Pedagogical Agents Research Articles

The Impact of Emotional Feedback and Elaborated Feedback of a Pedagogical Agent on Multimedia Learning.

This study aimed to explore the impact of emotional feedback and elaborated feedback provided by a pedagogical agent (PA) on learners' emotions, intrinsic motivation, agent perception, cognitive load, and transfer performance in multimedia learning. The experiment was conducted based on an actual undergraduate course. Undergraduate students (N = 117) were randomly assigned to one of the four conditions, where PA's feedback differed by emotional feedback (with vs. without) and elaborated feedback (elaborated feedback vs. knowledge of results). Results revealed that emotional feedback reduced learners' confusion, activated intrinsic motivation, and enhanced agent perception. In addition, elaborated feedback improved intrinsic motivation, agent perception, and transfer performance but reduced germane cognitive load. Surprisingly, there was no significant interaction between emotional feedback and elaborated feedback. These findings had implications for designing a PA with a feedback fulfilling learners' emotional and cognitive needs to maximize multimedia learning.

A diffractive and decolonising reading methodology for education research

AbstractFor white settler researchers aiming to contribute to the work of decolonising education, actively seeking ways to disturb and destabilise long-held onto-epistemological assumptions associated with colonial modernity is important. In this article I investigate how these disturbances might occur in a diffractive and decolonising reading methodology. I outline two prior diffractive reading experiences that drew on decolonial theory and Barad’s diffraction theory: A situated inquiry of the Great Barrier Reef as a pedagogical agent; and a reading of Australian teacher education policy through military imaginaries. In this article I read these prior diffractive reading experiences through one another, attending to further methodological patterns. I identify two connected methods of defamiliarisation that are generative for destabilising colonising ways of knowing, norms and thinking in education. These are: Bringing ostensibly different phenomena together in diffractive relations with one another; and reading difference in the spirit of companionship, that is, in an orientation to learning from difference rather than to master difference. I suggest that if education continues to rely on and wield the same modern critical tools that support colonial-capitalist systems it will be unable to recognise, address and reimagine the continued violence of these systems.

How pedagogical agents communicate with students: A two-phase systematic review

Technological advancements have improved the capabilities of pedagogical agents to communicate with students. However, an increased use of pedagogical agents in learning environments calls for a deeper understanding of student–agent communication to assess the effectiveness of pedagogical agents in learning. This study is a two-phase systematic review of scientific papers on pedagogical agent communication research published between 2010 and 2020, including review papers and original research papers. In the first phase, this study analyses literature reviews and meta-analyses to find the status and research gaps. The findings indicate that pedagogical agents' characteristics and impact on learning have been reviewed, but pedagogical agent communication and its relation to learning have not. In the second phase, the empirical studies of pedagogical agent communication are reviewed and classified into three categories that describe how pedagogical agent communication facilitates students' learning through (1) students' intrapersonal communication processes, (2) interpersonal communication between students and a pedagogical agent, and (3) by facilitating learning in a group. The findings show that pedagogical agent communication can enhance learning through intrapersonal communication of motivation, self-regulation, self-efficacy, and metacognition. At the interpersonal level, pedagogical agents aim to scaffold learning by giving feedback, prompts, and hints from learning processes and learning results. Pedagogical agents also support learning in a group by facilitating discussions and directing students' collaboration. Despite rapid technological advancements, pedagogical agents are not on the level to communicate fluently and human-like, which is likely to reduce their effectiveness and usability in learning. The review concludes that pedagogical agents’ communication needs to be developed toward adaptive, adequate, relational, and logical communication, which requires a multidisciplinary theoretical approach, the use of artificial intelligence, affective computing, and psychometric assessments. Recommendations for future research addressing the gaps identified in this systematic review are discussed.

THE IMPACT OF ANIMATED AND STATIC PEDAGOGICAL AGENTS ON STUDENTS’ LEARNING ACHIEVEMENT IN AN INTRODUCTORY PSYCHOLOGY COURSE

The use of a Pedagogical Agent (PA) to promote multimedia learning comes in various multimedia designs. Many PA designs from prior studies are found to exist without the combination of animation, human recorded voice, interactivity, and narration. In the current study, two interactive multimedia modules (IMMs) were developed, comprising two diferent types of PAs, namely animated PA (APA) with emotive responses and static PA (SPA) designs. Both IMMs conveyed the fundamental of psychology lessons to students. At the end of the intervention, a set of achievement tests was used to measure students’ learning achievement in Introductory Psychology knowledge acquisition. This study is to compare students’ learning achievement between APA and SPA groups in Introductory Psychology through a two-group posttest only experimental design. Thirty-two participants were recruited and equally allocated to APA and SPA groups to interact with the two diferent types of PA designs in the learning modules. The research fndings corroborate the hypothesis that learning through the use of APA with animated characters and spoken text/recorded voice features is more efective than SPA with static characters and written text features in acquiring Introductory Psychology knowledge.
 
