"Generation D"
learners
Every school district has set their sights
on developing computer and technology plans
concurrent with the massive advances and
developments in the technology field and the
requirements of students to live and cope in
a technology rich environment. "Generation
D" (digital) learners will need to cope in a
world where understanding technology and
using it appropriately is as essential as
knowing how to read and write. The ability
to use a computer will be just as
fundamental as the ability to use a pencil.
The virtual office and the information age
is challenging everyone whether in the
workplace or in the home, to learn and
function using technology tools.
For students with disabilities, this focus
on technology
is not only essential but is also in many
cases the gateway to extensive improvements in
communication, learning, recreation
and overall function. New technology is
opening the door to potential never before
available on this scale for people with
disabilities.
One of the most
promising technological advancements for
people with disabilities is the field of
speech recognition. During the last decade,
both the software and hardware required for
dictation to the computer has advanced
significantly, and become more affordable
and more accurate. For many people with
wide-ranging disabilities, speech
recognition has become an extremely viable
option. As the technology has improved over
the years, even those with 'not so perfect'
speech or cognitive skills are starting to
make use of the application. People with
different types of speech impediments,
attention problems, physical disabilities,
and cognitive challenges are now learning to
make use of the program with varying levels
of independence. In all cases, even for
those individuals who are unable to use
speech recognition completely independently,
the use of dictation to a computer
represents massive gains in overall levels
of independence, productivity and
achievement.
top
- - - - - - - - - - - - - -
- - - - - - - - -
Improve Writing
A study
exploring dictation and speech recognition
technology as test accommodations for
students with learning disabilities,
summarized that:
|
"Dictation has the potential to
improve the writing performance
of students with learning
disabilities by removing the
barrier created by the
difficulties with mechanics. " |
In this context, 'mechanics' includes the
processes of spelling, capitalization,
punctuation, and handwriting. The study
found that for students who are poor
writers, dictation and speech recognition
resulted in better papers, improve
composition and overall improved performance
than when using handwriting.
Dictation and Speech Recognition Technology
as Test Accommodations
©2004 Council for Exceptional Children.
Fall 2004
top
- - - - - - - - - - - - - -
- - - - - - - - -
Improve Reading
A recent research study (December 2004)
regarding the effectiveness of speech
recognition in training specific reading
skills is discussed in detail
here. This study was
conducted by Eleanor L Higgins and Marshall
H. Raskind in order to assess the
effectiveness of a computer Speech
Recognition-based Program (SRBP).
Studies conducted by the authors have shown
speech recognition technology
(speech-to-text) to be effective in the
remediation of reading and spelling deficits
of children with learning disabilities.
Raskind and Higgins (1999) found that
students with LD ages 9 to 18 who used
discrete speech recognition to write stories
and compositions 50 minutes a week, over a
16 week period, showed significantly higher
scores over a contrast group on measures of
word recognition, spelling, reading
comprehension, and phonological processing.
Under similar conditions, Higgins and
Raskind (2000) found continuous speech
recognition to be effective in remediating
the word recognition and reading
comprehension deficits of students with LD.
The hypotheses proposed by the authors
concerning the SRBP were confirmed for the
outcome measures of word recognition and
reading comprehension, but not for spelling.
After adjusting for age and IQ, significant
differences also were found for phonological
elision and nonword reading (in favor of
SRBP). These results are fairly consistent
with previous research using actual speech
recognition (Higgins & Raskind, 2000;
Raskind & Higgins, 1999), as significant
gains were found for two of the three
outcome measures: word recognition and
reading comprehension. Similar to the
Raskind and Higgins (1999) study, the
present research also found significant
increases in phonological awareness (i.e.,
phonological elision and nonword reading).
Overall, the findings of the present study
suggest that the SRBP was successful in
duplicating those elements of actual speech
recognition thought to be responsible for
gains in the previous studies (i.e., bimodal
presentation of text, practice in the
analysis and discrimination of similar
sounding and looking words, and
interest-driven, self-paced materials).
top
- - - - - - - - - - - - - -
- - - - - - - - -
Learning and Instruction
Programs that
utilize speech recognition to train reading
skills are starting to become available.
These are still in the early stages of
development but are starting to look
extremely promising in terms of their
potential benefit for training reading
skills in early learners. These programs
address specific aspects of learning such as
for the make a win is, fluency, and
comprehension. They use oral reading
exercises that allow students to read,
receive immediate feedback, and hear modeled
reading examples.
|
Soliloquy Learning is a company
which has developed a new program
for training reading fluency. As
students read into the microphone,
the speech recognition technology
tracks their reading performance and
offers help-by pronouncing the word
or providing a definition -when it's
needed. |
  |
- - - - - - - - - - - - - -
- - - - - - - - -
top
Teaching Aids
Rochester Institute of
Technology
Some teachers are using speech recognition in the
classroom as a teaching aid for students who have
hearing challenges, or for those who are unable to take
notes. A research and development group at the National
Technical Institute for the Deaf, in collaboration with
the Information Technology Laboratory at the Rochester
Institute of Technology, is working to adapt new speech
recognition technology to provide real-time
speech-to-text transcription as a support service for
deaf and hard-of-hearing students. In
this system, as the teacher speaks, a hearing
intermediary, or captionist, dictates the words of the
teacher into a microphone and the computer software
converts the dictated words into text.
