Gender-sensitive sentiment analysis for estimating the emotional climate in online teacher education

Abstract

Teacher training takes place in distance education to a large extent. Within these contexts, trainers should make use of all the information available to adapt and refine their instructional methods during the training process. Sentiment analysis (SA) can give immediate feedback of the emotions expressed and help in the training process, although it has been used infrequently in educational settings, slow to assess, and bound to interpretative issues, such as gender bias. This research aimed to design and evaluate a SA gender-sensitive method as a proxy to characterize the emotional climate of teacher trainees in an online course. An explanatory case study with mixed methods was implemented among students of the Interuniversity Master of Educational Technologies (N = 48). Participants’ messages were analyzed and correlated with learning achievement and, along with a qualitative study of participants’ satisfaction with the Master’s degree, to validate the effectiveness of the method. Results show that sentiment expression cannot be used to exactly predict participants’ achievement, but it can guide trainers to foresee how participants will broadly act in a learning task and, in consequence, use SA results for tuning and improving the quality of the guidance during the course. Gender differences found in our study support gendered patterns related to the emotional climate, with female participants posting more negative messages than their counterparts. Last but not least, the design of well-adjusted teaching–learning sequences with appropriate scaffolding can contribute to building a positive climate in the online learning environment.

Sentiment analysis for formative assessment in higher education: a systematic literature review

“…The vast majority of publications perform SA at word or sentence level (i.e., attributing a sentiment score to each word or sentence and then computing the overall result), but motivation is a particular mental state, directed towards a particular topic, and usually related to behavioral responses. For this reason, we argue that measuring sentiment polarity provides information about the emotional climate of a group or general emotional state of a student (Usart et al, 2022), rather than providing information about a particular mental or emotional state. Hence, attributing polarity results to a mental state may carry the risk of falling into spurious relationships.…”

Natural Language Processing in Teacher Training : a systematic review

“…For example, Guadalupe et al developed SocialMining, a model that uses a Spanish comment corpus to evaluate teacher performance [27]. Mireia Usart et al designed and evaluated a gender-sensitive SA method to characterize the emotional atmosphere of teacher trainees in online courses and improve their learning [28]. Somdeep Das et al investigated the feedback module of an online learning management system (OLMS) and found that combining machine learning algorithms and NLP techniques with OLMS data can improve teaching efficiency [29].…”

Baseline Study of Different Sentiment Analysis Computing Methods to Enhance Quality Assurance In Teaching and Learning

“…Supervised Machine Learning Algorithms Articles Decision Tree(DT) [1], [2], [5], [6], [8], [12], [13], [15], [16]], [17], [18], [21], [22], [26], [30], [33], [38], [40], [45], [47], [50],[ Support Vector Machine(SVM) [51], [52], [55], [58], [59], [61], [67], [68], [71], [72], [73],…”