Interrater reliability of the scoring of the Screening Test of Adolescent Language

Abstract

Group administrations of the Screening Test of Adolescent Language have been successful in identifying students with English-language problems among groups of university students who include many recent immigrants from southeastern Asia. However, scoring several items requires subjective judgement. Accordingly, interrater reliability was investigated by having two independent examiners score the written responses of 299 first-year medical students at two Australian universities. The examiners produced very similar distributions of total scores with means of 20.36 and 19.36 and achieved a high agreement in the categorisation of students with English problems. The Spearman rank-order correlation of 0.83 was high and statistically significant from zero.