The 2002 CUE Roundtable on Leveling, Streaming and Placement Testing

Curriculum Innovation, Testing and Evaluation: Proceedings of the 1st Annual JALT Pan-SIG Conference.
May 11-12, 2002. Kyoto, Japan: Kyoto Institute of Technology.

The CUE Roundtable on Leveling, Streaming and Placement Testing

by Alan S. Mackenzie, Deborah Bollinger,
Amanda Gillis-Furutaka, and Nobuko Sakurai

Mixed ability classes have been the norm in many universities in Japan for some time. There are a variety of reasons for this, many of which are administrative in nature rather than academic. Increasingly, however, institutions have started to place students in classes based on some form of test or placement procedure given at the beginning of the academic year. This roundtable came out of discussions at the JALT International Conference 2001 between some of the participants on exactly why and how different institutions were leveling students, what options were available to help judge student levels, and what problems there were in implementing such a system within the EFL curriculum.
During the roundtable, discussion had to be halted at one point because of the hostile nature of questions from the floor. Many in the audience, especially those with strong testing backgrounds, seemed to think that the presenters were presenting their ideas as "good" solutions to the problems inherent in leveling students. However, this was far from the intention of the presenters, who simply wanted to share what their universities were doing and discuss the problems and specific institutional constraints to adopting different approaches. Indeed, it was clear from some sections of their presentations that the presenters openly disagreed with some of their institutions' approaches and wanted to explain why they disagreed with it.
The resulting discussion between all involved was animated and brought to the surface many issues. The first of which was the important difference between the terms leveling and streaming. Leveling was defined for the purposes of this discussion as the placement of students into appropriate classes based on their English ability so that they might study with students at a similar level. This could be a global measure of English ability or a skill-specific measure. Movement between levels is fluid, dependent on changes in the measured English ability. Streaming was defined as the initial placement of students into classes based on a placement test. Students within these streams stay in them for the most part and follow courses of study for that stream. The distinction is subtle, but the main difference is the immutability of streams compared to the flexibility of levels. Streaming does not necessitate testing at later dates, whereas leveling requires constant monitoring throughout the course of study to ensure students are always at the appropriate level.
Other issues discussed involved:

Why bother? Mixed ability classes are not problematic if we set up the class in such a way that the higher level students help the lower level ones.
What are appropriate ways to determine levels? Published tests, in-house tests, interviews, university entrance exams, motivational measures, self-assessments, combinations of the above?
How can we make the assessment system fit the curriculum?
What effect does streaming or leveling have on student grades?

[ p. 113 ]

The summaries below of three of the four scenarios presented during the roundtable serve to illustrate the diverse nature of solutions taken and the variety of problems faced in different institutions.

2002 Pan SIG-Proceedings: Topic Index

Complete Pan SIG-Proceedings: Topic Index

[ p. 114 ]