A Multidimensional Analysis of Student Evaluations of College English Instruction: An Application of Nonparametric Methods

Abstract

A number of researchers give reason for an exercise of caution
in interpretations of stUdent-supplied evaluative data, Cross -culturally,
student characteristics may be systematically extraneously re lated to
student ratings, Arguably, EFL students with diverse backgrounds may
have their own theories of good teaching which may be revealing more
information about themselves than about their instructors. It may therefore
be hypothesized that extraneous van'ables have a differential impact on
English as a Foreign Language students' ratings. This study attempted to
examine some of the most imporlant extraneous variables which explain
the variation in EFL student ratings of their instructors in Addis Ababa
Unrversify. The data were obtained using a pre-tested adapted version of
the Instructional Assessment Measure (lAM) developed by the Unrversity of
Otago in New Zealand and adapted by a number of unrversities worldwide.
The ten-item scale was rated on a 5 pOint Likert scale. T/7en the data were
duly analyzed using descriptrve and nonparametric inferential techniques.
The results showed that evaluations are a function of gender, level of
learning, program of learning and regional background