Hi all! Since I recently used a computer programme in my research on
essay writing in a timed, exam format, I would like to share my thoughts
on that process. The computer programme I used simply had to identify
those phrases of 3 words or more in the students' writing that were
identical to phrases in the reading that they based their essay on.
While the computer programme did this matching far more consistently,
and quickly than I could, its very consistency was a problem. For
example, it would identify phrases like "a lot of" or "in front of" as
coming from the reading, even though these are simply common
prepositional phrases. It could not recognize phrases where a word was
misspelled. It could not recognize acronyms (e.g CAMPFIRE) if they were
spelled in lower case (Campfire). It could not recognize close
paraphrases where a synonym was substituted for one of the words of the
text, or where the tense of the verb was changed. It could not
recognize the difference between a comma in the middle of a phrase and
the comma in the number 700,000. Both were counted as the end of a
phrase. All of the written texts had to be typed into the programme.
As you can see, even with something as simple as using a computer
programme for counting the number of phrases in the written text, there
are huge logistical and judgemental issues involved. I shall be very
interested to see what kind of rating judgements this wonder software
will be able to carry out, and what format the written standardized test
will take so that the software can be used.
By the way, I missed you all and thought of the fun you were having in
Kamloops. You can be sure I'll be at the next Inkshed!!! Theresa Hyland
-=-=-=-=-=-=-=-=-=-
To leave the list, send a SIGNOFF CASLL command to
[log in to unmask] or, if you experience difficulties,
write to Russ Hunt at [log in to unmask]
For the list archives and information about the organization,
its newsletter, and the annual conference, go to
http://www.stu.ca/inkshed/
-=-=-=-=-=-=-=-=-=-
