This paper provides a summary of research that shows a substantial
performance gap between native and non-native speakers of
assessment language. Such a performance gap increases with the
level of language demand of the test items. Results of studies
indicate that unnecessary linguistic complexity may impact
reliability and validity of the assessment outcomes. To control for
language factors as a source of bias, this paper introduces the
concept of linguistic modification of assessment through which
unnecessary linguistic complexity of test items is reduced or
eliminated. The paper recommends that test developers be aware
of the impact of linguistic complexity in assessment and should
plan to control for such possible source of bias. |