Videomining for the Assessment of Teacher Skills in Higher Education

Abstract

Recently, learning analytics and the use of big data to help teaching practices has increased. However, despite there is an easy access to teachers’ big data regarding internet, MOOCs, VLEs and other logs; data from onsite teacher classroom activity is also needed, both to evaluate teacher skills both in face-to-face courses, and also in order to triangularize data to help having a complete view of the teacher competencies in blended learning contexts. With the implementation of videomining in onsite contexts, qualitative data from the teacher activity in class can turn into quantitative logs that are useful for real-time assessment and also as part of a bigger learning analytics processes. This study proposes a new methodology, based on the integration of video analysis data mining (videomining), for measuring teacher skills in the classroom. As a first step in the implementation of videomining for the assessment of teacher skills, we explain the process of a video analysis system, MED1A software, an evolution from the existing SME instrument named MED1C; observations conducted in classroom, and the later big data preparation description. This video analysis tool can manage big data from logs and catalogue teacher actions in real-time, multi-observer sources. In particular, we will show the example of teacher Digital Competence (DC) case. Conclusions show that videomining could help retrieving real-time, quantitative and qualitative data from teacher activity without having to spend hours of post-visioning and cataloguing data. This will help researchers, institutions and teachers to retrieve, analyse and later evaluate teacher’s skills and guide them to a real reflection and better practice of their profession.MED1C (Call: H2020-SMEINST-1-2014_24-09-2014 Funding scheme: SME instrument phase. Proposal number: 663957. Proposal acronym: MED1C. Duration (months): 4. Proposal title: Video Management in Medical Practice Activity: ICT-37-2014-1);