 
 Keywords: animation, human recorded voice, learning achievement, multimedia learning, pedagogical agents, psychology
 
 
 
 
 
 

Choice of Pedagogical Agents as Virtual Math Tutors: Perspectives from Children and College Students

Choice of Pedagogical Agents as Virtual Math Tutors: Perspectives from Children and College Students

The Role of Pedagogical Agents in Personalised Adaptive Learning: A Review

Personalised adaptive learning is becoming increasingly popular as a method of providing each student on an online platform with learning experiences that are tailored to their own requirements and peculiarities. This enables learners to go along many learning routes with the shared objective of information and skill development. In such systems, adaptivity and intelligence play distinct roles, with adaptivity being a more data-driven decision-making approach and intelligence being the emulation of human traits in a learning setting. Pedagogical agents, as defined in the field of artificial intelligence, are virtual characters with anthropomorphic features that are introduced for educational reasons. Because e-learning is a continuously growing area, the responsibilities of pedagogical agents change based on the goals that have been established for them. This article provides a systematic evaluation of pedagogical agents’ research and empirical data in e-learning from 2015 to 2022. Their responsibilities will be examined specifically in terms of flexibility and variety, realistic simulation, and their influence on learning: performance improvement, improved motivation, and engagement. The article finishes with a discussion and recommendations on pedagogical agents’ future directions in this ever-changing world of individualised adaptive e-learning.

A Meta-Analytic Review on Embodied Pedagogical Agent Design and Testing Formats

Over the past decade, meta-analyses have been conducted to evaluate the impact of embodied pedagogical agents on learning outcomes. Most review studies have evaluated learning outcomes from the perspective of testing labels such as transfer, retention, recognition, free recall, and various other classifications. This is problematic, because studies can be excluded due to the labeling of the assessment, even though the testing format and test taking strategy are the same. This meta-analysis assesses how embodied pedagogical agents impact learning when measured from the testing format and test taking strategy perspective. The new approach, along with new moderating variables to evaluate agent design and content information, supports previous findings that embodied pedagogical agents significantly increase learning; but there are limitations depending on the agent’s design and the type of information being learned. This meta-analysis advances the field by exploring new ways to think about design elements associated with embodied pedagogical agents.

Can an angry pedagogical agent enhance mental effort and learning performance in a multimedia learning environment?

PurposeThis study aims to examine if a pedagogical agent’s expressed anger, when framed as a feedback cue, can enhance mental effort and learning performance in a multimedia learning environment than expressed happiness.Design/methodology/approachA between-subjects experiment was conducted in which learners engaged with a multimedia learning material that taught programming algorithms, featuring a pedagogical agent who expressed anger or happiness as a feedback cue in response to the learners’ prior performance. Learners completed a self-reported scale and post-test for measuring mental effort and learning performance, respectively.FindingsFemale learners reported higher mental effort and had better learning performance when the pedagogical agent expressed anger than happiness. Male learners reported marginally lower mental effort when the pedagogical agent expressed anger than happiness.Originality/valueThis study focuses on a pedagogical agent’s expressed emotion as social information to learners. Extending from research advocating a pedagogical agent’s positive emotional expression, this study highlights the potential benefits of a pedagogical agent’s negative emotional expression, such as anger, as a cue for learners to enhance learning effort and performance in a multimedia learning environment.

MineObserver: A Deep Learning Framework for Assessing Natural Language Descriptions of Minecraft Imagery

This paper introduces a novel approach for learning natural language descriptions of scenery in Minecraft. We apply techniques from Computer Vision and Natural Language Processing to create an AI framework called MineObserver for assessing the accuracy of learner-generated descriptions of science-related images. The ultimate purpose of the system is to automatically assess the accuracy of learner observations, written in natural language, made during science learning activities that take place in Minecraft. Eventually, MineObserver will be used as part of a pedagogical agent framework for providing in-game support for learning. Preliminary results are mixed, but promising with approximately 62% of images in our test set being properly classified by our image captioning approach. Broadly, our work suggests that computer vision techniques work as expected in Minecraft and can serve as a basis for assessing learner observations.

