Page 57 - Composition in Convergence The Impact of the New Media on Writing Assessment

24 CHAPTER 1

worked writing assignments for the sake of grading efficiency, there
would be significant difficulties beyond what compositionists al-
ready have found. Given the fluid nature of writing in hypertext,
MOO, or other electronic environments, how will the software pack-
age discern when an error is an error or when the "mistake" is a rhe-
torical or linguistic improvisation designed to play to an audience
beyond the teacher? There is far more rhetorical and linguistic play
in electronic communication, and the current predicate analysis or
key-word-in-context formats used by these software programs are
unable to make these "subjective" decisions in papertext situations.
Moreover, because of the multiple contexts available to a reader or
writer in electronic texts, how does the computer program distin-
guish which contexts are appropriate for the material under review?
Even a live instructor holistically reading an electronic text is
weighed down with problems if she is using the typical rubric gener-
ated to read written essays. Although it is used in many ways, the
holistic model is best used for short expository essays written under
very specific conditions. Holistic essay scoring certainly was never
designed for use with public texts. Although certain fields, like public
relations and advertising, use various readability scales, like the
Gunning-Fog Index or Flesch's scale, to determine how reader-
friendly a text is, these gauges do little to help students improve their
writing beyond isolating surface constructions.
Still other questions emerge when we explore how traditional
writing assessment tools could function in a networked writing en-
vironment. How can an individual teacher argue with colleagues to
build a consensus if she is reading a set of electronic texts without
wrenching authority from the student writer who understands the
community for which he writes? How can that same individual
teacher set all the correct parameters for the machine's reading of
that student's electronic text? If one of the purposes for Composi-
tion is to move students into taking more authorial stances in their
writing through public reception of their work, then machine read-
ing of students' electronic work clearly runs counter to that pur-
pose — regardless of how efficient some may believe the evaluation
process may be.
A further concern compositionists should have about the use of
current writing assessment tools being used for students' electronic
communication centers on issues of who controls the text. Because
these various writing and essay-grading software packages make