No abstract

This paper presents a retrospective analysis of students' use of self-regulated learning strategies while using an educational technology that connects physical and digital learning spaces. A classroom study was carried out in a Data Structures & Algorithms course offered by the School of Computer Science. Students' reviewing behaviors were logged and the associated learning impacts were analyzed by monitoring their progress throughout the course. The study confirmed that students who had an improvement in their performance spent more time and effort reviewing formal assessments, particularly their mistakes. These students also demonstrated consistency in their reviewing behavior throughout the semester. In contrast, students who fell behind in class ineffectively reviewed their graded assessments by focusing mostly on what they already knew instead of their knowledge misconceptions.

The proliferation of educational technology systems has led to the advent of a large number of datasets related to learner interaction. New fields have emerged which aim to use this data to identify interventions that could help the learners become efficient and effective in their learning. However, these systems have to follow user-centered design principles to ensure that the system is usable and the data is of high quality. Human factors literature is limited on the topics regarding Educational Data Mining (EDM) and Learning Analytics (LA). To develop improved educational systems, it is important for human factors engineers to be exposed to these data-oriented fields. This paper aims to provide a brief introduction to the fields of EDM and LA, discuss data visualization and dashboards that are used to convey results to learners, and finally to identify where human factors can aid other fields.

The present research examines a pattern-based measure of communications based on Closed Loop Communications (CLC) and non-content verbal metrics to predict Loss of Separation (LOS) in the National Airspace System (NAS). This study analyzes the transcripts from six retired Air Traffic Controllers (ATC) who participated in three simulated trials of various workloads in a TRACON arrival radar simulation. Results indicated a statistically significant model for predicting LOS based on CLC deviations (CLCD), word count in transmission, words per second, and traffic density. However, more research is required to evaluate the significance of each communication variable to predict LOS.

Air Traffic Controllers (ATCs) communicate with pilots through radio communication. Speech intelligibility is vital in ensuring that the message is conveyed accurately. Factors such as speech rate affect this. Additionally, workload and stress have been shown to affect how people communicate significantly. In this paper, we attempt to analyze the voice data of ATCs who participated in a simulated experiment in the context of these non-verbal aspects of communication, particularly transmission length and speech rate. To better understand, we analyzed our data at two levels: aggregate and individual. Moreover, we focused on a single participant to see how such non-verbal characteristics evolve. Understanding these intricacies would contribute to building automated detectors in real-time voice transmissions that would leverage technology to avert any incidents brought about by stress and workload.