Design and implementation of an intelligent tutoring system in the view of learner autonomy

PurposeIn intelligent tutoring systems (ITS), learners were often granted limited authority and are forced to obey the decision of the system which might not satisfy their needs. Failure to grant learners sufficient autonomy could yield unexpected effects that hinder learning, including undermining learners’ motivation, priming learners’ aversion to the algorithm. On the contrary, granting learners overwhelming autonomy could also be harmful as the absence of learning support would also have a negative impact on learning. As such, this study aims to design and implement an intelligent tutoring system that offers learners proper autonomy.Design/methodology/approachThe main learning activity in the system is doing exercises, and by finishing exercises learners could earn virtual coins. Based on item response theory, exercises are administered to learners with proper difficulty. Based on a recommended difficulty parameter predicted by the system, learners could manually modify the difficulty of the exercises, they could earn more credits by finishing more challenging exercises. Meanwhile, a pedagogical agent is embedded. Learners could customize the agent’s personality jointly with the system to create the learning context they prefer.FindingsA intelligent tutoring system with proper learner autonomy (LA) is designed and implemented.Originality/valueFew previous researches have noticed the potentially important role that LA plays in ITS. Learning might be facilitated using such a design.

Motivation effect of animated pedagogical agent's personality and feedback strategy types on learning in virtual training environment

BackgroundThe personality and feedback of an animated pedagogical agent (APA) are vital social-emotional features that render the agent perceptually believable. Their effects on learning during virtual training need to be examined. MethodsIn this paper, an explanation model is proposed to clarify the underlying mechanism of how these two features affect learners. Two studies were conducted to investigate the model. In Study 1, the effect of the APA's personality type and feedback strategy on flow experience and performance was reexamined, revealing significant effects of the feedback strategy on flow and performance and a marginally significant effect of the personality type on performance. To explore the mechanism behind these effects, a theoretical model is proposed by distinguishing between intrinsic and extrinsic motivation effects. In Study 2, the model was evaluated, and the APA's personality type was found to significantly influence the factors in the path of the extrinsic motivation effect rather than those in the path of the intrinsic motivation effect. ResultsIn contrast, the feedback strategy affected factors in the path of the intrinsic motivation effect. ConclusionsThese results validated the proposed model. Further distinguishing the two motivation effects is necessary to understand the respective effects of an APA's personality and feedback features on learning experiences and outcomes.

Benefits of Affective Pedagogical Agents in Multimedia Instruction.

The goal of the present study is to explore whether the affective states (happy or neutral) of a pedagogical agent (PA) in an online multimedia lesson yields different learning processes and outcomes, and whether the effects of affective PAs depend on the learners’ emotion regulation strategies and their prior knowledge. In three experiments, undergraduates were asked to view a narrated animation about synaptic transmission that included either a happy PA (smiling expression and enthusiastic voice) or a neutral PA (neutral expression and calm voice) and subsequently took emotions, motivation, cognitive outcomes tests. Across three experiments, the happy PA group reported more positive emotions (ds = 0.70, 0.46, and 0.60) and higher level of motivation (ds = 0.76, 0.49, and 0.51) than the neutral PA group. Moreover, the happy PA prompted higher germane load (d = 0.41) than a neutral PA in Experiment 3. However, adding a happy PA to the screen did not improve learning performance. In addition, in Experiment 2, learners’ usage of emotion regulation strategies moderated the effectiveness of affective PA on positive emotions in learners. Specifically, happy PAs increased the positive emotions of students who used expressive suppression strategy (d = 0.99) but not those who used cognitive reappraisal strategy (d = 0.13). In Experiment 3, the effectiveness of affective PAs was not moderated by learners’ prior knowledge. Results support the cognitive affective theory of learning with media (CATLM) that students are happier and more motivated when they learn from happy PAs than from neutral PAs.

