Voice recognition software


The ‘Just Ask’ campaign from the Scottish Government urges parents to ask for help for their children ‘to help them navigate their way through issues affecting them’: these issues include dyslexia.

The advert in the local paper declares: ‘Support can be provided in many different ways and is entirely individualised to the child. For example, a pupil with dyslexia may benefit from use of voice recognition software’.

Oh dear. While I am in compete agreement that young people with dyslexia need a ‘plan to ensure success’, to be so specific about such a very difficult area as speech to text software is going to cause all sorts of genuine problems for us in schools. I can hear the parents bashing my door down as I write.

I have not used voice recognition software myself nor have I seen it in action. There are mostly good reviews here  but none focus on its deployment in schools.

The only review I can find that specifically mentions voice recognition software and young people was last updated in December 2006. There may, of course, have been considerable changes for the better since then. I’d be grateful if anyone with knowledge and experience in this area would join the discussion.

 I have tried to summarise the major points in that review as far as they relate to enabling youngsters to circumvent the barrier of writing. The article does state that: ‘where spelling, handwriting and composing are major problems, then Speech Recognition can be hugely liberating and allow children to express their ideas on paper fluently for the first time in their lives’. Studies have shown that students with learning difficulties who use speech recognition:


  •  Use longer and richer words
  • Write more creatively
  • Organise work better
  • Complete more work
  • Improve reading
  • Improve spelling
  • and produce better hand-written work’.


‘A ‘normally clear speaker [and here they are referring to adult users], using a recent computer with a decent microphone and with a little experience should get very good recognition results and gain real productivity benefits’. (My emphasis).

However, there are also many caveats, summarized thus:

  • Motivation is a key component. It would be impossible for children to learn dictation at the same time as learning the basics of computing, of Windows and of word processing. This all adds up to considerable information overload and time implications. On the whole children don’t produce masses of written work, so are less likely to have the motivation to persevere with speech recognition. If it’s not used often the child will need to relearn it each time; and is likely to stop bothering.
  • Patience and accuracy are needed. Time training the software to recognise the voice is necessary. [And in young people, especially boys, the voice can change, requiring frequent up-dates]. Accuracy is hugely important. Each ‘mistake’ [or misinterpretation by the computer] made takes many times longer to correct compared with dictating a word correctly. [Children with ‘regional accents’ – i.e. accents from areas outside the south of England – are likely to find training the software to recognise their voices even more difficult]. Fluent speakers with a wide vocabulary are most successful at using speech recognition software. [Not all learners with dyslexia fall into this category].
  • Good, fast word retrieval, finding the words needed easily to express ideas, is essential. Multi-tasking – using the software whilst composing text – is also likely to be hard for many children with dyslexia. [Slow processing speed and poor automaticity are very common characteristics].
  • An understanding of word processing & punctuation is crucial. [No comment necessary].
  • A quiet, relatively private environment for confident dictation for training and using the software [is not always available in busy classrooms]. 
  • Proof reading is particularly difficult for dyslexic people. They are liable to have more difficulty finding and correcting an error than somebody who reads and spells well. Even the best dictation system, after you have spent a long time training it and working with it, will make recognition mistakes.
  • Ongoing support is essential and this has implications for training for those who work with the child. For somebody new to dictation there are a lot of things to get right: diction style, microphone adjustment and positioning, making corrections, punctuation and the voice commands. Modifications to speech style (pace, clarity, particularly of unstressed words, evenness of volume) make a big difference.

The review stresses that, while a helpful resource in many instances, Speech Recognition software ‘can still lead to frustration and a lack of success. The main reasons for this will be human, not technical’. However, there are inevitably technical issues that would need to be addressed if adoption of the software is to be effective:

  • [The microphone that comes with the software is universally derided]. A better microphone than that supplied in the box may make the difference between success and failure. In addition, it is absolutely critical to have the microphone properly adjusted, and the authors of the review suspect that this is the single most likely cause of frustration and failure at dictation.
  • It is generally the case that a laptop computer will be slower and noisier than a desktop machine of the same specification although both might be so slow as to be virtually unusable. It follows that it is all the more important (and, consequently all the more expensive) to have more than the minimum spec if you want a computer to perform well. It is safer to choose a machine that has been certified for use with speech recognition. [In an ideal world schools would up-date their hardware to demand but this is unfeasible].

This makes for depressing reading. We would all dearly love to find the miracle solution to the problem so beautifully described by one girl I know: ‘I can write 3 sentences, but speak 3 pages’. Alas, it seems that voice recognition software is not the panacea claimed in the adverts.

I feel that the ’Just Ask’ team has performed a severe disservice to schools and parents in promoting this as a ‘support’ so thoughtlessly.

(I was also somewhat bewildered by the case study of a boy with dyslexia on the Just Ask site. It stated, ‘Jamie was offered a place at a literacy unit’. Do any of these exist nowadays? What about the philosophy of Inclusion and Equality? I feel another post coming on.)





1 thought on “Voice recognition software

  1. Thanks Hilery, for your informative blog. I too was surprised to see the article in the Courier and I imagine that parents/carers of a dyslexic child were wondering why they had not heard of speech recognition software and why it had not been offered at school.

    Probably because it’s not the right product! You have highlighted many of the reasons why it’s not suitable and I would impress again that it’s not easy to train a child’s voice, it’s complicated to correct errors on misheard words resulting in missspelt text and it does involve a level of multi tasking that would only cause more frustration to the pupil.

    Pupils can record their work using an Easi Speak microphone. Very easy to use – click and record, click and playback. Or, download and save your files directly to the PC with the attached USB. Files can be saved as WAV or MP3 format and later transcribed. Ok, it’s speech to sound file not speech to text but could be used for dyslexic pupils to answer questions that otherwise need to be written down.

Leave a Reply

Your email address will not be published.