Corpus-aided error analysis of the development of accuracy between CEFR levels A2–B2
Keywords:
accuracy, corpus-aided error analysis, development of language proficiency, potential occasion analysisAbstract
In this article, the development of an individual’s language proficiency, based here on the language levels of Common European Framework for Reference, is examined with the help of Potential Occasion Analysis (Thewissen, 2015). The development is analyzed from the viewpoint of accuracy, and it is measured by the amount of forms that differ from the conventions of form and usage in the target language. The research represents corpus-aided error analysis (Dagneaux et al., 1998), and it is based on the statistical testing of the amount of errors seen in the nine categories of errors found in the reference levels of language proficiency. The research material consists of the International Corpus of Learner Finnish (ICLFI). The analysis shows that the most remarkable progress happens between levels B1 and B2. In addition, in the progress between levels A2 and B1, there is regression in the form of, for example, errors in adjuncts and phraseology as well as stabilization, which can especially be seen in morphosyntactic and syntactic errors. However, the change between B1 and B2 is so distinct that the increase in morphosyntactic and syntactic accuracy could be made one of the defining aspects of level B2. Earlier studies have indicated level B1 as a distinct culmination point in the development of language (see e.g. Kajander, 2013, p. 93–95; Seilonen, 2013, p. 59–61; Thewissen, 2015), and on the basis of this study's results, this is also the case for the development of language accuracy.
