It is about combining different formats in which the teacher provides information in class and in which students work on assignments. For example, video and text, 3D models and audio.
As the report’s authors point out, the idea of multimodality is self-evident, but it hasn’t seemed so important before. Now, the development of generative neural networks is increasingly challenging text-based learning tasks. In particular, multimodal evaluation methods are being disseminated.
Preparing content and assessment tools in non-text formats can be difficult. After all, it’s not just about adding pictures or videos to a standard lesson – the point of the approach is to completely transfer some of the learning into other formats. And this requires a good understanding of the capabilities and limitations of different modalities.
It can also be difficult to keep students’ attention – having to switch to different modalities can be distracting.