Semantically Annotated Lesson observation Data in Learning Analytics Datasets: a Reference Model

Maka Eradze, María Jesús Rodríguez-Triana, Mart Laanpere
pp. 75 – 91, download
(https://doi.org/10.55612/s-5002-033-004)

Abstract

Learning analytics (LA) and lesson observations are two approaches frequently used to study teaching and learning processes. In both cases, in order to extract meaningful data interpretations, there is a need for contextualization. Previous works propose to enrich LA datasets with observation data and to use the learning design as a framework to guide the data gathering and the later analysis. However, the majority of lesson observation tools collect data that is not compliant with LA datasets. Moreover, the connection between the learning design and the data gathered is not straightforward. This study reflects upon our research-based design towards an LA model for context-aware semantically annotated lesson observations that may be integrated in multimodal LA datasets. Six teachers (out of which 2 were also researchers) with previous experience in lesson observation were engaged in a focus group interview and participatory design session that helped us to evaluate the LA model through the conceptual design of Observata (a lesson observation tool that implements our model). The findings show the feasibility and usefulness of the proposal as well as the potential limitations in terms of adoption.

Keywords: learning design, learning analytics, lesson observations, multimodal learning analytics, semantic annotations.

back to Table of Contents