Learner-system Interaction Research Articles

System design for using multimodal trace data in modeling self-regulated learning

Self-regulated learning (SRL) integrates monitoring and controlling of cognitive, affective, metacognitive, and motivational processes during learning in pursuit of goals. Researchers have begun using multimodal data (e.g., concurrent verbalizations, eye movements, on-line behavioral traces, facial expressions, screen recordings of learner-system interactions, and physiological sensors) to investigate triggers and temporal dynamics of SRL and how such data relate to learning and performance. Analyzing and interpreting multimodal data about learners' SRL processes as they work in real-time is conceptually and computationally challenging for researchers. In this paper, we discuss recommendations for building a multimodal learning analytics architecture for advancing research on how researchers or instructors can standardize, process, analyze, recognize and conceptualize (SPARC) multimodal data in the service of understanding learners' real-time SRL and productively intervening learning activities with significant implications for artificial intelligence capabilities. Our overall goals are to (a) advance the science of learning by creating links between multimodal trace data and theoretical models of SRL, and (b) aid researchers or instructors in developing effective instructional interventions to assist learners in developing more productive SRL processes. As initial steps toward these goals, this paper (1) discusses theoretical, conceptual, methodological, and analytical issues researchers or instructors face when using learners' multimodal data generated from emerging technologies; (2) provide an elaboration of theoretical and empirical psychological, cognitive science, and SRL aspects related to the sketch of the visionary system called SPARC that supports analyzing and improving a learner-instructor or learner-researcher setting using multimodal data; and (3) discuss implications for building valid artificial intelligence algorithms constructed from insights gained from researchers and SRL experts, instructors, and learners SRL via multimodal trace data.

What roles do quality and cognitive absorption play in evaluating cloud-based e-learning system success? Evidence from medical professionals

Purpose The purpose of this study is to propose a hybrid model integrating the expectation-confirmation model with the views of cognitive absorption (CA) theory and updated DeLone and McLean information system success model to examine whether quality factors as antecedents to medical professionals’ beliefs can affect their continuance intention of the cloud-based e-learning system. Design/methodology/approach This study’s sampling frame was taken from among medical professionals working in hospitals with over 300 beds in Taiwan which had implemented the cloud-based learning management system (LMS) with a blend of asynchronous and synchronous technologies. Sample data for this study were collected from medical professionals at six hospitals in Taiwan. The data for this study were gathered by means of a paper-and-pencil survey, and each sample hospital that participated in this study was asked to identify a contact person who could distribute the survey questionnaires to medical professionals who had experience in using the cloud-based LMS in their learning. A total of 600 questionnaires were distributed, and 378 (63.0%) usable questionnaires were analyzed using structural equation modeling in this study. Findings This study proved that medical professionals’ perceived learner–content interaction quality, learner–system interaction quality, service quality, cloud storage service quality and learner–human interaction quality all positively caused their perceived usefulness, confirmation and CA elicited by the cloud-based e-learning system, which jointly explained their satisfaction with the system, and resulted in their continuance intention of the system. Research limitations/implications Several limitations and suggestions may open avenues for future research. First, the limitation of self-reported measures should be considered; future research may combine with qualitative data (e.g. semi-structured, narrative, in-depth interviews, focus group interviews and open-ended questions) to get more complete interpretations of medical professionals’ cloud-based e-learning continuance intention. Next, this study’s data were collected from hospitals in Taiwan only; given this study’s limited scope, future research may generalize this study’s sample to the respondents of other national cultural backgrounds and make cross-country comparisons to enhance the completeness of this study. Finally, this study’ results were based on cross-sectional data; future research may use a longitudinal analysis by taking into account the evolution of medical professionals’ cloud-based e-learning continuance intention over time. Originality/value This study fully evaluates interaction-related and cloud-related quality determinants through an understanding of medical professionals’ state of CA in explaining their cloud-based e-learning system continuance intention that is difficult to expound with only their utilitarian perception of the system. Hence, the results contribute to deep insights into an all-round quality evaluation in the field of medical professionals’ cloud-based e-learning continuance intention, and extrinsic and intrinsic motivators are both taken into consideration in this study’s theoretical development of medical professionals’ cloud-based e-learning continuance intention to acquire a more comprehensive and robust analysis.