Say What? Learner Reactions to Unexpected Agent Dialogue Moves

For maximally efficient and effective conversation-based intelligent tutoring systems, designers must understand the expectations carried by their intended learners. Strategies programmed into the agents may be interpreted differently by the learners. For example, conversational heuristics in these systems may be biased against false alarms in identifying wrong answers (potentially accepting more incorrect answers), or they may avoid directly answering learner-generated questions in an attempt to encourage more open-ended input. Regardless of pedagogical merit, the learner may view these agents’ dialogue moves as bugs rather than features and respond by disengaging or distrusting future interactions. We test this effect by orchestrating situations in agent-based instruction of electrical engineering topics (through an intelligent tutoring system called AutoTutor) where the pedagogical agent behaves in ways likely counter to learner expectations. To better understand the learning experience of the user, we then measure learner response via think-aloud protocol, eye-tracking, and direct interview. We find that, with few exceptions, learners do not reason that the actions are meant as instructional or technical strategies, but instead broadly understood as errors. This indicates a need for either alteration of agent dialogue strategies, or else additional (implicit or explicit) introduction of the strategies to productively shape learners’ interactions with the system.

Exploring Persona Characteristics in Learning: A Review Study of Pedagogical Agents

In digital interactive learning environments, pedagogical agents are digital characters that assist instruction, providing learners with coaching, feedback, as well as emotional and social support. Recently, increased work has explored how animated, actual, or artificial pedagogical agents impact learning. Some studies found that agents promote learning, while some revealed that agents increased the workload and caused learners to lose focus. Especially poorly designed pedagogical agents would aggravate learners’ external cognitive load, negatively affecting learning performance and experience. This work reviewed how features of pedagogical agent roles affect learners’ retention and transfer performance, as well as the learning experience. It extracted five features, including appearance, gender, facial expression, sound, and movements, then discussed their impact. The survey concludes with the implications for designing effective agents.

An Investigative Study on the Effects of Pedagogical Agents on Intrinsic, Extraneous and Germane Cognitive Load: Experimental Findings With Dyscalculia and Non-Dyscalculia Learners

Several studies have warranted the need to explore the relationship between pedagogical agents in the learning environment and the cognitive load on the learners, specifically, to study the influence on learner populations with different learning and motivation needs. The present work investigates the effects of the presence of a pedagogical agent in the learning environment on intrinsic, extraneous, and germane cognitive load for the dyscalculia and non- learner population. The proposed system intelligently investigates the learner, recommends an exclusive learning dyscalculia path and after tutoring assesses learning gain, and retention. Learner experience and effects of the pedagogical agent on types of cognitive loads are discussed on basis of post tutoring analysis. Samples of 82 learners have been studied, experimental findings based on research questions have been reported and conclusions discussed. Our assumption that ‘a well-articulated and well-designed instructional design exerts a minimal extraneous cognitive load on learners and facilitates learning gain and retention’ is consistent with the obtained results. The result concludes that pedagogical agent does not add to intrinsic and extraneous cognitive load. An improved Germane Cognitive load and a good amount of knowledge retention are noticed post learning.

The effectiveness of gamification in programming education: Evidence from a meta-analysis

This paper aimed at constructing a systematic framework and examining the effect of gamification in programming education through a meta-analysis conducted on 21 empirical studies published in the last decade. We examined the effects of game types, gamification applications, pedagogical agents, programming types, and schooling levels on students' academic achievement, cognitive load, motivation, and thinking skills in programming education by cross-tabulation analysis. Results verified the positive impact of gamification in programming education. Gamification has the largest effect on students' motivation, followed by academic achievement, whereas it has the least effect on students' cognitive load. As for game types, the reasoning strategy game is most effective on academic achievement, while the puzzle game is most effective on motivation. As for gamification application, the games as a competitive mechanism has the greatest impact on students’ thinking skills and motivation. However, when games were adopted as teaching tools or student works, the effects are mainly represented in academic achievement. Pedagogical agents have a limited effect on programming education. With regard to programming types, the effect of gamification is more pronounced in text-based programming rather than graphical programming. This study provided an analytic framework and shed light on potential directions for further studies in the field.

Intelligent Interfaces: Pedagogical Agents and Virtual Humans

Little by little, we are entering the new era, intelligent interfaces are absorbing us more and more every day, and artificial intelligence makes its presence in a stealthy way. Virtual humans that represent an evolution of autonomous virtual agents; they are computer programs and in the future capable of carrying out different activities in certain environments. They will give the illusion of being human; they will have a body, and they will be immersed in an environment. They will have a set of senses that will allow them: 1) Sensations and therefore associated expressions; 2) Communication; 3) Learning; 4) Remembering events, among others. By integrating the above, they will have a personality and autonomy, so they will be able to plan with respect to objectives; allowing them to decide and take actions with their body, in other words, they will count on awareness. The applications will be focused on environments that they will inhabit, or as interfaces that will interact with other systems. The application domains will be multiple; one of them being education. This article shows the design of OANNA like an avatar with the role of pedagogical agent. It was modeled as an affective-cognitive structure related to the teaching-learning process linked to a pedagogical agent that represents the interface of an artilect. OANNA, has the necessary animations for intervention within the teaching-learning process.

