Using Speech to Text to Support Written Expression in Students with ADHD

Written expression is complex task dependent upon the simultaneous integration of multiple language skills, working memory, graphomotor skills, and processing speed. Students with Attention Deficit Hyperactivity Disorder( ADHD) often exhibit pervasive difficulty meeting writing demands in the classroom environment.

Very often, these students are able to express their thoughts and ideas verbally but do not have the required transcription skills to meet the demands of the writing task. Evidence supports that youth with ADHD generate less organized written work, write fewer words, and make more mechanical errors (e.g., misspelled words and poor handwriting) in comparison to their peers (Molitor, 2016).

Unfortunately the profile of most students with ADHD often includes:

  • Writing avoidance and frustration
  • Simple word choice that does not reflect vocabulary abilities
  • Lack of punctuation, capitalization, and paragraph indentation
  • Written output that does not match oral language skills

Assistive Technology for Written Expression

Speech to Text technology (STT) converts spoken word into text. STT is otherwise known as dictation or speech recognition software. Thanks to Siri, Alexa and Hey Google ,speech recognition software has become a widely accessible and universal productivity tool in the lives of most people.

For many neurodiverse students, the addition of STT can remove substantial barriers to learning and expressing their knowledge. Speech recognition works well for students who have difficulty performing the output demands of writing (transcription) but who can express thoughts and ideas well verbally (expressive language). 

Speech-to-text tools are becoming more common on IEP and 504 accommodation plans these days. Do you know enough about how these technologies work and how to realistically introduce them to your students?

This week OT4ADHD, will be taking a deep dive into Speech to Text (STT) technology. This post will describe free and easily accessible speech to text options and provide evidenced based resources and an actionable intervention plan.

Benefits of Speech to Text Software as Assistive Technology

Written expression requires the simultaneous integration of many different task-demands. Written expression is a complex process of orthographic integration that is governed and managed by working memory, an area of significant weakness in children with ADHD. 

Assistive Technology benefits of Speech to text for Students with ADHD

Speech To Text ( STT) reduces many of the task demands impacted in learners by ADHD including :

  • Supports working memory Speech To Text immediately supports working memory by decreasing cognitive load. STT isolates the thought generation task from the transcription task. Students are not simultaneously overloaded by grammar, sentence structure, spelling demands – these can be dealt with after students get their thoughts on paper.
  • Reduce motor demands  STT supports students with delayed motor skills, enables them to write more comfortably and accurately. STT removes physical endurance and fatigue from the writing process.
  • Reduce demands of visual motor integration Students with poor handwriting ( transcriptions skills) can use dictation to create text that’s easier for others to read.
  • Reduce demand on orthographic motor integration
  • Reduce demand on encoding (spelling)
  • Supports slow processing speed: Students who are strong in verbal expression but have difficulties organizing their thoughts on paper can use STT to demonstrate their knowledge.
  • STT can be used for note taking.
  • STT can empower students to compose longer responses.
  • STT creates a positive learning and writing environment , reducing writing avoidance and frustration.

Built-in Accessibility Tools- Speech to Text

There are several types of speech to text applications available to students today.  Chromebooks, Macbooks and Ipads have built in speech to text assessibility features. The microphone and dictation tools come with the device, but you may need an internet connection.

Picture of Device with Speech to text Available as Built In Programs

Chromebooks: 

Many school use Google and Chromebooks widely.  Chromebook has a free built in text to speech accessibility feature called Google Voice Typing. Available on Chromebooks and the Chrome Browser on all other devices.

Macbooks:

Mac OS built in Macbook also has built in Speech to Text features. I am using right now to write this article. Pressing the “fn” button twice “Function Function” is the shortcut on my MacBook

IPads 

IPads have Siri built in right next to the keyboard. Most children in 2022 have had a chat or two with Siri.

You Can’t Just Hand Them The Device 

Often I see, access to laptop, or speech to text software as an accommodation on IEP’s and 504 plans without the student being fully trained on how to use it. Like any tool, if you do not know how to use it fluently, it will not help the task. A student will require 8-10 sessions of training before being able to determine if speech to text software is appropriate to complete classroom assignments.

Daniel Cochrane & Kelly Key, co authored Speech Recognition as AT for Writing,  A Guide for K-12 Education. The authors propose a scaffolded approach to teaching speech recognition software to our students.

Cochrane and Key, emphasize that the key to teaching speech recognition with students is to explicitly teach the speech recognition writing process. Which you will learn, is different then both conversational speech or the typical pencil and paper writing process.

The Speech Recognition Writing Process for STT

The Speech Recognition Process Adapted from Daniel Cochrane & Kelly Key, Speech Recognition as AT for Writing,  A Guide for K-12 Education
Adapted from Daniel Cochrane & Kelly Key, Speech Recognition as AT for Writing,  A Guide for K-12 Education

Speech recognition software uses the context of the sentence for word recognition and spelling. There are four key elements to the sequential process .