AI in Online-Learning Research: Visualizing and Interpreting the Journal Publications from 1997 to 2019

This study reviews the journal publications of artificial intelligence-supported online learning (AIoL) in the Web of Science (WOS) database from 1997 to 2019 taking into account the contributing countries/areas, leading journals, highly cited papers, authors, research areas, research topics, roles of AIoL, and adopted artificial intelligence (AI) algorithms. Results indicate that, from 1997 to 2009, AIoL research focused on the combination of intelligent tutoring systems and distance learning. In 2010–2014, AIoL research emphasized learner-oriented learning. In 2015–2019, learner-system interactions to facilitate personalized, adaptive, and collaborative learning became the main focus. “Intelligent tutoring systems” have played the most important role in AIoL, followed by “profiling and prediction,” and “adaptive systems with personalization.” Accordingly, the roles and research trends as well as several suggestions for future research in the field of AIoL are provided as a reference for researchers and policy makers.

Continued use of an interactive computer game-based visual perception learning system in children with developmental delay

This study developed an interactive computer game-based visual perception learning system for special education children with developmental delay. To investigate whether perceived interactivity affects continued use of the system, this study developed a theoretical model of the process in which learners decide whether to continue using an interactive computer game-based visual perception learning system. The technology acceptance model, which considers perceived ease of use, perceived usefulness, and perceived playfulness, was extended by integrating perceived interaction (i.e., learner-instructor interaction and learner-system interaction) and then analyzing the effects of these perceptions on satisfaction and continued use. Data were collected from 150 participants (rehabilitation therapists, medical paraprofessionals, and parents of children with developmental delay) recruited from a single medical center in Taiwan. Structural equation modeling and partial-least-squares techniques were used to evaluate relationships within the model. The modeling results indicated that both perceived ease of use and perceived usefulness were positively associated with both learner-instructor interaction and learner-system interaction. However, perceived playfulness only had a positive association with learner-system interaction and not with learner-instructor interaction. Moreover, satisfaction was positively affected by perceived ease of use, perceived usefulness, and perceived playfulness. Thus, satisfaction positively affects continued use of the system. The data obtained by this study can be applied by researchers, designers of computer game-based learning systems, special education workers, and medical professionals.

Developing Emotion-Aware, Advanced Learning Technologies: A Taxonomy of Approaches and Features

A growing body of work on intelligent tutoring systems, affective computing, and artificial intelligence in education is exploring creative, technology-driven approaches to enhance learners’ experience of adaptive, positively-valenced emotions while interacting with advanced learning technologies. Despite this, there has been no published work to date that captures this topic’s breadth. We took up this grand challenge by integrating related empirical studies and existing conceptual work and proposing a theoretically-guided taxonomy for the development and improvement of emotion-aware systems. In particular, multiple strategies system developers may use to help learners experience positive emotions are mapped out, including those that require different amounts and types of information about the user, as well as when this information is required. Examples from the literature are provided to illustrate how different emotion-aware system approaches can be combined to take advantage of different types of data, both prior to and during the learner-system interaction. High-level system features that emotion-aware systems can tailor to learners in order to elicit positive emotions are also described and exemplified. Theoretically, the taxonomy is primarily informed by the control-value theory of achievement emotions (Pekrun 2006, 2011) and its assumptions about the relationship between distal and proximal antecedents and the elicitation and regulation of emotion. The taxonomy expands upon a dichotomy of emotion-aware systems proposed by D’Mello and Graesser (2015) and is intended to guide the design of emotion-aware systems that can foster positive emotions during learner-system interactions through the use of varied approaches, data sources, and design features.

Mining learner–system interaction data: implications for modeling learner behaviors and improving overlay models

A growing body of empirical evidence suggests that the adaptive capabilities of computer-based learning environments can be improved through the use of educational data mining techniques. Log-file trace data provides a wealth of information about learner behaviors that can be captured, monitored, and mined for the purposes of discovering new knowledge and detecting patterns of interest. This study aims to leverage these analytical techniques to mine learner behaviors in relation to both diagnostic reasoning processes and outcomes in BioWorld, a computer-based learning environment that support learners to practice solving medical problems and receive formative feedback. In doing so, hidden Markov models are used to model behavioral indicators of proficiency during problem solving, while an ensemble of text classification algorithms are applied to written case summaries that learners' write as an outcome of solving a case in BioWorld. The application of these algorithms characterize learner behaviors at different phases of problem solving which provides corroborating evidence in support of where revisions can be made to provide design guidelines of the system. We con- clude by discussing the instructional design and pedagogical implications for the

