Learning analytics como apoio na avaliação diagnóstica dos docentes no ambiente virtual de aprendizagem

Abstract

The virtual learning environments (VLE) have been, over the years, an instrument in the dynamization of the teaching-learning process of distance education online. These are called Big Data, and these environments have a huge range of data. These data can be increased in potentially qualified information that, with the aid of technologies such as Learning Analytics (LA), stimulate the generation of relevant indicators for a diagnostic evaluation of teachers during the course of a given discipline, always seeking to elucidate better strategies to achieve the success of the student, consequently promoting an education with better quality. LA in conjunction with Educational Data Mining (EDM) techniques can represent a major breakthrough to resources already existing in the current VLE of higher education institutions, which by default are rather scarce and complex. This thesis identified the difficulties teachers had in assessing VLE, based on a methodology that they used in addition to the literature review and the Goal-Question-Metric (GQM) planning, also developed and evaluated an experiment carried out with teachers who work in distance education at higher education institutions, which were addressed tools that include the objectives of LA, including a new contribution called LAnalize, and validate these for the evaluation evaluation. The results, based on the objectives achieved, showed that teachers need adequate qualification to take advantage of the more sophisticated information and communication technologies (ICT), which managers (managers) have, however, without much assertiveness, motivation for the implantation of new tools in the VLE institutional and that there are tools of LA available in the VLE-Moodle for the diagnostic evaluation support to the profile of the teacher. The tools used to promote actions to achieve the purpose of the teaching evaluation practice in VLE through the LA were: Completion Progress (18.5%), Course Dedication (18.1%) and Level Up! (17.8%), followed by "evaluation / feedback" (13.0%) and "intervention" (12.0%) %). The tool developed, a plugin for LA called LAnalize, stood out among all the tools used in the experiment, being classified by the System Usability Scale (SUS) as excellent with the score of 91.1 points in the usability question, becoming a contribution for the VLE-Moodle community.