Development of Speech Recognition and Computer
Communications as a Support for Deaf And Hard-of-Hearing
Students Rochester Institute of
Technology
TypeWell Educational
Transcription System
TypeWell is a transcription system that allows a
transcriber to provide communication access and
notetaking services for students using speech
recognition on a computer, during classroom lectures.
A transcriber uses a notebook computer with abbreviation
software to transcribe meaning-for-meaning what is said
in class lectures and discussions. The students read the
transcription in real-time from a second computer.
Students can also type questions and comments to the
transcriber during class, and even take their own notes
on the reader computer.
Liberated Learning
In
LIBERATED LEARNING classrooms,
lectures are transcribed in real
time using automated speech
recognition and projected to the
class, enabling students to "see"
the lecture. After the lecture,
comprehensive, software-generated
notes are provided in a variety of
formats. |
 |
iCommunicator
ICommunicator™ is a program
which allows teachers to talk to
their students, and have their
speech translated into sign
language, text or speech in 'real
time' as they are teaching!!!
|
 |
Information, Communication and Learning Technologies and
Widening Participation
- - - - - - - - - - - - - -
- - - - - - - - -
Types of Disabilities
and Challenges
Speech
recognition can benefit students with the
following disabilities:
- learning
and attention challenges
- physical
disabilities
-
cognitive disabilities
- speech
impediments.
Challenges:
Visual Motor
- The mechanical aspects of
keyboarding are reduced
significantly. Apart from the
reduced motor load with reduced need
to use the keyboard, there is also a
reduced visual motor load since
there is far less movement of the
eyes from the keyboard to the
screen.
Spelling - The student who
has good spelling and good site
word recognition, has the potential
of a high level of independent use
of the program. However, for those
students to have very poor spelling
and who are unable to visually
isolate incorrect words on the
screen, additional add-on programs
such as
Keystone ScreenSpeaker are
invaluable in allowing the students
to use speech recognition. No longer
do students need to have high levels
of literacy in order to use
NaturallySpeaking. As long as the
student has the potential to use
both NaturallySpeaking and Keystone
ScreenSpeaker in conjunction with
one another, there is potential for
using speech recognition.
Ergonomics - Using speech
recognition reduces the amount of
keyboard and mouse entry required.
This allows the student to be more
flexible in terms of positioning and
posture and also allows for more
movement or alteration of posture
while working. In addition, the
prevalence of repetitive stress
injuries with extended use of the
keyboard or mouse is significantly
reduced. Students who have
difficulty maintaining attention and
posture, will find the more relaxed
postural requirements of using
speech recognition more conducive to
working and producing written
documents. Ergonomics and body
posture including head, body, arms
and legs are still very important
while working. Students should still
maintain a good body posture with
appropriate height of chair, table
and monitor.
Hands-free use - While using
speech recognition, the user has the
option of dictating most or all of
the text using the microphone as
well as performing some or all
computer commands and control
functions. What this means is that
the option of moving the mouse,
opening and closing programs, moving
from one place to the next on the
computer and clicking on buttons are
all options provided by speech
input. Users with learning
disabilities often experience subtle
problems with fine motor
coordination and control. These
users may choose to perform more
functions by voice rather than using
the mouse or keyboard. Users
with significant physical
disabilities may want to use full
command and control options
available in NaturallySpeaking.
Apart from the options available in
NaturallySpeaking, there are other
programs such as QPointer Voice
which primarily offer command and
control functions with speech input.
For students with physical
disabilities wanting to use voice
input to control the mouse as well
as for computer functions, the use
of Dragon NaturallySpeaking
Professional is suggested.
NaturallySpeaking Professional
allows for custom macros and
commands to be added which make the
program extremely powerful for users
with significant physical
disabilities. Users are able to add
their own macros, or can save a lot
of time and efforts by purchasing
third-party add on commands modules
such as the
KnowBrainer Command Software by
Lunis Orcutt. Detailed
information regarding computer
commands and controlled by voice is
provided in the curriculum on this
web site.
Endurance - For those
students who experience greater
levels of fatigue while working,
using speech recognition is of
benefit since reduced physical
energy expenditure is required with
more flexible and relaxed postural
requirements. The option of setting
in a more comfortable seated
position with greater body support,
leaning backwards, results in
increased endurance for written work
and greater focus on the content of
work rather than focus on
controlling and maintaining posture
and arm movements required for
keyboarding or writing.
Concentration and Attention -
Learning how to use speech
recognition and using the program
over time requires good amounts of
concentration and focus. Students
who have concentration difficulties
may find initial training difficult,
and the need for sustained
concentration for editing and
corrections challenging. However,
through work over many years with
learning disabled students, it is
noted that for many students the
positive outcomes in terms of
increased quality of work and rate
of work, as well as the process of
improved rate and production of text
results in improved motivation and
also improved focus of attention.
