Conference PaperPDF Available

Learning Analytics in the Context of COVID-19: A Case Study of Using Network Analysis to Guide Campus Course Offering Plans.

Authors:

Abstract

Network analysis simulations were used to guide decision-makers while configuring instructional spaces on our campus during COVID-19. Course enrollment data were utilized to estimate metrics of student-to-student contact under various instruction mode scenarios. Campus administrators developed recommendations based on these metrics; examples of learning analytics implementation are provided.
A preview of the PDF is not available
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
Current radical changes in the Police service internationally and in England and Wales are being driven by movements to adopt an evidence-based practice (EBP) approach to policing. However this poses a challenge as early adopters have experienced resistance to EBP, a relatively unknown, and more importantly misunderstood approach for policing. This resistance is not limited to police with international research highlighting implementation issues for evidence based medicine, evidence based management, and evidence based teaching. One reason is the lack of training in EBP, which is coupled with recent concerns over the general quality of training and level of professionalism within UK police organisation. There have been international initiatives aimed at increasing learning around evidence based practice. Some UK police forces have adopted approaches from other domains to counteract these problems (e.g. champions, enquiry visits). Mapping clear pathways that link training, experience and evidence-based practice is crucial to developing the capacity for an evidence-based workforce. This paper presents evidence from recent research that used Evidence Cafés and Practitioner Cafés connected to online resources as a route to increase understanding and awareness of evidence based practice amongst frontline police ofcers. Evidence Cafés are coordinated by a knowledge exchange expert with an academic and a police practitioner who facilitate the translation of research into practice. This paper presents evidence of the beneft and limitations of these events. Analytics and learning analytics of events’ online resources also provide insights into these approaches and identify triggers for increased engagement across a wide geographical context.
Technical Report
Full-text available
Collaboration is increasingly identified as an important educational outcome, and most models of twenty-first-century skills include collaboration as a key skill. This report outlines definitions and models of collaboration, teaching of collaboration, skills, assessment methods and a set of recommendations and implications for classroom practice.
Conference Paper
Full-text available
Recent findings in the field of learning analytics have brought to our attention that conclusions drawn from cross-sectional group-level data may not capture the dynamic processes that unfold within each individual learner. In this light, idiographic methods have started to gain grounds in many fields as a possible solution to examine students’ behavior at the individual level by using several data points from each learner to create person-specific insights. In this study, we introduce such novel methods to the learning analytics field by exploring the possible potentials that one can gain from zooming in on the fine-grained dynamics of a single student. Specifically, we make use of Gaussian Graphical Models —an emerging trend in network science— to analyze a single student's dispositions and devise insights specific to him/her. The results of our study revealed that the student under examination may be in need to learn better self-regulation techniques regarding reflection and planning.
Conference Paper
Full-text available
One of the main obstacles impeding the widespread use and adoption of learning analytics is the threat that it poses to students’ data privacy. In this article, we present a proposal for generating person-centered insights for learners by combining data from multiple sources while preserving students' privacy. The basis of our approach is idiographic learning analytics, in which data are collected and insights are generated for each student individually. On the one hand, all the data collection and processing are performed locally on the student’s device, thus preserving student privacy. On the other hand, being based on person-based methods, the idiographic approach helps deliver personalized insights.
Thesis
Full-text available
Gamification, the use of game elements in non game contexts, is becoming widely used in the educational field to enhance learner engagement, motivation, and performance. Many current approaches propose systems where learners use the same game elements. However, recent studies show that learners react differently to different game elements, and that learner motivation, engagement, and performance can vary greatly depending on individual characteristics such as personality, game preferences, and motivation for the learning activity. Results indicate that in some cases game elements that are not adapted to learners can at best fail to motivate them, and at worst demotivate them. Therefore, adapting game elements to individual learner preferences is important. This thesis was part of the LudiMoodle project, dedicated to the gamification of learning resources to enhance learner engagement and motivation. In this thesis, I propose a new system that adapts relevant game elements to learners using individual characteristics, as well as learner engagement. This work is based on previous results in the general gamification field, as well as more specific results from gamification in education. Our main goal is to propose a generic adaptation engine model, instantiated with specific adaptation rules for our educational context. This manuscript presents four major contributions: (1) A general adaptation engine architecture that can be implemented to propose relevant game elements for learners, using both a static and dynamic adaptation approach; (2) A design space and design tools that allows the creation of relevant and meaningful game elements, in collaboration with the various actors of the gamification process (designers, teachers, learners etc.); (3) A static adaptation approach that uses a compromise between both learners' player profile (i.e. preferences for games) and their initial motivation for the learning task; (4) A dynamic learner model built on a trace-based approach to propose an adaptation intervention when an abnormal decrease in engagement is detected. The adaptation engine was implemented in a prototype for the LudiMoodle project, that was used by 258 learners in 4 different secondary schools