Most language educators would agree that testing spoken language is often challenging in educational settings due to its logistically-complex and resource-intensive nature. Uruguay was not an exception, despite the success of the Plan Ceibal one computer-tablet per child initiative in schools.
This presentation reports on a recent collaborative project that has developed and validated a localised semi-direct speaking test for young learners at the end of primary and beginning of secondary education in Uruguay targeting CEFR levels pre-A1 to A2. In this talk, we provide an overview of the project and the results of (i) a pilot with 210 learners and 8 examiners and (ii) a CEFR-linking exercise with 14 expert panellists, while focusing on some of the challenges encountered as well as the technological innovations made in the project.
For example, to counter a lack of interlocutor scaffolding (e.g., Field, 2018; Hasselgreen & Caudwell, 2016) in an online environment, the test incorporated instructional videos in the learners’ L1 with a young teacher-like figure, and allowed repetition of prompts. It also featured game-like tasks to enhance learners’ motivation. Given the beginner-level language ability of the candidature, another challenge was to divide the CEFR A1 category into sub-levels for score-reporting purposes. The CEFR linking panel offered insights into how many granular levels can be reliably evaluated with the specific output language on the test.. As well as sharing the validation evidence that we have obtained in this test development project, we will discuss implications for other speaking assessments for young learners particularly those at CEFR A-levels