Learning theories for artificial intelligence promoting learning processes

This paper discusses a three-level model that synthesizes and unifies existing learning theories to model the roles of artificial intelligence (AI) in promoting learning processes. The model, drawn from developmental psychology, computational biology, instructional design, cognitive science, complexity and sociocultural theory, includes a causal learning mechanism that explains how learning occurs and works across micro, meso and macro levels. The model also explains how information gained through learning is aggregated, or brought together, as well as dissipated, or released and used within and across the levels. Fourteen roles for AI in education are proposed, aligned with the model's features: four roles at the individual or micro level, four roles at the meso level of teams and knowledge communities and six roles at the macro level of cultural historical activity. Implications for research and practice, evaluation criteria and a discussion of limitations are included. Armed with the proposed model, AI developers can focus their work with learning designers, researchers and practitioners to leverage the proposed roles to improve individual learning, team performance and building knowledge communities.

Practitioner notes

What is already known about this topic

• Numerous learning theories exist with significant cross-over of concepts, duplication and redundancy in terms and structure that offer partial explanations of learning.
• Frameworks concerning learning have been offered from several disciplines such as psychology, biology and computer science but have rarely been integrated or unified.
• Rethinking learning theory for the age of artificial intelligence (AI) is needed to incorporate computational resources and capabilities into both theory and educational practices.

What this paper adds

• A three-level theory (ie, micro, meso and macro) of learning that synthesizes and unifies existing theories is proposed to enhance computational modelling and further develop the roles of AI in education.
• A causal model of learning is defined, drawing from developmental psychology, computational biology, instructional design, cognitive science and sociocultural theory, which explains how learning occurs and works across the levels.
• The model explains how information gained through learning is aggregated, or brought together, as well as dissipated, or released and used within and across the levels.
• Fourteen roles for AI in education are aligned with the model's features: four roles at the individual or micro level, four roles at the meso level of teams and knowledge communities and six roles at the macro level of cultural historical activity.

Implications for practice and policy

• Researchers may benefit from referring to the new theory to situate their work as part of a larger context of the evolution and complexity of individual and organizational learning and learning systems.
• Mechanisms newly discovered and explained by future researchers may be better understood as contributions to a common framework unifying the scientific understanding of learning theory.

     

Third wave of measurement in the self-regulated learning field: when measurement and intervention come hand in hand

Measurement is a central issue for the self-regulated learning (SRL) field as SRL is a phenomenon difficult to measure in a reliable and valid way. Here, 3 waves in the history of SRL measurement are identified and profiled. Our focus lies on the third and newest one, which combines measurement and intervention within the same tools. The basis for this approach is located in the reactivity principle via students’ self-monitoring: when students are aware of their actions, they can react and change what is needed. That happens when the measurement tools promote students' self-monitoring which turn part of the intervention then. Examples of this new approach to SRL measurement and guidelines for implementing it are presented.     

Cultivating student skills in self-regulated learning through evaluation of task complexity

In order to self-regulate, students need to honestly reflect on their learning and to take appropriate corrective action. A simple procedure to cultivate student skills in self-regulated learning, known as the Task Evaluation and Reflection Instrument for Student Self-Assessment (TERISSA) is discussed in this paper. TERISSA guides students through two evaluations of the complexity of a task: the first is undertaken just before solving the task and the second straight after completing the task. This study involved 63 undergraduate students and observed a statistically significant difference (p = 0.007) in performance between the students who did (6.1/10) and did not (4.1/10) use TERISSA during tutorials leading up to an assessment task.     

Examining the ontological and epistemic assumptions of research on metacognition,self-regulation and self-regulated learning

While research on metacognition, self-regulation and self-regulated learning is quite mature, these studies have been carried out with varying methodologies and with mixed results. This paper explores the ontological and epistemological assumptions of theories, models and methods used to investigate these three constructs to examine the underlying assumptions of all three. Using oft-cited theories and models of the three constructs along with highly cited studies identified in a previous review of these constructs, this paper examined facets of two popular frameworks: Cartesian-split-mechanistic tradition (CSMT) and the relational tradition specifically looking at the role of intra-individual development, the inclusiveness of categories and notions of causality in these theories, models and methods. While the theories and methods contained elements of both traditions, methods to investigate these constructs relied almost exclusively on assumptions from CSMT. Future directions for research include incorporating more studies examining intra-individual change and multiple notions of causality. Future directions for practice include better contextualisation of research results to strengthen the link between theory and practice.     

