Kevin Pack, Communication Intern for The Wheel
Dictation Software in Education
Dear Faculty,
This article was written by Kevin Pack, a communication intern for The Wheel. Kevin Pack is a third year computer science student at UC Davis. He is interested in software development, music, and education. As a new communication intern for The Wheel, Kevin is looking forward to exploring the applications and effects of instructional technology at UC Davis.
"Speech-to-text" or "dictation" software has been available to consumers since 1990, when Dragon Systems released "Dragon Dictate." This speech recognition software made it possible for users to employ their computers as a personal scribe that recorded their every word. However, Dragon's product originally cost an astonishing $9,000 per copy (almost $18,000 in today's money), and it required the speaker to pause between each word. Thirty years have passed since the birth of "Dragon Dictate," and dictation software has come a long way in terms of usability and potential applications. UC Davis faculty might consider the following five ways to use speech-to-text tools to support student learning.
However, speech-to-text is still imperfect. Dictation software does not have a streamlined method of editing sentences. If one were to make an error in one’s speech, one could attempt to select and remove it using voice commands (or say "backspace" ten times fast!), but it is inevitably easiest to edit the words manually, especially those larger mistakes.
Additionally, voice typing is much less convenient in public spaces than standard typing. First, it is almost impossible to filter out background voices and sounds. These will likely get picked up by a computer's microphone, thus adding extraneous words or even sentences to one's work. Second, using speech-to-text means that a user's neighbors can listen in on what is being transcribed. In addition to a privacy concern for the user, the user's speaking could be a nuisance for others in a quiet space. This makes voice-to-text difficult to use in public spaces like libraries or coffee shops.
Lastly, all punctuation must be inputted with one's speech when using dictation software. Google attempted to add auto-punctuation to voice typing on GBoard in an update at the end of January 2020. However, as of February 10, 2020, The Verge reports that GBoard's automated punctuation is "overly aggressive" and generally faulty. Google seems to be working on a solution for punctuation with voice typing, but it is uncertain when it will be complete.
Although speech-to-text software is impressively accurate and quick in 2020, it may still need time before it reaches its full potential. Nevertheless, such software can save time in specific scenarios, and can help faculty prompt student reflection on their personal productivity, as well as their writing and presentation skills. Period.
This article was written by Kevin Pack, a communication intern for The Wheel. Kevin Pack is a third year computer science student at UC Davis. He is interested in software development, music, and education. As a new communication intern for The Wheel, Kevin is looking forward to exploring the applications and effects of instructional technology at UC Davis.
"Speech-to-text" or "dictation" software has been available to consumers since 1990, when Dragon Systems released "Dragon Dictate." This speech recognition software made it possible for users to employ their computers as a personal scribe that recorded their every word. However, Dragon's product originally cost an astonishing $9,000 per copy (almost $18,000 in today's money), and it required the speaker to pause between each word. Thirty years have passed since the birth of "Dragon Dictate," and dictation software has come a long way in terms of usability and potential applications. UC Davis faculty might consider the following five ways to use speech-to-text tools to support student learning.
- Help students reflect on their writing. Dictation software prompts users to speak precisely. As I wrote parts of this article with voice typing, I felt compelled to dictate to my laptop with exact grammar and diction, for I knew that everything I said was instantly recorded. This forced me to slow down and think about what I wanted to say before I began speaking.
- Support students’ development of presentation skills. As users read the transcript of their words while using speech-to-text software, they are prompted to identify counterproductive speaking patterns that they wish not to use in presentations (e.g. overuse of ums or likes).
- Help students who may struggle with typing speed. Although the average person speaks about 150 words per minute, according to the National Center for Voice and Speech, the average typing speed is only 52 words per minute, according to a study by Cambridge University. By making use of current speech-to-text software, one could triple their typing speed.
- Create records of lectures, club meetings, or interviews. Students and instructors can easily make searchable transcripts of lectures, club meetings, or interviews. Such transcripts allow users to focus on the tasks at hand, rather than on their notes.
- Improve Accessibility. Dictation software improves accessibility for all students by creating captions of lectures. However, the UC Davis Accessibility site recommends that users review the program's captions for accuracy whenever using automated captioning.
However, speech-to-text is still imperfect. Dictation software does not have a streamlined method of editing sentences. If one were to make an error in one’s speech, one could attempt to select and remove it using voice commands (or say "backspace" ten times fast!), but it is inevitably easiest to edit the words manually, especially those larger mistakes.
Additionally, voice typing is much less convenient in public spaces than standard typing. First, it is almost impossible to filter out background voices and sounds. These will likely get picked up by a computer's microphone, thus adding extraneous words or even sentences to one's work. Second, using speech-to-text means that a user's neighbors can listen in on what is being transcribed. In addition to a privacy concern for the user, the user's speaking could be a nuisance for others in a quiet space. This makes voice-to-text difficult to use in public spaces like libraries or coffee shops.
Lastly, all punctuation must be inputted with one's speech when using dictation software. Google attempted to add auto-punctuation to voice typing on GBoard in an update at the end of January 2020. However, as of February 10, 2020, The Verge reports that GBoard's automated punctuation is "overly aggressive" and generally faulty. Google seems to be working on a solution for punctuation with voice typing, but it is uncertain when it will be complete.
Although speech-to-text software is impressively accurate and quick in 2020, it may still need time before it reaches its full potential. Nevertheless, such software can save time in specific scenarios, and can help faculty prompt student reflection on their personal productivity, as well as their writing and presentation skills. Period.