Overview

• Word prediction software, sometimes called “predictive text”, provides users with a list of suggested words based on what they’ve previously written in a sentence

• Word prediction is frequently used in word processing programs (e.g., Word, Pages, Google Docs), but is also seen in smartphone texts and emails. It requires one letter to be typed for prediction to activate

• Can help with word selection and sentence composition skills when writing

To see Word Precition in action, check out this video: https://www.youtube.com/watch?v=zwz3YNJ_Mwg

Research

• Evmenova & Regan, 2019

  • Reviewed, among other technologies, the efficacy of word prediction software in students with learning disabilities (LDs)

  • Provide strong evidence of numerous research studies showing that word prediction software improves word generation, spelling, and sentence structure in students with LDs

• Miranda et al., 2006

  • Examined the impact of word prediction software on the written output of 8-19 year olds with a physical disability impacting their graphomotor ability through measures of words produced in 10 minutes

  • Findings demonstrated that word prediction software helped students writer faster and neater, and led to improvement in word variety usage and sentence structure

• Brown & Almond, 2021

  • Examined the usage and impact of word prediction software with bilingual children in first grade

  • Results suggest that, following word prediction software usage over 1 year, improvements were present in spelling, vocabulary, and written expression length

Quick Facts

Advantages

• Improves vocabulary, spelling, and sentence structure

• Can be used independently, in multiple applications and environments

Disadvantages

• Requires thorough training and familiarity with software

• May occasionally predict non-useful words

To Consider

• Different softwares present word prediction differently - training in one program may not translate to another

Insights from Practice

Word prediction is an extremely useful tool for students with poor spelling skills, but they are most useful when the latency (the time between when they student types and the options appear) is quick, if the student is waiting for words to be generated, they will often discontinue use. It can be helpful to limit the number of words offered in the output, typically 5 is good, however, if they are poor spellers they may need a longer output list to ensure the desired word appears. If they have decoding issues, ensure the product also has Text to Speech built into it so that it can present the options orally so that students can properly consider the offered words.