Learning analytics to uncover inequality in behavioural engagement and academic attainment in a distance learning setting

Abstract

Although the attainment gap between black and minority ethnic (BME) students and white students has persisted for decades, the potential causes of these disparities are highly debated. The emergence of learning analytics allows researchers to understand how students engage in learning activities based on their digital traces in a naturalistic setting. This study investigates the attainment gap by analysing the differences in behavioural engagement between different ethnic groups. Using multilevel models of academic performance, demographics, and online traces of 149,672 students enrolled in 401 modules in a distance learning setting, we confirmed the existing attainment gap. After controlling for other demographics, module characteristics and engagement, BME students were between 19% and 79% less likely to complete, pass or achieve an excellent grade compared to white students. Given the same academic performance, BME students spent 4-12% more time on studying than white students. While the attainment gap remained persistent after controlling for academic engagement, our study further highlighted the inequality of attainment between BME and white students.     

Using learning analytics to assess student engagement and academic outcomes in open access enabling programmes

Curriculum design, teaching methods, assessments and range of academic support need to be inclusive in Open Access Enabling courses. The findings of this study confirm a correlation between student access to online learning materials and a positive impact on grades in science courses. More specifically, students who frequently use the online learning system to access materials have better assessment and exam results.     

The mediating effects of needs satisfaction on the relationships between prior knowledge and self-regulated learning through artificial intelligence chatbot

The anthropomorphic characteristics of artificial intelligence (AI) can provide a positive environment for self-regulated learning (SRL). The factors affecting adolescents' SRL through AI technologies remain unclear. Limited AI and disciplinary knowledge may affect the students' motivations, as explained by self-determination theory (SDT). In this study, we examine the mediating effects of needs satisfaction in SDT on the relationship between students' previous technical (AI) and disciplinary (English) knowledge and SRL, using an AI conversational chatbot. Data were collected from 323 9th Grade students through a questionnaire and a test. The students completed an AI basic unit and then learned English with a conversational chatbot for 5 days. Confidence intervals were calculated to investigate the mediating effects. We found that students' previous knowledge of English but not their AI knowledge directly affected their SRL with the chatbot, and that satisfying the need for autonomy and competence mediated the relationships between both knowledge (AI and English) and SRL, but relatedness did not. The self-directed nature of SRL requires heavy cognitive learning and satisfying the need for autonomy and competence may more effectively engage young children in this type of learning. The findings also revealed that current chatbot technologies may not benefit students with relatively lower levels of English proficiency. We suggest that teachers can use conversational chatbots for knowledge consolidation purposes, but not in SRL explorations.

Practitioner notes

What is already known about this topic

• Artificial intelligence (AI) technologies can potentially support students' self-regulated learning (SRL) of disciplinary knowledge through chatbots.
• Needs satisfaction in Self-determination theory (SDT) can explain the directive process required for SRL.
• Technical and disciplinary knowledge would affect SRL with technologies.

What this paper adds

• This study examines the mediating effects of needs satisfaction in SDT on the relationship between students' previous AI (technical) and English (disciplinary) knowledge and SRL, using an AI conversational chatbot.
• Students' previous knowledge of English but not their AI knowledge directly affected their SRL with the chatbot.
• Autonomy and competence were mediators, but relatedness was not.

Implications for practice and/or policy

• Teachers should use chatbots for knowledge consolidation rather than exploration.
• Teachers should support students' competence and autonomy, as these were found to be the factors that directly predicted SRL.
• School leaders and teacher educators should include the mediating effects of needs satisfaction in professional development programmes for digital education.

     

Temporal networks in collaborative learning: A case study

Social Network Analysis (SNA) has enabled researchers to understand and optimize the key dimensions of collaborative learning. A majority of SNA research has so far used static networks, ie, aggregated networks that compile interactions without considering when certain activities or relationships occurred. Compressing a temporal process by discarding time, however, may result in reductionist oversimplifications. In this study, we demonstrate the potentials of temporal networks in the analysis of online peer collaboration. In particular, we study: (1) social interactions by analysing learners' collaborative behaviour, part of a case study in which they worked on academic writing tasks, and (2) cognitive interactions through the analysis of students' self-regulated learning tactics. The study included 123 students and 2550 interactions. By using temporal networks, we show how to analyse the longitudinal evolution of a collaborative network visually and quantitatively. Correlation coefficients with grades, when calculated with time-respecting temporal measures of centrality, were more correlated with learning outcomes than traditional centrality measures. Using temporal networks to analyse the co-temporal and longitudinal development, reach, and diffusion patterns of students' learning tactics has provided novel insights into the complex dynamics of learning, not commonly offered through static networks.     