Speech recognition can be a
significant motivator for a student
who is experienced years of failure
with producing written work. In
addition, although speech
recognition requires the cognitive
skills of memory for commands and
understanding when and how to make
corrections, it reduces the amount
of divided attention normally found
in regular keyboarding tasks.
Regular keyboarding requires the
user to focus attention not only on
the screen and the text which
appears, but also attention on the
keyboard and location of individual
keys, as well as focus on
maintaining good body posture and
alignment for keyboarding. Speech
recognition actually reduces the
amount of divided attention by
allowing the user to focus on their
speech and producing text /
brainstorming with editing and
corrections taking place at a later
stage in the writing process.
Reading and Speech - the use
of speech output together with
speech input (speech recognition) is
an essential part of the whole
speech recognition process. Speech
output in Dragon NaturallySpeaking
occurs in two different forms:
Digitized speech - in which the
user's voice is recorded and can be
played back. As the user dictates
his voice is recorded and during
playback each word is indicated to
show what has been produced as text
related to the user's speech.
Synthesized speech - this is purely
a text-to-speech process in which
the computer reads the text which
has been highlighted on the screen.
This is played back in a
computer/synthesized voice.
The value of digitized speech output
is that each user, regardless of
their consistency of articulation or
speech, is able to have their own
speech or dictation read back to
them in their own voice as they had
initially dictated it. This allows
the user to determine if an error
was an error of dictation or an
error of recognition. Many times,
people may make errors of
articulation and not be aware of it
until they play back their own
dictation.
Synthesized speech or screen
reading, allows users to have
difficulty reading or editing their
work, to have the computer or read
back the exact words on the screen.
If the computer has made an error in
recognizing words that was said, the
user will realize these errors
either through reading the words or
three hearing the words read aloud
through the synthesized speech
option. Synthesized speech is
extremely valuable for students who
have difficulty reading or isolating
text errors on the screen.
NaturallySpeaking has synthesized
speech reading built in to the
program (preferred version and
higher), but this will only read
text that has been recognized on the
screen. An additional programs such
as Keystone ScreenSpeaker is
required to read other parts of text
in NaturallySpeaking such as the
training texts and the correction
lists.
Pronunciation and Articulation -
in previous versions of
NaturallySpeaking, users with poor
articulation and was inconsistent
pronunciation found the use of
speech recognition extremely
frustrating and in most cases
nonfunctional. With the improvements
in the accuracy of the program, more
and more individuals with speech and
articulation problems are managing
to use the program and use it more
successfully. An individual with
speech challenges will find facts
additional training and higher
levels of ongoing corrections are
required. However, these may be
acceptable trade-offs for someone
who is unable to type fast or who
has difficulty with their spelling.
The potential for using speech
recognition if a user has speech
challenges can only be determined on
a one-to-one basis with each
individual. At trial of speech
recognition with significant support
and a structured training program as
well as much practice may result in
an acceptable, functional level of
use. For users with more significant
speech in payments, Dragon
NaturallySpeaking can be used in
spell mode, together with add-on
word prediction software, so that
NaturallySpeaking is almost being
used as a keyboard with individual
keystrokes provided by voice input
rather than finger on a keyboard.
top
- - - - - - - - - - - - - -
- - - - - - - - -
School
Districts and State Standards
Many school districts are developing State
standards which include the use of digital
tools or communication tools. Developing
voice-recognition skills or voice typing at
acceptable speed and accuracy levels as well
as the understanding and use of varied input
devices / technologies are becoming
common-place in State standards.
In
Educational State Standards for
Washington, voice recognition is not
only listed as part of the digital
communication tools coursework, but it also
has the goal of 100 wpm and 95% accuracy.
These are educational standards mandated by
each state, for the benefit of all students.
The inclusion of speech-recognition and a
variety of input devices as part of regular
State standards is of enormous benefits to
students with disabilities! This means that
students with disabilities should be exposed
to a variety of input systems as a matter of
routine.
North
Carolina Public School Standards list
the following standards for keyboarding,
speech recognition and handwriting
recognition:
- DC08.01
Key from straight copy for five
minutes with a minimum of 30 wpm and
five or less errors while making all
corrections.
-
C3PC/MSuppDC08.02 Dictate from
straight copy for three minutes with a
minimum of 100 wpm and 97% accuracy
while making all corrections.
-
C3PC/MSuppDC08.03 Input using
handwriting recognition from
straight copy for one minute with a
minimum of 30 wpm and 95% accuracy.
The standards for the digital tools
framework curricula in many Educational
State Standards are taken from the
National Business Education Standards.
Read this excellent article by Karl
Barksdale:
CHANGING INPUT STANDARDS
Other applications
VOICE ACTIVATED DISPLAY OF AMERICAN SIGN
LANGUAGE FOR AIRPORT SECURITY
|