The effects of word prediction on communication rate for AAC

Abstract

Word prediction systems can reduce the number of keystrokes required to form a message in a letter-based AAC system. It has been questioned however, whether such savings translate into an enhanced communication rate due to the additional cognitive load (e.g., shifting of focus and scanning of a prediction list) required in using such a system. Our hypothesis is that word prediction has great potential in enhancing communication rate, but the amount is dependent on the accuracy of the word prediction system. Due to significant user interface variations in AAC systems and significant differences between communication rates achieved by different users on even the same device, this hypothesis is difficult to verify. We present a study of communication rate and word prediction systems using "pseudo-impaired" participants and two different word prediction systems compared against letter-by-letter entry. We find that word prediction systems can in fact speed communication rate, and that a more accurate word prediction system can raise communication rate higher than is explained by the additional accuracy of the system alone.

User Interaction with Word Prediction

“…This decrease is consistent with our previous findings as well as those of other researchers [Trnka et al 2007;Venkatagiri 1993;Koester and Levine 1994b] and can be explained by an increase in the cognitive overhead needed to deal with the prediction list. The decrease in input rate between basic and advanced word prediction is explained by an increasing reliance on the prediction system.…”

“…The major finding of this study and our previous work [Trnka et al 2008;2007] is that a good word prediction system can improve communication rate, much more so than a basic system. The increase in communication rate is not only due to increased potential keystroke savings, but also due to higher utilization of the system when users trust the predictions.…”

“…We have also focused on both a high-quality prediction system as well as a basic system, whereas other research seems to produce results similar to our basic system. This article is an expansion of our previous studies Trnka et al [2008;2007] using 33 participants, including much more detail regarding our experimental methods as well as expanded analysis. We compare three different text entry methods (also referred to as input methods) for AAC text: typing on a keyboard with no word prediction, typing on a keyboard with a basic word prediction algorithm available, and typing on a keyboard with a more advanced word prediction algorithm available.…”

Corpus studies in word prediction

“…Although an end user is more concerned with communication rate (words per minute), communication rate has been shown to increase with increased keystroke savings [20].…”

“…[11] varied ngram order and training text size rather than domain, validating intuitions regarding the relation between training text size and keystroke savings. [20] seeks to answer the same general question as this work (how will word prediction help real users?) but studies the connection between keystroke savings and communication rate.…”