Aligning input and assessment: Pictorial measures in audiovisual vocabulary research

Abstract

A common methodological issue in audiovisual vocabulary research is input-test incongruence. This is particularly evident in studies comparing captioned and non-captioned viewing, where the reliance on written vocabulary tests essentially biases results in favor of captioned groups. Grounded in the complementary frameworks of Transfer-Appropriate Processing and Dual Coding Theory, this brief report argues that the reliance on unimodal (written or aural) tests overlooks the visual channel, a defining element of the audiovisual input. Consequently, a portion of vocabulary knowledge encoded via imagery risks going unmeasured, threatening construct validity. A methodological case is made for integrating pictorial measures to achieve greater authenticity and alignment with the multimodal learning condition. The report concludes by outlining a proposed study to empirically test whether pictorial assessments can capture unique learning gains, thereby providing a more valid and complete account of vocabulary learning from audiovisual input.