Towards Novel Researcher Tooling Based on Multimodal AnalyticsReport as inadecuate

Towards Novel Researcher Tooling Based on Multimodal Analytics - Download this document for free, or read online. Document in PDF available to download.

Presented at: Cross Multimodal Learning Analytics Workshop, Tallinn, Estonia, September 13, 2017 Publication date: 2017

Much of educational research today employs a complex mixture of qualitative and quantitative analyses, both during the exploratory and confirmatory phases. However, researchers are still stuck with tools that were developed mainly for single-perspective research. The problem is even more acute in emergent areas with complex, high-frequency data, such as multimodal learning analytics (MMLA). In this position paper, we posit that a new generation of researcher tools are needed to account for this new complexity of research processes. We also anticipate that such new wave of tools should leverage recent advances in computing technology, while keeping humans in the loop. The paper presents a proof-of-concept ongoing study focusing on one of the main points of friction in social sciences research: the manual coding of audio/video. The results of this study, to be presented in the workshop, will illustrate some of the advantages and unsolved challenges in the development of computationally-enhanced researcher tools that can lead to MMLA solutions usable in the real world.

Keywords: research tools ; Multimodal Learning Analytics ; video coding ; content analyses ; automation Reference EPFL-CONF-230187

Author: Prieto, Luis P.; Rodriguez Triana, Maria Jesus


Related documents