Measuring the complexity of upper primary school children’s self-regulated learning: A multi-component approach

Balancing theoretical and practical issues in the measurement of SRL remains a challenge. This is especially the case for large-scale studies among primary school children’s SRL. In this respect, the present study describes the development and validation of the Children’s Perceived use of Self-Regulated Learning Inventory (CP-SRLI) consisting of nine components. A multistep process was used to develop the questionnaire, including reviews by a teacher and expert panel, cognitive interviews with upper primary school children, and a large-scale administration. The original 109-item questionnaire was then presented to 504 fifth and 463 sixth graders (sample 1). Subsequent to exploratory factor analyses on each component, the factor structure of each component was confirmed by confirmatory factor analyses using an independent second sample (409 fifth and 314 sixth graders), leading to a questionnaire of 75 items. Further, the factor structure of the different components is found to be invariant across boys and girls. The implications of the results and potential avenues for future research are presented and discussed.     

Assessing approaches to learning with nonparametric multidimensional scaling

This article reports on a trace-based assessment of approaches to learning used by middle school aged children who interacted with NASA Mars Mission science, technology, engineering and mathematics (STEM) games in Whyville, an online game environment with 8 million registered young learners. The learning objectives of two games included awareness and knowledge of NASA missions, developing knowledge and skills of measurement and scaling, applying measurement for planetary comparisons in the solar system. Trace data from 1361 interactions were analysed with nonparametric multidimensional scaling methods, which permitted visual examination and statistical validation, and provided an example and proof of concept for the multidimensional scaling approach to analysis of time-based behavioural data from a game or simulation. Differences in approach to learning were found illustrating the potential value of the methodology to curriculum and game-based learning designers as well as other creators of online STEM content for pre-college youth. The theoretical framework of the method and analysis makes use of the Epistemic Network Analysis toolkit as a post hoc data exploration platform, and the discussion centres on issues of semantic interpretation of interaction end-states and the application of evidence centred design in post hoc analysis.

Practitioner notes

What is already known about this topic

• Educational game play has been demonstrated to positively affect learning performance and learning persistence.
• Trace-based assessment from digital learning environments can focus on learning outcomes and processes drawn from user behaviour and contextual data.
• Existing approaches used in learning analytics do not (fully) meet criteria commonly used in psychometrics or for different forms of validity in assessment, even though some consider learning analytics a form of assessment in the broadest sense.
• Frameworks of knowledge representation in trace-based research often include concepts from cognitive psychology, education and cognitive science.

What this paper adds

• To assess skills-in-action, stronger connections of learning analytics with educational measurement can include parametric and nonparametric statistics integrated with theory-driven modelling and semantic network analysis approaches widening the basis for inferences, validity, meaning and understanding from digital traces.
• An expanded methodological foundation is offered for analysis in which nonparametric multidimensional scaling, multimodal analysis, epistemic network analysis and evidence-centred design are combined.

Implications for practice and policy

• The new foundations are suggested as a principled, theory-driven, embedded data collection and analysis framework that provides structure for reverse engineering of semantics as well as pre-planning frameworks that support creative freedom in the processes of creation of digital learning environments.

     

Exploring differences between self-regulated learning strategies of high and low achievers in open distance learning

Open distance students differ in their preparedness for higher education studies. Students who are less self-regulated risk failure and drop out in the challenging milieu of open distance learning. In this study, the differences between the application of self-regulated learning strategies by low and high achievers were explored. A multi-method research design was applied. Quantitative data were statistically analysed by factor analysis (n = 246) and effect sizes. Medium to small effect sizes were found in quantitative data. Qualitative data were collected by means of semi-structured interviews. Low achievers rated their self-regulatory behaviour higher than high achievers, yet qualitative data revealed that high achievers are more self-regulated. The value of this research lies in the identification of low achievers’ use of self-regulated learning, and recognising the need to create awareness of the self-regulated learning skills necessary to support these students.     

Temporal analysis of multimodal data to predict collaborative learning outcomes

The analysis of multiple data streams is a long-standing practice within educational research. Both multimodal data analysis and temporal analysis have been applied successfully, but in the area of collaborative learning, very few studies have investigated specific advantages of multiple modalities versus a single modality, especially combined with temporal analysis. In this paper, we investigate how both the use of multimodal data and moving from averages and counts to temporal aspects in a collaborative setting provides a better prediction of learning gains. To address these questions, we analyze multimodal data collected from 25 9–11-year-old dyads using a fractions intelligent tutoring system. Assessing the relation of dual gaze, tutor log, audio and dialog data to students' learning gains, we find that a combination of modalities, especially those at a smaller time scale, such as gaze and audio, provides a more accurate prediction of learning gains than models with a single modality. Our work contributes to the understanding of how analyzing multimodal data in temporal manner provides additional information around the collaborative learning process.     