Note-taking within MetaTutor: interactions between an intelligent tutoring system and prior knowledge on note-taking and learning

Hypermedia learning environments (HLE) unevenly present new challenges and opportunities to learning processes and outcomes depending on learner characteristics and instructional supports. In this experimental study, we examined how one such HLE—MetaTutor, an intelligent, multi-agent tutoring system designed to scaffold cognitive and metacognitive self-regulated learning (SRL) processes—interacts with learner’s prior domain knowledge to affect their note-taking activities and subsequent learning outcomes. Sixty (N = 60) college students studied with MetaTutor for 120 min and took notes on hypermedia content of the human circulatory system. Log-files and screen recordings of learner-system interactions were used to analyze notes for several quantitative and qualitative variables. Results show that most note-taking was a verbatim copy of instructional content, which negatively related to the post-test measure of learning. There was an interaction between prior knowledge and pedagogical agent scaffolding, such that low prior knowledge students took a greater quantity of notes compared to their high prior knowledge counterparts, but only in the absence of MetaTutor SRL scaffolding; when agent SRL scaffolding was present, the note-taking activities of low prior knowledge students were statistically equivalent to the number of notes taken by their high prior knowledge counterparts. Theoretical and instructional design implications are discussed.

Exploring the roles of interaction and flow in explaining nurses' e-learning acceptance

To provide safe and competent patient care, it is very important that medical institutions should provide nurses with continuing education by using appropriate learning methods. As compared to traditional learning, electronic learning (e-learning) is a more flexible method for nurses' in-service learning. Hence, e-learning is expected to play a pivotal role in providing continuing education for nurses. This study's purpose was to explore the role and relevance of interaction factors, intrinsic motivator (i.e., flow), and extrinsic motivators (i.e., perceived usefulness (PU) and perceived ease of use (PEOU)) in explaining nurses' intention to use the e-learning system. Based on the technology acceptance model (TAM) with the flow theory, this study's research model presents three types of interaction factors, learner-system interaction, instructor-learner interaction, and learner-learner interaction to construct an extended TAM to explore nurses' intention to use the e-learning system. Sample data were gathered from nurses at two regional hospitals in Taiwan. A total of 320 questionnaires were distributed, 254 (79.375%) questionnaires were returned. Consequently, 218 usable questionnaires were analyzed in this study, with a usable response rate of 68.125%. First, confirmatory factor analysis was used to develop the measurement model. Second, to explore the causal relationships among all constructs, the structural model for the research model was tested by using structural equation modeling. First, learner-system interaction, instructor-learner interaction, and learner-learner interaction respectively had significant effects on PU, PEOU, and flow. Next, flow had significant effects on PU and PEOU, and PEOU had a significant effect on PU. Finally, the effects of flow, PU, and PEOU on intention to use were significant. Synthetically speaking, learner-system interaction, instructor-learner interaction, and learner-learner interaction can indirectly make significant impacts on nurses' usage intention of the e-learning system via their extrinsic motivators (i.e., PU and PEOU) and intrinsic motivator (i.e., flow).

THE EFFECTofTypeofLearning EnvironmentonPERCEIVED LEARNING Among Graduate Nursing Students

A quasi-experimental study was conducted to examine the effect of type of learning environment (Internet only, compressed video remote-site, and compressed video host-site) on perceived learning among graduate nursing students. A convenience sample of 49 students enrolled in a graduate nursing research course was selected for the study. The findings showed no differences in perceived learning and final course grades among students enrolled in the three sections of the research course. While the effects of type of learning environment, feelings of connectedness, learner-learner, and learner-system interactions on perceived learning were not statistically significant, learner-instructor interactions had a significant effect on perceived learning. The study findings are encouraging in that students felt that they learned the course content and these perceptions were the same regardless of type of learning environment. The findings also suggest that the quality of instruction is more of an influence on learning than type of teaching modality.

