004) using a computer animated display to train participants. The current study used a 2x2x2 factorial design to determine why participants who view an animated display of a diagram of the grammar perform well at test. The results suggest that the diagram informs participants of which letters, or chunks of letters can appear in each position, as well as where they cannot appear. Animating the diagram focuses attention on the relevant portion of the complex display and leads to the best performance.
Document Availability at the Time of Submission
Release the entire work immediately for access worldwide.
Sallas, Bill, "Animation in artificial grammar learning: can animation facilitate learning?" (2006). LSU Master's Theses. 872.
Robert C. Mathews