Helppoa vai vaikeaa arvioida? Yleisten kielitutkintojen suomen kielen arvioijien käsityksiä vironkielisten arvioinnista

Abstract

Finnish raters’ beliefs about assessing Estonian speakers in a Finnish language test: easy or difficult? This paper aims to contribute to our understanding on rater beliefs and views and to promote transparency in rater’s work in the form of reflection. The paper focuses on professional Finnish raters’ beliefs and views of assessing L1 Estonian speakers in the criteria-referenced Finnish language test of the National Certifi- cates of Language Proficiency test system. Finnish raters (N = 27) were asked via email to shortly reflect on what they find easy and what difficult in rating Estonian speakers’ writing and speaking skills in the Finnish language test. The raters reflected on many issues which may or may not have direct influence in their rating thus bringing about assessment issues relating to reliability and validity: rating influenced by views and beliefs outside the established CEFR-based criteria. Content analysis was used to classify and organize the answers. Rater views highlighted the closeness of Estonian and Finnish cultures and language relations, which had both positive and negative impact on their attitudes. The raters regarded this closeness as a merit in that the test tasks and instructions were often easily understandable by the candidates and that the candidates appeared to be trained to taking tests, i.e. had considerable test-taking skills compared to candidates from more distant cultures. As the two languages are closely related, there is also considerable overlap between them, hence the raters pointed out that it was at times difficult to identify whether the candidates were speaking or writing Finnish or Estonian and more importantly whether this influences and should influence their assessment or not. There was negative interference in particular in words, idioms and grammar but in general mistakes made by these candidates were often minor and did not impede understanding. An unsolved issue from the rating perspective was the fact that raters could understand the candidates even though it was not clear which of the two languages was in use. While at the same time individual mistakes were not the focus in assessment but rather understanding the candidate product, the raters wished to point out this fairness issue in rating the skills in a language. The raters generally considered the candidate products easy to read and listen to, thus a common assessment criterion, fluency was usually fulfilled.