MacArthur, C. A., & Cavalier, A. (1999). Evaluation of the feasibility of dictation with speech recognition technology to support the assessment of all students . University of Delaware Education Research and Development Center.

Attachment study from the Inclusive Comprehensive Assessment Systems Project

Broken Link: http://www.doe.state.de.us/aab/DSTP_research.html

Research assistants trained subjects on the use of speech recognition technology, namely the Dragon Naturally Speaking for Teens. Students dictated passages and word lists to the system, using speech recognition technology.

Students in middle school (mostly grade 8) and high school (all in grade 10) in Delaware (U.S.) participated. Students with learning disabilities (LD) and students without disabilities comprised the sample. The students with LD had been identified in part with the WISC III full scale IQ by their schools as having LD with difficulties in writing.

At the conclusion of each training session, participants completed a probe of approximately 100 words, which they dictated to the computer. The passages were taken from a collection of brief expository passages at the grade 4 reading level.

The researchers were interested in whether high school and middle school students could use the speech recognition software with adequate accuracy. After three initial hours of training, the median accuracy was 89%. Seventy-six percent of students met the criterion of 81% accuracy, and 40% met the criterion of 90% accuracy. The researchers were also interested in whether accuracy of the speech recognition system differed by special education status. The difference between normally achieving and special education students did not quite meet the conventional standards for statistical significance.