Classifying educational forum posts is a longstanding task in the research of Learning Analytics and Educational Data Mining. Though this task has been tackled by applying both traditional Machine Learning (ML) approaches (e.g., Logistics Regression and Random Forest) and up-to-date Deep Learning (DL) approaches, there lacks a systematic…

In Technology Enhanced Learning field, learning analytics cover multiple research challenges, among which tracking data analysis and data indicator design and visualization. Part of our research effort is dedicated to changing their design process, in order to capitalize them. This would allow us to meet a need in cost savings of design workflow…

In the higher education context, an increasing concern on the technical or instrumental approach permeates attention to academics' data literacies and faculty development. The need for data literacy to deal specifically with the rise of learning analytics in higher education has been raised by some authors, though in spite of some focus on the…

Collapsing categories is a commonly used data reduction technique; however, to date there do not exist principled methods to determine whether collapsing categories is appropriate in practice. With ordinal responses under the partial credit model, when collapsing categories, the true model for the collapsed data is no longer a partial credit…

Like other parts of the social system, education is becoming an information-driven venture: data technologies pervade all levels of the system. This datafication of education seems to take place alongside a general turn to learning that Gert Biesta has called learnification: a progressively singular focus on the manipulable features of individual…

An important aspect of qualitative research is reaching saturation--loosely, a point at which observing more data will not lead to discovery of more information related to the research questions. However, there has been no validated means of objectively establishing saturation. This article proposes a novel quantitative approach to measuring…

Physical literacy is developing as a contested concept with definitional blurring across international contexts, confusing both practitioners and researchers. This paper serves the dual purpose of reporting on an interrogation of concepts associated with physical literacy in academic writing and exploring the use of a text mining data analysis…

This study presents a model for the early identification of students who are likely to fail in an academic course. To enhance predictive accuracy, sentiment analysis is used to identify affective information from text-based self-evaluated comments written by students. Experimental results demonstrated that adding extracted sentiment information…

Phenomenology as a philosophy and a method of inquiry is not limited to an approach to knowing, it is rather an intellectual engagement in interpretations and meaning making that is used to understand the lived world of human beings at a conscious level. Historically, Husserl' (1913/1962) perspective of phenomenology is a science of understanding…

Cognitive diagnosis models (CDMs) allow for the extraction of fine-grained, multidimensional diagnostic information from appropriately designed tests. In recent years, interest in such models has grown as formative assessment grows in popularity. Many dichotomous as well as several polytomous CDMs have been proposed in the last two decades, but…

Big data analytics are prevalent in fields like business, engineering, public health, and the physical sciences, but social scientists are slower than their peers in other fields in adopting this new methodology. One major reason for this is that traditional statistical procedures are typically not suitable for the analysis of large and complex…

We introduce a model which combines principles from psychometric and connectionist paradigms to allow direct Q-matrix refinement via backpropagation. We call this model dAFM, based on augmentation of the original Additive Factors Model (AFM), whose calculations and constraints we show can be exactly replicated within the framework of neural…

A common motivation for adding control variables to statistical models is to reduce the potential for spurious findings when analyzing non-experimental data and to thereby allow for more reliable causal inferences. However, as I show here, unless "all" potential confounding factors are included in an analysis (which is unlikely to be…

Learning analytics has been increasingly outlined as a powerful tool for measuring, analysing, and predicting learning experiences and behaviours. The rising use of learning analytics means that many educational researchers now require new ranges of technical analytical skills to contribute to an increasingly data-heavy field. However, it has been…

Work-integrated learning, also known as co-operative education, allows students to alternate between on-campus classes and off-campus work terms. This provides an enhanced learning experience for students and a talent pipeline for employers. We observe that co-operative job postings are a rich source of information about the required skills,…