A new composition editing program

[anonymous]

I would like to describe some new software that
I have written for ESL/EFL teachers.
The Composition Editing Program by Russ Hodge
(copyright 1994) is software for IBM-compatible
PC's designed to assist teachers in correcting
texts written by their students. Originally this
software was designed for use in ESL/EFL class-
rooms, but a built-in authoring system allows
each user to modify the program to fit a wide
variety of different needs; it can easily be
adapted for use in native language classrooms or
for instruction in other languages.
This program is NOT a "grammar-checker." In
designing the software, I sought a way to inte-
grate a computer into the "traditional" method
of text correction, where a teacher makes comments
on a hand-written student text. With the pro-
gram, a teacher uses a databank containing
several hundred pre-written comments to create a
"comments sheet" for the author of a paper. The
instructor can also compose free comments on
problems that are not covered in the databank.
The student uses this comments sheet (which
is returned with the original text) to find
his or her errors and correct the paper.
With practice, a teacher can correct a paper
in about the same time it takes to correct a
text by hand (or perhaps a little less time).
The correction process creates "error data"
that is stored in files.
Up to this point, the only benefit offered by
the program is that the comments on
errors are very explicit---although the time
needed to make the comment is minimal. In addition,
the format of the comments sheet allows students
to get a very direct, visual impression of the
frequency of their errors. If an error has been
made several times, the different occurrences
appear together on the comments sheet; the student
may correct all the errors of one type before moving
to a completely different problem, and experience
with my students has shown that this method en-
courages them to become sensitive to the "environ-
ment" in which an error is likely to occur in the
future.
Information about students' errors is stored in
data files, and this allows a whole range of further
pedagogical possibilities that "correction by hand"
does not permit. (It is possible to do these things,
but the time required is unbelievable.) The program
has built-in features which allow the teacher to do
the following things instantaneously:
- Create an "error profile" for a whole
group on one assignment---how many students
made specific errors, how many times?
- Chart the "error history" of an indi-
vidual or group to look for signs of progress
or areas of weakness;
- Create a databank of exercises which are
automatically cross-referenced with the errors
in the error databank, and have the computer
automatically search for exercises appropriate
for an individual based on his/her mistakes;
- Search for links between certain themes and
certain grammatical or discursive problems;
- Search for links between students' native
-language backgrounds and specific grammatical
or discursive problems.
If several instructors use the same error data, it
will also be possible to accumulate huge quantities
of data on large groups of students, because the
data will all be in the same format. This will
especially be interesting in investigating relations
between structures and themes or language background
and errors.
An extensive built-in authoring system allows each
teacher to personalize the system, integrating his/her
own concepts of the structure of the language, tailor-
fitting the databanks to his/her students' specific
needs, and integrating the materials that he/she wants
to use.
The error databank can even be erased and rewritten
completely from scratch (for example, in a different
language).
I have been using this system at the University of
Heidelberg for 2 years and have had some very interesting
results.
If you would like more information, or would like to
discuss this approach or my research, feel free to
contact me at
ho...@novell1.gs.uni-heidelberg.de
Russ Hodge