An Improved Display for Word Prediction Software


  • David Wells Bloorview Research Institute
  • Cynthia Tam Bloorview Research Institute
  • Paul O'Brien Bloorview Research Institute
  • Colin Ward Bloorview Research Institute
  • Tom Nantais Bloorview Research Institute


Word prediction software is designed to enhance rate and accuracy of text entry. A key factor affecting the use of word prediction software seems to be the associated visual cognitive load. Children with fine motor and visual tracking difficulties often use vision to guide their fingers, but they commonly have reduced oculo-motor control. Moving their eyes between the keyboard and the monitor costs time, reducing the benefit of word prediction. To address this, it is felt that the prediction window should be placed as close to the keyboard as possible. In a pilot project, we took an important step in reducing the load and improving performance by developing a prototype with the prediction window made available at the keyboard. The prototype display consists of a Personal Digital Assistant (PDA) with an appropriately-sized LCD module. We developed a Dynamic Link Library (DLL) to provide the software interface to the word prediction software. A pilot study evaluated the effect of the display location on performance. Participants had a physical disability affecting fine motor functions (mostly Cerebral Palsy). Clients using the prototype improved their performance (rate and accuracy) and expressed a clear preference for it. As a group, the pilot study participants showed a significant improvement in Canadian Occupational Performance Measure (COPM) of more than 2 for both performance and satisfaction. The pilot work has laid the foundation for further development of the display and validation of the results with additional client populations 




How to Cite

D. Wells, C. Tam, P. O’Brien, C. Ward, and T. Nantais, “An Improved Display for Word Prediction Software”, CMBES Proc., vol. 30, no. 1, Dec. 2007.