FACILITATING LEARNING FROM COMPUTER-SUPPORTED COLLABORATIVE INQUIRY: THE CHALLENGE OF DIRECTING LEARNERS' INTERACTIONS TO USEFUL ENDS

Chemistry problems can be tough nuts to crack for students, particularly when the problems are of a mathematical type. Collaborative inquiry activities that are embedded in computer-supported learning environments can provide students with opportunities to construct the kind of knowledge required to solve complex chemistry problems. Yet, directing learners' collaborative inquiry in computer-supported learning environments to such useful ends remains a challenge. In this article, we introduce a new approach to facilitating learning from computer-supported collaborative inquiry. The approach we present is a collaboration script that guides pairs of learners through a sequence of inquiry activities and prompts effective interactions in an adaptive fashion. We developed an initial version of a learning environment that included the collaboration script, a simulated chemistry laboratory and a note-taking and mind-mapping tool. In an experimental study involving two conditions (scripted collaboration, unscripted collaboration), we had pairs of students work on a stoichiometry problem using the learning environment. We compared the learning outcomes and learning experiences of the two conditions, and explored learner-learner and learner-system interactions employing a case study methodology. Based on the quantitative and qualitative results, we derived a set of design recommendations for the next development cycle. Our main recommendations are to (a) inform learners about the value of the script support, (b) add algebra practice items and some general chemistry instruction to the learning environment as well as support for interrelating the multiple external representations in the simulated chemistry laboratory, (c) develop tools for assessing learners' prior knowledge and skills in collaborative inquiry learning, general chemistry and basic algebra, (d) increase the degree to which the learning environment can adapt to different learner needs, and (e) improve the interface and interaction design of the simulated chemistry laboratory.

Automatically identifying learners' problem solving strategies in-process solving algorithmic problems

Learners often use learning and programming environments to practice basics of programming and solving algorithmic problems. To adapt the software feedback to each learner's problem solving process, the learning environment will be enhanced with a tool to identify the individual problem solving strategy in-process automatically. Former attribution of problem solving strategies to patterns in recorded learner-system interactions by human researchers should be validated with the help of questionnaires based on the Theory of Reasoned Action. To identify the patterns automatically, speech recognition methods can be used. In this research project each single problem solving strategy is modelled by a hidden markov model.

How do 7th graders solve algorithmic problems?

Informatics education, not only in higher but also in secondary education, is often assisted by special learning software to teach the fundamental ideas of algorithms [2]. In this context pupils also learn the basics of programming using didactically reduced, textbased or visual programming languages. Therefore in Germany, in some federal countries (for example Bavaria), where the basics of algorithms are already taught in the 7 th grade (age 12 to 13 years), age-based learning and programming environments, such as Karel, the robot and Kara, the programmable ladybug [1], are used. Although the design of these environments is age-based, working with them to solve algorithmic problems often causes problems in the classroom. These tools give feedback to the learners based on the analysis of a current solution attempt without taking the previous problem solving process into account. The system messages are often rather technical and therefore hardly helpful especially for weaker learners to enable them to correct arisen problems by themselves. In order to give optimal support to pupils in these situations and therefore improve the learning processes, the learner-system interaction of the used educational software environments should be enhanced and better be adapted to the learners? individual problem solving strategies. The main objective of this research project is to find out, to what extent the automated diagnosis of a problem solving strategy of a learner is possible, and to what extent this knowledge can be used to enhance the learner-system interaction. Starting from the advantages and disadvantages of standardized process observation methods, two software-based research instruments for the system supported diagnosis of the individual proceedings, using the learning environment Kara, were designed and implemented. With the first component learner-system interactions are recorded, the second one provides functions to analyse the collected data. Using test-cases gives a first idea of the quality of the solution attempts. The requirements for the software components resulted from several test scenarios with a small number of participants with different qualification in computer science (from novices to graduating computer science students). During these tests each individual was observed by a researcher and additionally interviewed afterwards. A first version of the implemented instruments was tested in case studies with more than 100 participants (12 to 13 years old) from Bavarian grammar schools to evaluate the suitability for daily use. During the studies the learners were asked to solve three given tasks in a session of 45 minutes, provided by the Kara system, individually (one pupil per computer), but communication between the test persons was allowed. The tasks required knowledge of the control structures (sequence, selection, iteration). The results of these studies indicate that it is possible to identify and to evaluate different problem solving patterns with the help of the developed instruments. To identify different types of learners? strategies it is necessary to combine the various kinds of visualizations of the collected data. To support automatic categorization pattern-recognition methods will be used. The collected ordinal (test-case results) and nominal data can be used for analyses of the correlation between different factors (for example number of error messages or program executions compared with the assessment of the solution attempt) with methods of descriptive statistics.

