Information for OHS Case Managers and Rehabilitation Providers
If you are organising speech recognition software for a client or staff member, this information will help you to:
- check that everything is in place;
- understand how speech recognition software can assist with the management of some health conditions and disabilities;
- understand the limitations of speech recognition software; and
- ensure that the process of learning to work by voice is as smooth as possible.
Equipment checklist
Software
Dragon Professional 16 is the latest version, and can be purchased from Voice Recognition. Other retailers can be found by an Internet search for 'Dragon professional 16 '. Note that upgrade pricing is available (usually a discount of around 50%) for those upgrading from Dragon 15 to 16.
A suitable microphone
Any microphone that works with a computer will work with Dragon. However, for best performance (especially if working in an open plan office with background noise), it is important to use a high-quality noise-cancelling headset microphone. Not all microphones suitable for use with computer-based telephony or videoconferencing will work well with Dragon. Headset microphones recommended for use with Dragon.
A suitable computer
Dragon requires Windows 10 or 11 operating system, it doesn't work on Mac computers.
Dragon is resource hungry software and for best performance requires at least 16 GB of RAM, and a fast processor. Using Dragon on an under-spec computer can be a slow and frustrating experience.
Click here for system requirements/recommended computer specs.
Training
Click here for more information on training.
Dragon Professional 16 is the latest version, and can be purchased from Voice Recognition. Other retailers can be found by an Internet search for 'Dragon professional 16 '. Note that upgrade pricing is available (usually a discount of around 50%) for those upgrading from Dragon 15 to 16.
A suitable microphone
Any microphone that works with a computer will work with Dragon. However, for best performance (especially if working in an open plan office with background noise), it is important to use a high-quality noise-cancelling headset microphone. Not all microphones suitable for use with computer-based telephony or videoconferencing will work well with Dragon. Headset microphones recommended for use with Dragon.
A suitable computer
Dragon requires Windows 10 or 11 operating system, it doesn't work on Mac computers.
Dragon is resource hungry software and for best performance requires at least 16 GB of RAM, and a fast processor. Using Dragon on an under-spec computer can be a slow and frustrating experience.
Click here for system requirements/recommended computer specs.
Training
Click here for more information on training.
Other issues
Time
As with learning anything new there is a learning curve, and it is important that the Dragon user has time to focus on Dragon away from the pressure of deadlines. Exactly how much time is required will differ from person to person, depending on how quickly they need to learn to work by voice, what their normal workload is, how challenging their computing tasks are to do by voice etc. Having at least 20 to 30 minutes each day to focus entirely on Dragon for the first 2 to 3 months will help to ensure that your client has time to learn to use Dragon well.
While most experienced Dragon users are able to do a significant proportion of their work by voice at a good speed, it is likely that your client will be less productive than normal during the learning period (usually around 2 to 4 months).
While most experienced Dragon users are able to do a significant proportion of their work by voice at a good speed, it is likely that your client will be less productive than normal during the learning period (usually around 2 to 4 months).
Medical certificates
If your client is working restricted hours, it can be useful if their medical certificate is flexible enough to allow them to be at work during the times I am available for training. For example if a medical certificate states four hours per day, rather than 9 AM to 1 PM, it will be easier to organise training appointments.
Payment for Dragon software, microphones and training
Payment of invoices by the organisation or rehabilitation provider is requested, rather than sending invoices directly to Comcare, as this often results in long delays in payment. The organisation or rehabilitation provider can then seek reimbursement from Comcare if appropriate. Payment terms are 30 days.
Organising training appointments
For government agencies - if Dragon software is being installed by your IT area, it is generally best to wait until Dragon is installed before organising a time for training.
The learning process
How long will it take to learn to work by voice?
For most people, learning to work by voice involves considerable adjustment. It is more than likely that you will be up and running, and able to do some computing tasks by voice, after the first session. However, it will be some time before you can use Dragon for the bulk of your work. For most people, it takes several months to learn to use Dragon confidently.
Training will be most effective if you can dedicate some time to Dragon each working day—even if it is only for 15 minutes or so. Try to find time away from the pressure of deadlines to focus on customising Dragon and reading through the training documents. This is very important in the first few months of learning to work with Dragon.
To keep Dragon software running well, you will need to spend some time on a regular basis for maintenance and customisations—for example by adding new words and phrases to the vocabulary and correcting recognition errors.
How much training is required?
The amount of training required varies from person to person and will depend on:
Training sessions Most training sessions are 1.5 to two hours long, with a minimum session length of 1.5 hours. Generally, training is delivered over 5 or 6 sessions, which are spaced between two and four weeks apart. This allows time to practice and consolidate between sessions. Depending on individual needs, it may take several months to complete Dragon training.
Training content
Training is tailored to specific needs, and can include:
For most people, learning to work by voice involves considerable adjustment. It is more than likely that you will be up and running, and able to do some computing tasks by voice, after the first session. However, it will be some time before you can use Dragon for the bulk of your work. For most people, it takes several months to learn to use Dragon confidently.
Training will be most effective if you can dedicate some time to Dragon each working day—even if it is only for 15 minutes or so. Try to find time away from the pressure of deadlines to focus on customising Dragon and reading through the training documents. This is very important in the first few months of learning to work with Dragon.
