Assessing academic speaking skills: from local rating scales to common reference points

The CEFR level descriptions provide fixed points of common reference with remarkable content coherence. While such descriptions offer a consistent and transparent basis for test development, they remain too vague and general for many purposes, particularly for testing language skills at the C levels in university contexts. The CEFR descriptors therefore require adaptation and refinement to suit the specific needs in academic settings. This presentation demonstrates how a local rating scale development and validation project has generated more detailed specifications for academic presentations and interactions at levels C1 and C2. It reports on the validation of two analytic rating scales developed by the Austrian English Language Teaching and Testing (ELTT) group consisting of linguists and language teachers at the universities of Graz, Klagenfurt, Salzburg and Vienna. The scales were examined in a series of studies investigating the proficiency continuum, employing both classical test theory and multi-faceted Rasch analysis. The synthesis of the findings resulted in two modified versions of the scales. The descriptors were then analysed for their common ground. Shared points of reference across the scale categories were abstracted from the specific activities and competencies described on each level, complementing the salient characteristics at levels C1 and C2 as specified in the CEFR and offering a more fine-grained set of specifications for speaking in academic contexts.

Referent/in - Presenter/s

DR. ARMIN BERGER (Vienna University, Austria)

Zeit / Time

  • Freitag / Friday, 24.02.17
  • 14:15 - 14:30 (presenter available for questions)
  • the poster itself is accessible throughout the conference

Raum / Room

Bereich Kaffeestände 3. Ebene

coffee break area 3rd floor

Sprache / Language

English