Say What? Learner Reactions to Unexpected Agent Dialogue Moves

For maximally efficient and effective conversation-based intelligent tutoring systems, designers must understand the expectations carried by their intended learners. Strategies programmed into the agents may be interpreted differently by the learners. For example, conversational heuristics in these systems may be biased against false alarms in identifying wrong answers (potentially accepting more incorrect answers), or they may avoid directly answering learner-generated questions in an attempt to encourage more open-ended input. Regardless of pedagogical merit, the learner may view these agents’ dialogue moves as bugs rather than features and respond by disengaging or distrusting future interactions. We test this effect by orchestrating situations in agent-based instruction of electrical engineering topics (through an intelligent tutoring system called AutoTutor) where the pedagogical agent behaves in ways likely counter to learner expectations. To better understand the learning experience of the user, we then measure learner response via think-aloud protocol, eye-tracking, and direct interview. We find that, with few exceptions, learners do not reason that the actions are meant as instructional or technical strategies, but instead broadly understood as errors. This indicates a need for either alteration of agent dialogue strategies, or else additional (implicit or explicit) introduction of the strategies to productively shape learners’ interactions with the system.

Enhancement of online education system by using a multi-agent approach

Multi-Agent System (MAS) is popular in the fields where cooperative effort is required to fulfil the purpose of the end product. This study creates a Multi-Agent System as a potential solution for the implementation of an online education system in order to bring about swift and accurate responses from the system and thus a meaningful conversation between the learners and the system. The existing work on online education system employing MAS does not depict the means of flow of information from the browser to the internal MAS and also lacks the adaptivity element with respect to the changing demands of the learners. The study attempts to bridge this gap by employing a system with Event-Condition-Action model and intelligent agents which include a message passing agent to depict the means of message flow from the browser to the MAS and an adaptive course organiser agent which organises adaptive educational content with respect to the changing needs of the learner. The outcome of the study is the design and simulation of the said system as a standalone entity that can be attached to a virtual learning environment (VLE), with the addition of pedagogical agents performing different functionalities to form an adaptive system. The system is evaluated by validating the results generated for various case studies using a validation tool. The expected behaviour is that the resulting course agenda agrees with the learning mode preference of the user provided initially and the results are expected to change according to the changing user preferences through a self-assessment questionnaire. • A Multi-Agent system (MAS) is created as a potential solution for the implementation of an online education system. • Solutions for virtual learning concerns are presented. • System is standalone and can be attached to an existing virtual learning environment (VLE) to make the resulting system more effective in providing appropriate educational resources. • Resulting course agenda agrees with the learning style preference of the user. • Results change according to the changing user preferences through a self-assessment questionnaire.

Animated pedagogical agents enhance learning outcomes and brain activity during learning

AbstractBackgroundWith the rapid popularization of e‐learning, how to improve online learning has aroused widespread concern. A human‐like pedagogical agent (PA) that displays eye gaze and gestures, is often added to online multimedia lessons to increase social connection and improve learning in e‐learning environments. However, there has been a debate about how PAs affect learning processes and learning outcomes.ObjectivesSocial agency theory holds that learners can build a social connection with PAs, which affects learning processes and outcomes. This study seeks to reveal how PAs influence learning by exploring the influence of PAs on learning outcomes and brain activity during learning.MethodsCollege students viewed a multimedia lesson on the process of chemical synaptic transmission either with or without an embodied PA on the left side of the screen that pointed to the graphic as she lectured. During learning students' brain cortical activity was measured by a functional near‐infrared spectroscopy (fNIRS) system, and after learning students completed learning outcomes post‐tests.Results and ConclusionsConsistent with social agency theory, students learning with a PA performed better on learning outcome tests and showed greater brain activity in the social areas of the brain during learning based on fNIRS measures.ImplicationsBy using fNIRS technology, this study provides preliminary new support for the idea that learners engage in social processing during online learning with an embodied PA that leads to improved learning outcomes.