Cognitive load theory vs. constructivist approaches: which best leads to efficient,deep learning?

Computer‐assisted learning, in the form of simulation‐based training, is heavily focused upon by the military. Because computer‐based learning offers highly portable, reusable, and cost‐efficient training options, the military has dedicated significant resources to the investigation of instructional strategies that improve learning efficiency within this environment. In order to identify efficient instructional strategies, this paper investigates the two major learning theories that dominate the recent literature on optimizing knowledge acquisition: cognitive load theory (CLT) and constructivism. According to CLT, instructional guidance that promotes efficient learning is most beneficial. Constructivist approaches, in contrast, emphasize the importance of developing a conceptual understanding of the learning material. Supporters of these theories have debated the merits and shortcomings of both positions. However, in the absence of consensus, instructional designers lack a well‐defined model for training complex skills in a rapid, efficient manner. The current study investigates the relative utility of CLT and constructivist‐based approaches for teaching complex skills using a military command and control task. Findings suggest that the acquisition of procedural, declarative, and conceptual knowledge, as well as decision‐making skills, did not differ as a function of the type of instruction used. However, integrated knowledge was slightly better retained by the group provided with CLT‐based instruction. These results are contrary to our expectation that constructivist approaches, which focus on the development and integration of information, would yield better performance in an applied problem‐based environment. Thus, while contemporary researchers continue to defend the use of constructivist strategies for teaching, our research supports earlier findings that question the utility, efficiency, and impact of these strategies in applied domains.     

The effect of computer science on physics learning in a computational science environment

College and high-school students face many difficulties when dealing with physics formulas, such as a lack of understanding of their components or of the physical relationships between the two sides of a formula. To overcome these difficulties some instructors suggest combining simulations' design while learning physics, claiming that the programming process forces the students to understand the physical mechanism activating the simulation. This study took place in a computational-science course where high-school students programmed simulations of physical systems, thus combining computer science (CS) and mathematics with physics learning. The study explored the ways in which CS affected the students' conceptual understanding of the physics behind formulas. The major part of the analysis process was qualitative, although some quantitative analysis was applied as well. Findings revealed that a great amount of the time was invested by the students on representing their physics knowledge in terms of computer science. Three knowledge domains were found to be applied: structural, procedural and systemic. A fourth domain which enabled reflection on the knowledge was found as well, the domain of execution. Each of the domains was found to promote the emergence of knowledge integration processes (Linn & Eylon, 2006, 2011), thus promoting students’ physics conceptual understanding. Based on these findings, some instructional implications are discussed.     

Understanding digital transformation in advanced manufacturing and engineering: A bibliometric analysis,topic modeling and research trend discovery

Digital transformation (DT) is the process of combining digital technologies with sound business models to generate great value for enterprises. DT intertwines with customer requirements, domain knowledge, and theoretical and empirical insights for value propagations. Studies of DT are growing rapidly and heterogeneously, covering the aspects of product design, engineering, production, and life-cycle management due to the fast and market-driven industrial development under Industry 4.0. Our work addresses the challenge of understanding DT trends by presenting a machine learning (ML) approach for topic modeling to review and analyze advanced DT technology research and development. A systematic review process is developed based on the comprehensive DT in manufacturing systems and engineering literature (i.e., 99 articles). Six dominant topics are identified, namely smart factory, sustainability and product-service systems, construction digital transformation, public infrastructure-centric digital transformation, techno-centric digital transformation, and business model-centric digital transformation. The study also contributes to adopting and demonstrating the ML-based topic modeling for intelligent and systematic bibliometric analysis, particularly for unveiling advanced engineering research trends through domain literature.     

Relating research to curriculum and course development

One problem with research on how students learn (using new technology) is that it is usually concerned with a specific topic, i.e. it has rather a narrow domain and it is not clear to what extent the findings are applicable to other domains. This paper discusses research in two rather different domains: novices learning programming and school pupils interpreting cartesian graphs. We discuss the findings in each domain but our main focus is to examine the extent to which general principles and issues apply across both domains. Finally, we discuss some of the implications from our research, for instructional design.In the case of novices learning programming we are concerned with the development of instructional materials for teaching computer concepts. In the case of graph interpretation the research shows that interpreting trends in graphs is a complex skill which is often performed badly by pupils—but which teachers often assume they can do. Here we are concerned with how computers can be used to help pupils to develop graph interpretation skills and basic concepts.In order to gain a detailed understanding of the learning processes involved in both domains we are using Artificial Intelligence techniques to model the cognitive processes. We shall, therefore, also discuss how such techniques can help us improve instructional design.     

Are two heads always better than one? Differential effects of collaboration on students�� computer-supported learning in mathematics

While some studies found positive effects of collaboration on student learning in mathematics, others found none or even negative effects. This study evaluates whether the varying impact of collaboration can be explained by differences in the type of knowledge that is promoted by the instruction. If the instructional material requires students to reason with mathematical concepts, collaboration may increase students’ learning outcome as it promotes mutual elaboration. If, however, the instructional material is focused on practicing procedures, collaboration may result in task distribution and thus reduce practice opportunities necessary for procedural skill fluency. To evaluate differential influences of collaboration, we compared four conditions: individual vs. collaborative learning with conceptual instructional material, and individual vs. collaborative learning with procedural instructional material. The instruction was computer-supported and provided adaptive feedback. We analyzed the effect of the conditions on several levels: Logfiles of students’ problem-solving actions and video-recordings enabled a detailed analysis of performance and learning processes during instruction. In addition, a post-test assessed individual knowledge acquisition. We found that collaboration improved performance during the learning phase in both the conceptual and the procedural condition; however, conceptual and procedural material had a differential effect on the quality of student collaboration: Conceptual material promoted mutual elaboration; procedural material promoted task distribution and ineffective learning behaviors. Consequently, collaboration positively influenced conceptual knowledge acquisition, while no positive effect on procedural knowledge acquisition was found. We discuss limitations of our study, address methodological implications, and suggest practical implications for the school context.     

