Enhancing student GAI literacy in digital multimodal composing through development and validation of a scale

Abstract

It is widely acknowledged that Generative Artificial Intelligence (GAI) has exerted a greater influence on EFL learners' digital multimodal composing (DMC) process. GAI focuses on creating new textual and multimodal content using large language models (LLMs), and it puts different demands on EFL learners. Although much research has been conducted on EFL learners' AI literacy in various socio-cultural contexts, more attention should now be paid to EFL learners' GAI literacy in the DMC context, a new autonomous model of literacy. It should be noted that even though some studies may concentrate on users' perceptions and experiences with GAI, which may be closely tied to GAI literacy, there lacks the development of a scale for assessing GAI literacy in DMC. Thus, this study attempted to fill these research gaps by developing and validating an applicable and generalizable instrument to measure Chinese EFL learners' multimodal GAI literacy in their DMC process. Two subsamples (n1 = 296, n2 = 294) were randomly invited to respond to the GAIDMCS, and the data were subjected to exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) to test the validity and reliability of the instrument. The findings suggested that a four-factor solution with 17 items can help explain Chinese EFL learners’ GAI literacy in DMC in terms of affective learning, behavior learning, cognitive learning, and ethical learning. Our GAI literacy in DMC scale may help improve GAI education for researchers and practitioners by providing a comprehensive and plausible framework that can serve as an outline for further syllabus design.