The role of learning theory in multimodal learning analytics

This study presents the outcomes of a semi-systematic literature review on the role of learning theory in multimodal learning analytics (MMLA) research. Based on previous systematic literature reviews in MMLA and an additional new search, 35 MMLA works were identified that use theory. The results show that MMLA studies do not always discuss their findings within an established theoretical framework. Most of the theory-driven MMLA studies are positioned in the cognitive and affective domains, and the three most frequently used theories are embodied cognition, cognitive load theory and control–value theory of achievement emotions. Often, the theories are only used to inform the study design, but there is a relationship between the most frequently used theories and the data modalities used to operationalize those theories. Although studies such as these are rare, the findings indicate that MMLA affordances can, indeed, lead to theoretical contributions to learning sciences. In this work, we discuss methods of accelerating theory-driven MMLA research and how this acceleration can extend or even create new theoretical knowledge.

Practitioner notes

What is already known about this topic

• Multimodal learning analytics (MMLA) is an emerging field of research with inherent connections to advanced computational analyses of social phenomena.
• MMLA can help us monitor learning activity at the micro-level and model cognitive, affective and social factors associated with learning using data from both physical and digital spaces.
• MMLA provide new opportunities to support students' learning.

What this paper adds

• Some MMLA works use theory, but, overall, the role of theory is currently limited.
• The three theories dominating MMLA research are embodied cognition, control–value theory of achievement emotions and cognitive load theory.
• Most of the theory-driven MMLA papers use theory ‘as is’ and do not consider the analytical and synthetic role of theory or aim to contribute to it.

Implications for practice and/or policy

• If the ultimate goal of MMLA, and AI in Education in general, research is to understand and support human learning, these studies should be expected to align their findings (or not) with established relevant theories.
• MMLA research is mature enough to contribute to learning theory, and more research should aim to do so.
• MMLA researchers and practitioners, including technology designers, developers, educators and policy-makers, can use this review as an overview of the current state of theory-driven MMLA.

     

Epistemic beliefs: Setting the standards for self-regulated learning

Relations were examined between epistemic beliefs, achievement goals, learning strategies, and achievement. We sought to empirically test Muis’ [Muis, K. R. (2007). The role of epistemic beliefs in self-regulated learning. Educational Psychologist, 42, 173–190] hypothesis that epistemic beliefs influence processes of self-regulated learning via the standards students set for learning once goals are produced. Two hundred one undergraduate students from an educational psychology course completed questionnaires designed to measure the various constructs. Students’ final grades were also collected at the end of the semester. Students’ recollections of course tasks revealed that their epistemic beliefs are activated during learning. Results from structural equation modeling revealed epistemic beliefs influenced the types of achievement goals students adopted, which subsequently influenced the types of learning strategies they used in their education course, and their achievement. Moreover, achievement goals mediated relations between epistemic beliefs and learning strategies, and learning strategies mediated relations between achievement goals and achievement.     

Promoting self-regulated learning to improve achievement: A case study in higher education

Abstract

The aim of this article is to report on a study conducted to assess the effect of an intervention programme to improve SRL and the achievement of a group of poorly performing undergraduate students at the Tshwane University of Technology. SRL was used as theoretical framework. The case study reports on 20 Engineering students who attended learning skills intervention sessions and wrote a college version of the learning and study strategies inventory (LASSI) pre-test and post-test. The intervention consisted of 12 workshop sessions presented over a period of three months. The LASSI pre-test showed that the group scored below the 50th percentile on four scales (anxiety, attitude, selecting main ideas and test-taking strategies). Observed improvements in the post-test scores of the LASSI scales for seven out of ten scales were statistically significant. The students’ academic achievements also improved. The findings are important for improving student success and throughput in South African higher education.     

Using log variables in a learning management system to evaluate learning activity using the lens of activity theory

As the advance of learning technologies and analytics tools continues, learning management systems (LMSs) have been required to fulfil the growing expectations for smart learning. However, the reality regarding the level of technology integration in higher education differs considerably from such expectations or the speed of advances in educational technologies. This research aimed to evaluate the current activation levels and usage patterns of a LMS. A large data-set was analysed, which included the online activity information from 7940 courses. Through data pre-processing, general indicators reflecting login frequencies of the virtual campus and activity-based indicators presenting the activation patterns of diverse functions provided by Moodle were derived. Activity theory was applied to interpret the results of analysis, since it has been recognised as a powerful framework to understand phenomena encompassing interactive systems. Further, time-series investigation over three consecutive semesters allowed observation of historical changes. The results revealed considerably low use of the virtual campus with only slight changes, as well as significantly different activity patterns across course attributes and colleges. Contradictions among components in the activity system are discussed, along with the implications for improving teaching and learning with LMS in higher education.