Higgins, E. L., & Raskind, M. H. (1995). Compensatory effectiveness of speech recognition on the written composition performance of postsecondary students with learning disabilities . Learning Disability Quarterly , 18 (2), 159–174. https://doi.org/10.2307/1511202

Participants wrote three essays, one for each of the following conditions: 1) using a speech recognition system, 2) dictating the essay to a human transcriber, and 3) without assistance. Students were allowed to handwrite or word process the 'no assistance' essay but were not allowed to use the spell-checking function.

Participants were 29 postsecondary students with learning disabilities enrolled at California State University, Northridge (CSUN). Twenty-three students were Caucasian, 3 were African-American, and 3 were Hispanic. The mean age was 24.9 years and the mean IQ was 97.

Students wrote essays from one of six possible questions. Essays were holistically scored on a scale of 1 to 6.

Speech recognition assists students with learning disabilities in compensating for their difficulties in written composition. When compared to receiving no assistance, students achieved higher holistic scores using the technology. Speech recognition apparently allowed students to use their more extensively developed oral vocabularies at a level that was statistically significant.
The single most sensitive predictor of the holistic score was words of seven or more letters.
The program was much better at making correct guesses for longer words than for short, unisyllabic ones.
The ratio of unique words to words was negatively correlated with composition length, a powerful predictor of holistic score.