Rater attitude towards emerging varieties of English: a new rater effect?

Abstract

BackgroundA strong interest in researching World Englishes (WE) in relation to language assessment has become an emerging theme in language assessment studies over the past two decades. While research on WE has highlighted the status, function, and legitimacy of varieties of English language, it remains unclear how raters respond to the results of the global spread of English. Also unclear is whether their attitudes towards the varieties of English constitute a biasing factor in the scores they award in English speaking tests. As such, this study investigates the relationship between rater attitudes towards Indian English as an example of WE, as measured by the “rater attitude instrument” (RAI), and scores that raters awarded to IELTS speech samples produced by Indian examinees.MethodsA total of 96 teacher raters rated six IELTS speech samples and then completed the RAI online. Correlation analysis, MANOVA, and Tukey contrasts were performed to test the extent to which rater attitudes towards Indian English as an example of WE affect rater scoring decisions on IELTS speech samples.ResultsModerate to strong correlations were observed between the RAI and IELTS speech sample scores. The MANOVA results suggest significantly different ratings, with the positive attitude group consistently awarding higher scores to IELTS speech samples in comparison to the negative attitude group on all of the four analytic rating criteria. Furthermore, the RAI appears to be a significant predictor of IELTS speech sample scores.ConclusionA link between rater attitude towards Indian English, as an example of WE, and scoring tendency for Indian examinees may exist in a language assessment context. Thus, as raters reoriented their views, broadened their grasp of WE, and as awareness of WE increased in the language testing community in recent decades, the findings here show that testing agencies must add an understanding of potential rater bias towards WE to the current relevant literature.