We had been wondering how to choose a starter set of vocabulary from
the 1,000's of common words stored on our son's communication aid. We
learned a lot about language and Minspeak by studying the above words
on our machine. We liked it that:
- the source of the sample is young speaking children, and these are
their most used words
- all these words, apart from Ah, Huh, Hum, Ya, Gonna and Guys, and
abbreviations like I'm, I'll, They're
(which are stored as whole phrases like I am, I will, They are), were in our device memory by default
- they are stored in just 30 (from a default set of 92) folders on our
device, so we only need learn an alphabet's worth of items folders to access
them
No wordlists were included in the document that the following excerpts
came from, AAC
Language Issues, pp 13-14, but a language sample analysis, performed
by the authors of the above wordlist, is described:
The study included six preschool children, all Caucasians, from three
classrooms. Samples of over 3,000 words* (please see *note: at end of
section) were collected from each child. The data collection time to
collect samples of this size ranged from 2 to 7 hours. The number of
different words used by these children ranged from 404 to 468.
The average commonality score for the six subjects using the 250 most
frequently occurring words in the composite sample is plotted against
the frequency of word occurrence. The average commonality score across
the subjects of the 25 most frequently occurring words was six, indicating
that all of the subjects used the first 25 words in their communication
samples.
The commonality score decreases with each 25-word set of frequently
occurring words until reaching an average commonality score of 3.68
for the 225th through the 250th most frequently occurring words.
David R. Beukelman, Rebecca S. Jones, and Mary Rowan, "Frequency
of Word Usage by Nondisabled Peers in Integrated Preschool Classrooms,"
Augmentative and Alternative Communication. Vol 5, No. 4, December 1989,
p. 245.
*note: The above information is from page 14 of AAC
Language Issues, a pdf (6,500 k) downloadable from Prentke Romich's
AAC Research & Resources section, where it is a photocopied page that
starts:
"4 for whom a 2,810 word sample was collected. The slight reduction
in sample size for Subject 4 resulted from a miscalculation of sample
size at the time of data collection."
The previous page to this is not included in the article, so we have
guessed and approximated the total number of words collected from each
of the other 5 Subjects to be around 3,000 and we have rearranged the
wording of the first paragraph slightly to make sense here.
|