The purpose of this study is to assess how the accuracy of advanced learners´ language can be influenced by a focused intervention. Even advanced learners of English frequently manifest rather high occurrence of errors which are believed to be caused by mother tongue (L1) interference and fossilization. These two concepts have recently seen renewed interest of researchers; studies of second language acquisition in advanced learners are, however, lacking. In this research project, two groups, experimental and control, of Czech advanced university students majoring in English were tested for accuracy. All students took a grammaticality judgement test (GJ), which has been proved to be a reliable measure of L2 knowledge, in order to assess their ability to identify errors, and they were also asked to express their certainty in answering (Certainty-based marking - CBM). Students´ written and spoken performance was also tested; short dialogues with participants from both groups were recorded, and all participants were asked to produce a short written text. The experimental group then participated in a 13 week blended course focused on the most problematic areas caused by L1 interference. Both groups were then tested again, the results were compared and analyzed, and the efficiency of the intervention evaluated. Despite a number of limitations, the low number of participants in particular, 29 students in the intervention and 14 students in the control group, and the lack of complete data in all participants, the preliminary results show an increase in accuracy in both groups. This paper presents an analysis of GJ and CBM testing and outlines how the other data will be assessed.
Keywords: accuracy, learner language, L1 interference related errors, fossilization, grammaticality judgement test, certainty-based marking;