Measuring interactional oral competence related to CEFR levels among young learners: results from a Dutch national survey

Abstract Summary

Abstract :

In 2017-2018, a national language survey was conducted in the Netherlands among young learners (aged 11-12) in the last year of primary school, aimed at determining their competence level in English in reading, listening, speaking and vocabulary. The results could be compared to those collected in a previous (2012) survey. Moreover, for the first time, the test results were related to CEFR levels, and spoken interaction skills were tested. Usually spoken interaction is tested using the OPI (Oral Proficiency Interview) model, in which the testee interacts with the testtaker, to avoid interlocutor effects (Sandlund, Sundqvist en Nyroos 2016; van Batenburg, 2018). Instead, due to ecological, psychological and logistic reasons, in the current survey pupils were tested through role-plays in pairs. A multistage (semi-adaptive) test, consisting of three tasks related to A1, A2, B1 CEFR levels, was designed for this purpose. Testees did not have to complete all tasks. Testtakers intervened according to a protocol if they estimated that the testee competence level was insufficient for a higher level task. Both individual performance (related to the CEFR) and pair interaction were assessed. The test was administerd to 288 pairs and led to a variety of combinations of tasks. These combinations were subsequently analysed according to item response theory (IRT). Tasks performances could be related to a competence scale and a standard setting took place to determine the CEFR levels. The test design and the research methodology were proven to be valid instruments for measuring interactional competence with young learners, and the performances in the sample lead to general conclusions for the entire population. 32% of the pupils perform below A1 level, 50% perform at A1 level, 9% at A2 level and 9% at least at B1 level.

Pre-recorded video :