AT for Writing for Students with High Incidence Disabilities

65% of students referred for learning disabilities have a writing disability (Mayes, Calhoun, and Crowell, 2000). The National Assessment of Educational Progress (2009) found only 6% of students with disabilities scored at a proficient level on writing tests. 46% scored below basic level, and 48% performed at basic level (Smith & Okolo, 2010).

Students with LD are more likely to have errors in spelling, punctuation, capitalization, and word usage. Their writing is more likely to be shorter and illegible (DeLaPaz, 1999).

Personal computer spell checkers, digitized text, word prediction software, speech or voice recognition, and alternative writing tools are the most common computer features used in schools to facilitate writing (Cullen, Richards, & Lawless-Frank, 2008; Barbetta & Spears-Bunton, 2007). Compared to handwriting, even word processing makes a difference. Hetzroni & Shcreiber (2004) found that students had fewer spelling errors, fewer reading errors, and improved organization and structure when using the computer. MacArther, graham, Schwartz, and Schafer (1995) found that improved performance depends on how well students are trained to use the features.

Is word prediction effective?

The term word prediction may include both word completion (guessing the remainder of a word based on the first letter or two) and true word prediction (guessing the next word based on the current word) aspects.  Studies of word prediction software prior to 2003 did not have phonetic spelling, so word prediction was less accurate. However, even older studies showed benefit. Word prediction alone and in combination with text-to-speech have had a positive impact on the written output of students with identified learning disabilities (Silió & Barbetta, 2010; Cullen, Richards, & Lawless-Frank, 2008; Tam, Archer, Mays, & Skidmore, 2005). Most studies look at number of words written, spelling accuracy, and writing rubric scores (including total unit length).

Do graphic organizers support writing?

Studies of the use of graphic organizers as AT show increases in number of words written, amount of time spent on planning, and common story elements (Blair, Ormsbee, & Brandes, 2002; Sturm & Rankin-Erickson, 2002; Unzueta & Barbetta, 2012; Gonzalez-Ledo, Barbetta, Unzueta, 2015). Changes in overall organization of the written product were found in some, but not all of these studies. In addition, electronic graphic organizers allow educators to change the visual representation of the images and text, convert the information in a concept map to an outline, and add audio and text, and allow students to manipulate text, alternate between concept map and outlining, and insert information (Englert, Wu, & Zhao, 2005; Sturm & Ranking-Erickson, 2002). Englert, Wu, and Zhao (2005) found that the 12 students in their study performed significantly better when using the graphic organizer than without it. Sturm and Rankin-Erickson (2002) found students demonstrated a more positive attitude to the computer based graphic organizer than to the hand drawn graphic organizer or the no organizer conditions.

Does speech recognition make a difference?

In several studies the use of speech recognition (referred to as voice recognition in older research) produced passages with more words and fewer errors than handwritten passages (Quinlan, 2004; MacArthur & Cavalier 2004; McCullum, Nation, & Gunn 2014).  However, the use of speech recognition (SR) requires the ability to plan phrases and sentences and to dictate without stopping to correct every error (Cullen, Richards, & Lawless-Frank, 2008).

SR has a large impact for some, but not all students. McCullum, Nation, & Gunn (2014) looked at total words, total multisyllabic words, and correct writing sequences. They determined writing sequences by looked at every pair of words. For some students, changes were dramatic, going from writing 18 words in the pre-test to 118 in the post-test. For others, the performance changes were much smaller.

When to try SR?

SR works for many, but not all students with Learning Disabilities. SR is particularly beneficial for those whose oral skills are stronger than their writing skills (Li & Hamel, 2003).

A good guide is Speech Recognition as AT for Writing (Cochrane & Key, 2014).. It can be downloaded from 

What is the impact of technology use on writing?

Batorowicz, Missiuna, & Pollock, (2012) did a very thorough review of the research. They looked at word processing software, spell checkers, word prediction, speech recognition, concept mapping/organizing software, and multimedia. Although the studies are limited, they suggest a positive influence on quality of written text, organization, transcription, and revising. AT combined with instruction yields the most positive results. Collaborating with peers when using technology appears beneficial for both composing and revising. They determined that technology motivates children and enhances their opportunities to practice writing.