To keep Dragon software running well, you will need to spend some time on a regular basis for maintenance and customisations—for example by adding new words and phrases to the vocabulary and correcting recognition errors.
How much training is required?
The amount of training required varies from person to person and will depend on:
- existing knowledge and skills
- which applications are used
- the nature and complexity of computing tasks
- whether you need to do as much as possible by voice, or are able to do a significant proportion of work by hand
- whether training in creating custom commands (macros) is required
- how much time you are able to devote to learning on your own (for example, by reading through the training documents)
Training sessions Most training sessions are 1.5 to two hours long, with a minimum session length of 1.5 hours. Generally, training is delivered over 5 or 6 sessions, which are spaced between two and four weeks apart. This allows time to practice and consolidate between sessions. Depending on individual needs, it may take several months to complete Dragon training.
Training content
Training is tailored to specific needs, and can include:
- dictating, formatting and editing text in Microsoft Word
- working with e-mails by voice
- customising Dragon to minimise use of keyboard and mouse
- accessing documents and applications by voice
- using Internet Explorer or Google Chrome
- working with any other applications in use ( eg: Microsoft Excel, PowerPoint)
Using speech recognition software as a rehabilitation tool - challenges and limitations
Speech recognition software is an effective tool for managing occupational overuse injury and reducing the risk of re-injury, by reducing keyboard and mouse use. However, it is important to be aware of the limitations and challenges. Some of these challenges become less significant over time as the user’s experience and proficiency increases:
- Speech recognition software is complex and temperamental. It occasionally malfunctions, particularly on a large network.
- Active involvement by local IT staff is essential to ensuring the software is properly installed on recommended hardware, and technical support is available when required.
- When working by hand, there is a degree of ‘user error’, for example when the user presses the wrong key. When working by voice, there is a degree of user error as well as a significant degree of system error. That is, when the user speaks clearly but the software does not correctly recognise what was said. In other words, working by voice is not as reliable as working by keyboard and mouse. For example, if you press the ‘k” key by hand, you will always get a ‘k’. When dictating a particular word, at best you are likely to get the word you want 9.5 times out of ten. That is, the best accuracy rate you can expect when using speech recognition software is between 95 and 98%, even with very clear speech. Most speech recognition software users don't achieve more than 95% recognition accuracy in the workplace due to background noise and other issues associated with using the software on a large network.
- Speech recognition software is more challenging to learn and to use than many people expect, particularly if their work is computer intensive and they wish to avoid the use of keyboard and mouse as much as possible. Other factors may add to the level of difficulty. For example — many users are faced with the task of learning to use speech recognition software during a period of acute stress; they are likely to be experiencing pain and discomfort which will affect their memory and ability to concentrate; they might be taking medication which could also affect their ability to concentrate.
- Using speech recognition software can require a lot more concentration and conscious mental activity than working by hand. As an example — punctuation and navigation, which would usually be done by hand without much conscious thought, all need to be verbalised using the correct command names when working by voice. Depending on the task, using speech recognition can sometimes require so much concentration that it becomes difficult to find enough focus for the work task. It can be challenging to use speech recognition when work is urgent and demands one’s full concentration. These difficulties tend to become less significant over time as users gain more experience with the system.
- Proofreading can be a more onerous task when text has been dictated, compared with typing. When working by voice, all words will be correctly spelled but might not be the correct word. ‘Sound-alike’ errors don't attract the eye in the way that a misspelled word does. Many speech recognition users report that they need to proofread more actively, reading for meaning rather than looking for misspelled words.
- Background noise can be a problem. While it is possible to use speech recognition software in an open plan office, the quieter the environment the better. Background noise picked up by the microphone translates into superfluous words on the screen which then have to be deleted, creating extra work. A high degree of concentration is required to use speech recognition software, and users may be more sensitive to background noise and distractions than the average computer user. Providing a quiet environment to work in and the use of a high quality, noise-cancelling microphone will assist in dealing with background noise.
- Not all computing tasks can be undertaken by voice. It is still necessary to use the keyboard and the mouse to some degree. Exactly how much can be achieved by voice will depend on the reliability of software and hardware, the nature of the task, the proficiency and patience of the user and how much time they have to complete the task.
- Dragon speech recognition software does not support all Windows-based applications. It is most effective and reliable in Microsoft applications, particularly in Word, Outlook, Excel and Windows Explorer. Click here for more information on Dragon’s functionality in different applications.
- It can be tiring working by voice for long periods and speech recognition users, like other computer users, need to take regular breaks. Care needs to be taken to avoid vocal strain.
- While speech recognition reduces reliance on keyboard and mouse, the user is still required to work in a fairly static position. The ergonomic problems associated with sedentary work involving little variation in movement are not addressed by speech recognition.
- To achieve a high level of accuracy and avoid constant deletions, it is necessary to avoid making any extraneous sounds while the microphone is turned on. It takes discipline and experience to remember to turn the microphone off any time you want to cough, sneeze, clear your throat, answer the phone, or talk to a colleague.
- To use voice recognition effectively, a significant investment in time is required for ongoing learning, customisation, troubleshooting and housekeeping.