Abstract:
New Zealand has recently introduced an assessment of L2 learners' writing proficiency in schools that allows the use of support resources like a bilingual dictionary. By contrast a move to disallow bilingual dictionaries in L2 tests has taken place in the United Kingdom. These contrasting directions reflect on-going debates about the role of dictionaries in tests.
This study investigated tests of writing proficiency in German at the intermediate level where a performance variable was the availability or non-availability of a one-volume bilingual dictionary. Three studies were carried out. The first two focused on a small number of participants (n=6 and n=5) studying German in a New Zealand tertiary institution. The third involved students (n=47) preparing for the intermediate-level Bursary examination in German, a high-stakes examination for 17-18 year-olds in New Zealand. Participants' writing products were analysed quantitatively in terms of scores, lexical richness, type of dictionary look-up and success with look-ups. A qualitative analysis of products investigated quality of language across the two conditions. In the first two studies, frequency and duration of dictionary use was investigated through observation. All participants were given questionnaires, and some were interviewed. This information was used to determine the range of strategies participants employed, and their perceptions of dictionary availability in tests of writing proficiency.
Results revealed that dictionary availability made no statistically significant difference to test scores at the target level. There was, however, some improvement in lexical sophistication. On the other hand, there were also often errors with look-ups, and writing quality when writing without the dictionary was also enhanced by previously learnt phrases. Participants generally felt more confident when they had the dictionary, particularly those with greater prior experience with using dictionaries. Similarly, those with less prior experience tended to prefer the test without the dictionary. The most frequently expressed drawback to having the dictionary was its negative impact on time. On average participants used the dictionary for around 20% of the available time, making around 9 look-ups. The results suggest that use of dictionaries in tests requires a careful weighing of benefits and drawbacks.