Some relevant work is in a volume edited by Diane Belcher and George Braine
_Academic Writing in a Second Langauge: Essays on Research and Pedagogy_
1995 Ablex. There is also a journal, English for Specific Purposes Journal
(ESPJ), that includes a number of articles, some quite useful, on EAP
(English for Academic Purposes). Some articles have also appeared in TESOL
Quarterly.
The issue of better tests is hard. Certainly, proficiency tests, like
those developed by ACTFL based on the foreign service and CIA testing
paradigms, are more contextualized and performative that paper and pencil
tests like TOEFL. There is also a pretty well established, but still
limited, tradition of performance testing, which invovles putting people
into quasi-naturalistic settings and asking them to do quasi-naturalistic
tasks. Often these are used in workplace settings, e.g., for certification
of TAs or for the English competency of medical workers (Sally Jacoby and
Tim McNamara have a nice piece on this in press that will appear in ESPJ,
they presented a paper version of it last year I believe at AAAL). But I
haven't heard of any tests that seriously seek to track something like the
extended ZOPED for people's langauge in use.
The other issue here, of course, is that communicative and/or task success
is shaped by others' uptake and contributions. Diane Belcher had a nice
piece in ESPJ that looked at the way different mentoring relationships
shaped NNES students' transition from dissertation to career work. I've
been struck at the difference between writing across the curriculum and
English for academic purposes. When the issue is improving the writing of
mainstream students, the response is almost entirely faculty and curriculum
development. When the issue is improving the writing of non-linguistic
mainstream students, the response is almost entirely courses (usually
non-credit) for the students.
There is a lot of literature in these areas, some more useful than other,
but the more you deal with the really meaningful questions of people's
literate practices in specific contexts of use and their trajectories of
participation in those practices, the farther you get from institutional
desires for "efficient" and "easy" tests.
Paul Prior
p-prior who-is-at uiuc.edu
University of Illinois at Urbana-Champaign