Technology‐enhanced research: educational ICT systems as research instruments

Determining the potential of educational information and communications technology for enhancing teaching and learning is the subject of a great deal of research activity. This paper argues that the technologies employed to enhance learning can also be exploited for the enhancement of research. Recorded learner–system interactions, dynamically captured computer screen activity and visual attention tracking all provide sources of rich process data that can be triangulated with tradition learning outcome (product) data. Such multiple data source approaches can be used in various research design and methodological contexts—design experiments, microgenetic experiments, case studies and so on. Process‐analytic methods can reveal the fine details of an individual’s learning trajectory and permit the effects of short‐timescale interventions upon longer term learning outcomes to be assessed. In this paper, three illustrations of technology‐enhanced research (TER) are described. The first example (switchER) illustrates an iterative learner‐centred design cycle. The second example (Hyperproof) illustrates how process data can elucidate individual differences in learning. The third example (PATSy, VL‐PATSy) demonstrates how an established online e‐learning system can be used as a tool for identifying students’ clinical reasoning difficulties and how it can then be extended by having additional learning support resources added to it together with a context‐sensitive, rule‐based delivery system. The paper concludes with a discussion of broader TER issues such as the need to anticipate and plan for the analysis of large amounts of rich data. It concludes by suggesting some further ways in which TER might be exploited and provides links to some potentially useful TER tools and resources.

The importance of participant interaction in online environments

An emerging body of research suggests that participant interaction is one of the strongest predictors of success in online environments. However, studies about the effects of participant interaction in a large sample of multiple online environments are rather limited. Using hierarchical modeling techniques, we examine a sample of 40 online MBA courses to determine whether learner–instructor, learner–learner, or learner–system interaction is most significantly related to online course outcomes. Our findings suggest that while collaborative environments were associated with higher levels of learner–learner and learner–system interaction, only learner–instructor and learner–system interaction were significantly associated with increased perceived learning.

Help seeking, learning and contingent tutoring

The focus of this paper is the application of the theory of contingent tutoring to the design of a computer-based system designed to support learning in aspects of algebra. Analyses of interactions between a computer-based tutoring system and 42, 14- and 15-year-old pupils are used to explore and explain the relations between individual differences in learner–tutor interaction, learners’ prior knowledge and learning outcomes. Parallels between the results of these analyses and empirical investigations of help seeking in adult–child tutoring are drawn out. The theoretical significance of help seeking as a basis for studying the impact of individual learner differences in the collaborative construction of ‘zones of proximal development’ is assessed. In addition to demonstrating the significance of detailed analyses of learner–system interaction as a basis for inferences about learning processes, the investigation also attempts to show the value of exploiting measures of on-line help seeking as a means of assessing learning transfer. Finally, the implications of the findings for contingency theory are discussed, and the theoretical and practical benefits of integrating psychometric assessment, interaction process analyses, and knowledge-based learner modelling in the design and evaluation of computer-based tutoring are explored.

Training for Instructional Uses of Multimedia at San Juan College: Toward the Campus of the Future

A multimedia training program for community college faculty includes research results in its training. Recent research in computer assisted multimedia instruction is examined and summarized. Changes in the technology of instruction are described and illustrated. A description of the pedagogical relevance of research to teacher training is offered. The facilities, activities, and outcomes of an instructional multimedia training project for faculty at San Juan College are described. The current principles of instructional systems design, identified by a survey of current research in cai, reflect a pedagogical shift from behaviorism through cognitivism to constructivism. Current research in cai reports that multimedia improves the effectiveness of cai, but provides little consistent empirical evidence explaining why. Questions about learner feedback and control have led to useful results in control/display ergonomics. Continued development of effective, artificially-intelligent tutoring systems has benefited from wider progress in expert systems. Research findings demonstrate that the immanent complexity of hypermedia requires methodical management, that cognitive factors limit the optimum number and effective types of hypermedia link displays, and that close monitoring of learner-system interaction reveals important cognitive differences in individual learners. That same close monitoring of interaction also enables complex instructional systems to respond differentially, thus effectively, across broad differences among individual learners. Different schools of pedagogical thought offer different theoretical explanations for cai's efficacy, but in practice cai's increasing capacity to address individual differences in learning processes is re-emphasizing the importance of teachers' ability to do the same.