in France for learning mathematics. To build this prototype we ran a real world study, where learners used this tool as a part of their normal mathematics course. From this study, we ran multiple analyses to better understand the factors that influence the motivational variations of the learners, and how their interaction traces could predict their engagement with the learning task. These analyses served to evaluate the impact of the adaptation of game elements on learner motivation and engagement, and to build the trace based model used for dynamic adaptation.This work represents a significant advancement for the adaptive gamification field, through a generic model for static and dynamic adaptation, with the former based on individual learner characteristics, and the latter on observed learner engagement. I also provide tools and recommendations for designers, to help explore different game element designs. Finally, I discuss these findings in terms of research perspectives, notably with regards to further possible advancements in the dynamic adaptation domain. The full text of my thesis can be found the open archive: https://tel.archives-ouvertes.fr/tel-03125624
Conference Paper
Full-text available
Automatic analysis of students’ collaborative interactions in physi- cal settings is an emerging problem with a wide range of applica- tions in education. However, this problem has been proven to be challenging due to the complex, interdependent and dynamic na- ture of student interactions in real-world contexts. In this paper, we propose a novel framework for the classification of student engage- ment in open-ended, face-to-face collaborative problem-solving (CPS) tasks purely from video data. Our framework i) estimates body pose from the recordings of student interactions; ii) combines face recognition with a Bayesian model to identify and track stu- dents with a high accuracy; and iii) classifies student engagement leveraging a Team Long Short-Term Memory (Team LSTM) neural network model. This novel approach allows the LSTMs to cap- ture dependencies among individual students in their collaborative interactions. Our results show that the Team LSTM significantly improves the performance as compared to the baseline method that takes individual student trajectories into account independently.
Article
In this paper, we review algorithmic bias in education, discussing the causes of that bias and reviewing the empirical literature on the specific ways that algorithmic bias is known to have manifested in education. While other recent work has reviewed mathematical definitions of fairness and expanded algorithmic approaches to reducing bias, our review focuses instead on solidifying the current understanding of the concrete impacts of algorithmic bias in education—which groups are known to be impacted and which stages and agents in the development and deployment of educational algorithms are implicated. We discuss theoretical and formal perspectives on algorithmic bias, connect those perspectives to the machine learning pipeline, and review metrics for assessing bias. Next, we review the evidence around algorithmic bias in education, beginning with the most heavily-studied categories of race/ethnicity, gender, and nationality, and moving to the available evidence of bias for less-studied categories, such as socioeconomic status, disability, and military-connected status. Acknowledging the gaps in what has been studied, we propose a framework for moving from unknown bias to known bias and from fairness to equity. We discuss obstacles to addressing these challenges and propose four areas of effort for mitigating and resolving the problems of algorithmic bias in AIED systems and other educational technology.
Article
With the widespread use of artificial intelligence (AI) systems and applications in our everyday lives, accounting for fairness has gained significant importance in designing and engineering of such systems. AI systems can be used in many sensitive environments to make important and life-changing decisions; thus, it is crucial to ensure that these decisions do not reflect discriminatory behavior toward certain groups or populations. More recently some work has been developed in traditional machine learning and deep learning that address such challenges in different subdomains. With the commercialization of these systems, researchers are becoming more aware of the biases that these applications can contain and are attempting to address them. In this survey, we investigated different real-world applications that have shown biases in various ways, and we listed different sources of biases that can affect AI applications. We then created a taxonomy for fairness definitions that machine learning researchers have defined to avoid the existing bias in AI systems. In addition to that, we examined different domains and subdomains in AI showing what researchers have observed with regard to unfair outcomes in the state-of-the-art methods and ways they have tried to address them. There are still many future directions and solutions that can be taken to mitigate the problem of bias in AI systems. We are hoping that this survey will motivate researchers to tackle these issues in the near future by observing existing work in their respective fields.
Article
This article reports on a study exploring student perspectives on the collection and use of student data for learning analytics. With data collected via a mixed methods approach from 2,051 students across six Australian universities, it provides critical insights from students as a key stakeholder group. Findings indicate that while students are generally comfortable with the use of data to support their learning, they do have concerns particularly in relation to the use of demographic data, location data and data collected from wireless networks, social media and mobile applications. Two key themes emerged related to the need for transparency to support informed consent and personal-professional boundary being critical. This supports findings from other research, which reflects the need for a nuanced approach when providing information to students about the data we collect, including what we are collecting, why and how this is being used. Implications for practice or policy: When implementing the use of dashboards, institutions should ideally include opportunities for students to opt in and out, rather than being set so that students have agency over their data and learning. When undertaking work in relation to learning analytics, staff need to ensure the focus of their work relates to student learning rather than academic research. When institutions and academic staff collect and use student data (regardless of the purpose for doing so), all aspects of these processes need to be transparent to students.