  1. Think it : Encourage students to mentally formulate the sentence ( not just the thought) before speaking it.
  2. Say it : Talk naturally and clearly. Try not to talk like a robot.
  3. Check it: Teach students how to catch recognition errors after each sentence.
  4. Fix it : Decide whether to fix a few words or re-dictate the whole sentence.

Scaffolded Steps to Teach the Speech to Text Writing Process

In Speech Recognition as AT for Writing,  A Guide for K-12 Education, the authors provide the below scaffolded approach to teaching speech recognition software to our students. I personally apply this model in my own practice and have had great success with it.

  • Introduction : Install and introduce the software, Model the STT writing process.
  • Level 1: Student writes a personalized sentence you provide
  • Level 2: Student writes 3-4 sentences on a personal topic.
  • Level 3 : Student writes and edits 1-2 paragraphs from motivating pictures or other personal topics.
  • Level 4: Student writes and edits 5-10 sentences using single words from an academic vocabulary list.
  • Level 5: Student writes multiple paragraphs after completing a pre-writing organizer using only key words or phrases.
  • Level 6: Student complete an academic writing assignment using speech to text
  • Level 7 : Student completes an independent writing assignment using speech to text (homework or at school).

School Based Occupational Therapy Speech to Text Intervention Plan

Below I outline the intervention plan I use for a small group learning how to use STT on school provided Chromebooks. The sessions are based upon the framework from Daniel Cochrane & Kelly Key, Speech Recognition as AT for Writing,  A Guide for K-12 Education. All sessions are provided in collaboration with the grade level resource room, classroom special education teacher or the students paraprofessional for implementation guidance in the classroom .

I find that my students, partially students with ADHD , really connect with the idea of “leveling up”. We do not compete against one another and use the visual to provide a beginning and end goal .

Introduction: Model the speech recognition writing process and software operational process.

This is harder than it looks, even for tech savvy therapists. Before you try and teach it, get your hands on the students device and ensure it is operational.

Therapist models the four step process and operational procedures.

  • Think It : Use a “think aloud” strategy as you choose a topic to write about and mentally compose your topic sentence.
  • Say it : Dictate the sentence and model the operational skills.
    • Turn on the microphone.
    • Dictate clearly and dictate punctuation.
    • Turn off the microphone.
  • Check It: Read the sentence out loud to model how to carefully check recognition accuracy after each sentence.
  • Fix it : Demonstrate how to correct recognition errors if there are any. Point out how there are no spelling errors.
The Speech Recognition Process Adapted from Daniel Cochrane & Kelly Key, Speech Recognition as AT for Writing,  A Guide for K-12 Education
Adapted from Daniel Cochrane & Kelly Key, Speech Recognition as AT for Writing,  A Guide for K-12 Education

Level 1 : Student uses Speech to Text to Dictate One Personalized Sentence.

  1. Provide think, say, check, fix visual ( pictured above )
  2. Select either a personalized fill-in-the-blank sentence. or a “Would you Rather ” question depending on the age and interests of your students.
    • “I have ___ (number or no) siblings and ____ (number or no) pets.
    • Would you rather live in the air or under the sea?
  3. Interview the student to fill in the blanks. DO NOT WRITE IT DOWN
  4. Say the sentence to the student with their choices.
  5. Ask the student to orally repeat the sentence. 
  6. Demonstrate how to turn on the microphone.
  7. Ask the student to dictate the sentence (or first phrase) into a microphone. 
  8. After dictation, read the transcribed sentence with the student to look for recognition errors.  
  9. Gently coach enunciation and compare output until the sentence is correct. ( minimum 80% enunciation or you may need to consider different software, positioning,or microphone).
  10. Practice with the next sentence , fading thinking out loud and fading support.
  11. Student levels up when they can silently think the sentence, say it, check it and fix it with 80% accuracy.

Level 2: Student Dictates 3-4 sentences on a personal topic. 

  • Interview students about the personal topic to get ideas flowing , encouraging descriptive terms. Avoid topics heavy with names and character names .
  • Topics I use include:
    • All About Me
    • If you could have any superpower, what would you choose?
    • Describe your ideal teacher?
  • Think It: At first, ask the student to practice their sentence out loud before turning on the mic.
  • Say It: Have the student say the sentence into the microphone.
  • Check It: Teach the editing process after each sentence.
    • Coach the student to visually check the transcribed sentence for recognition accuracy 
  • Fix or not to Fix: If a sentence has too many errors, it may be better to start over. 
  • Practice and Teach Punctuation at age appropriate level .
  • Repeat for 3-4 sentences about the topic 

Level 3 : Student Writes and Edits 1-2 Paragraphs from Motivating Picture 

Free Printable Picture Prompts for writing practice with Speech to text Technology
  • Model the use of a pre-writing outline as needed using only keywords to prompt dictation, not full sentences.
  • Continue to coach the SST process
  • Gradually fade coaching as the student follows the four tasks of the speech recognition writing process independently.
  • Collect data on independence and writing quality.

Level 4: Student Writes and Edits 5-10 sentences Using Single Words from an Academic Vocabulary List.

The purpose of this step is to push the student to create the more complex sentences that will be needed for academic writing. Many students have this as a homework assignment weekly.

  • Ask teacher for a list of vocabulary words.
  • Pick a word from the list that they know, define it and ask the student to create a sentence with it.
  • Repeat for 5-10 sentences.
  • Coach the student to create lengthier and more complex sentences if needed and observe the student’s performance. 
  • Sentence stretching can be performed during the think it phase or during the fix it phase.

Level 5 : Student Writes Multiple Paragraphs after Completing a Pre-writing Organizer using only Key words/phrases.

The purpose of this step is to practice using speech recognition to compose an organized essay. Most often students who use STT will complete the graphic organizer with the whole class then perform the dictation in a separate location like resource room.

  • Help the student identify an interesting topic.
  • Have the student to complete a pre-writing organizer.
    • Coach the use of keywords/phrases the graphic organizer rather than writing out full sentences so that speech recognition is still used to create the rough draft.
  • Coach the student through the process of turning key words from the map into full sentences using speech recognition.

Level 6: Test Drive: Student Completes Semi-Independent Assignment Using Speech to Text

  • Prepare the student for the Test Drive.
    • Give an assignment the student can do independently.
    • Make sure the technology is available for use.
    • Make sure someone is available to support the student with their assignment as needed.
  • Student reports back on how it went, and therapist and student problem solve as needed.

Level 7. Student Independently Completes an Academic Writing Assignment Using Speech to Text

  • Use an assignment the student needs to do for class.
  • Arrange for the student to complete the pre-writing organizer in class using keywords/phrases if possible.
  • Do not coach the writing process.
  • As you observe, collect data on operational skills.

Level 8. Implement: Student Independently Completes an Academic Writing Assignment Using Speech to Text in the Classroom

  • When the student is ready and feels confident, therapist can collaborate with the teacher , push in an observe / problem solve implementation barriers in the classroom context.

Exit Tickets: 

When considering , training or implementing assistive technology for a student, their preferences are highly important. Each session the therapist should collect data on the students perspective of the tool. At first, novelty may be motivating, but learning something new will be challenging and both aspects may impact their perception of the tool. Exit tickets are a quick and simple method of collecting data and communicating with parents and teachers. .

Speech to text OT Session Exit Slip

Speech to Text for Literacy Development

Many of our students with ADHD, especially those with comorbid Dyslexia, Dysgraphia or Developmental Coordination Disorder, will demonstrate signifiant developmental delay in the automaticity of spelling and writing by hand. These students will not have the skills required by the third grade to effectively use writing FOR learning .

When the barrier to written expression is identified in the transcription process (automaticity of spelling or handwriting) , practice alone will not sustainably improve the students classroom performance during that academic school year.

Today most classrooms have 1:1 devices for learners. Speech to text technology has incredible potential in reducing the most pervasive transcription barriers to learning and expressing knowledge.

Cochrane and Key, emphasize that the key to teaching speech recognition with students is to explicitly teach the speech recognition writing process using a scaffolded technique. A student will require 8-10 sessions of training before being able to determine if speech to text software is appropriate to complete classroom assignments .If speech recognition is effective as assistive technology for written expression therapists can implement and monitor effectiveness and regular use over time.

Providing speech to text as an option for demonstrating knowledge will reduce the extraneous cognitive load , reduce frustration, reduce avoidance and empower the development of higher level literacy skills.

I can not think of a more powerful use of 8 -10 sessions then introducing and trialing speech to text. Can you?

Resources and Practice Tools

Learn More about ADHD and Written Expression Here

References

  • Cochrane, Daiel, Key, Kelley. Speech recognition as AT for Writing, A Guide for K-12 Education. 2020
  • Molitor SJ, Langberg JM, Evans SW. The written expression abilities of adolescents with Attention-Deficit/Hyperactivity Disorder. Res Dev Disabil. 2016 Apr-May;51-52:49-59. doi: 10.1016/j.ridd.2016.01.005. Epub 2016 Jan 21. PMID: 26802631; PMCID: PMC5134244 PubMed
  • Hetzroni OE, Shrieber B. Word processing as an assistive technology tool for enhancing academic outcomes of students with writing disabilities in the general classroom. Journal of Learning Disabilities. 2004;37(2):143–154. PubMed
%d