Effects of different ratios of worked solution steps and problem solving opportunities on cognitive load and learning outcomes

A crucial challenge for instructional designers is to determine the amount of support that is most beneficial for learning. This experiment investigated effects of different ratios of worked solution steps (high assistance) and to-be-solved problem steps (low assistance) on cognitive skill acquisition in geometry. High-school students (N = 125) worked on a geometry lesson in a Cognitive Tutor under five different ratios (from zero worked steps and five to-be-solved steps to four worked steps and one to-be-solved step). Effects on cognitive load and learning outcomes were assessed. We expected the effectiveness of different ratios to vary with the type of learning outcomes (i.e., procedural vs. conceptual knowledge) and the difficulty of the to-be-learned principles. Results showed that for procedural knowledge (but not for conceptual knowledge) problem solving alone was most beneficial for the acquisition of procedural knowledge related to an easy principle. For a difficult principle, no ratio of worked steps and problem solving showed an advantage over another. Problem solving induced more extraneous load than studying worked examples. Thus, in determining optimal amounts of guidance type of knowledge and difficulty of the single to-be learned knowledge chunks should be considered.     

Design and evaluation of planning and mathematical models for generating learning paths

Education research has emphasized the need to develop instructional design tools to facilitate the generation of learning paths for students. Learning paths are important because they enable the personalization and optimization of the learning process. In this work, we present a flexible conceptual framework that allows the representation of curricula information as Artificial Intelligence Planning and Mathematical Programming models to facilitate the generation of learning paths by domain independent algorithms. The resulting models consider a rich set of properties from the education domain, like hierarchical learning structures, enabling conditions, temporal actions, mandatory activities, quality accumulation functions, and metric information. We show that the proposed mathematical models return optimal solutions very efficiently if we relax the total ordering constraints of learning paths. These relaxations allow evaluating greedy planning algorithms to identify the properties from the models that increase the complexity of solution synthesis. We expect that the results of this research can be helpful to education researchers and computer scientists in the quest of scalable systems that capture more flexible standards to model learning and compute more informed learning paths for students.     

Introduction to the special section on learning and security for multi-agent systems

This special section features six articles that provide an overview of the emerging research topics at the intersection learning, security, and multi-agent systems. Recent years have witnessed a surge in the number of works at their intersections, and they have appeared in system and control communities as well as many other communities in artificial intelligence, cyber–physical systems, and economics. The articles in this special section give accessible and comprehensive tutorials and surveys for a broad systems and control audience, covering topics including adversarial machine learning, multi-agent reinforcement learning, cyber resilience, resilient control systems, and game design. It is hopeful that this special section will spawn future interest and cross-disciplinary collaborations in this emerging transdisciplinary research area.     

Learning adaptivity in support of flipped learning: An ontological problem‐solving approach

In recent years, flipped learning has received tremendous attention from educational practitioners and researchers. However, this study argues that existing e‐learning systems mainly serve for learning management and content delivery purposes, although they lack support for flipped learning. As an innovative educational approach, flipped learning requires more pedagogical elements, such as integrated instructional design and adaptive content delivery, to achieve effective direct instruction. This study aims to create a learning adaptivity design to support effective learning in the flipped individual learning space in which the teacher is absent. Because teaching involves various pedagogical and content knowledge sources, we propose a conceptual model of teaching as a function of the knowledge triad of Guideline (G), Teaching Activity (T), and Material (M). To realize this conceptualization, an ontological problem‐solving approach is used for knowledge‐based systems development to integrate the relevant knowledge sources. The knowledge model is created using the Protégé platform to develop OWL‐based domain ontology, task ontology, and SWRL‐based semantic rules to enable inference in the GTM triad for learning adaptivity. The case illustration shows that the knowledge‐based system prototype can adaptively guide student learning in the flipped individual learning space with the knowledge sources integrated.     

An experimental analysis of end-user software training manuals*

Abstract. This paper describes two experimental studies that examine the difference between software training manuals for new users of a groupware software package. The manuals differed in terms of their elaboration of procedural, conceptual and usage information. It was expected that manuals with rich conceptual and usage elaborations would produce the best learning outcomes. No significant differences were found in the learning outcomes of subjects with either high or low levels of previous computer experience. The elements that influence the development of instructional manuals are analysed from the perspective of our findings. The paper suggests that examining the underlying relationships between procedural, conceptual and usage information is an important topic for understanding and evaluating design guidelines for instructional manuals. Research into the methods used to communicate this information may provide insight on the usage and development of instructional manuals.     

Uncovering research trends and topics of communities in machine learning

This paper aims to uncover the research topics in machine learning research communities in a scientific collaboration network (SCN) to enhance the characteristic of systems such as retrieval or recommendation in intelligence-based systems. The existing research mainly focuses on the community evolution and measurement of typical features of the network. It is however unexplored how to identify the research interest of the communities along with authors in each community. A dataset is prepared consisting of 21,906 scientific articles from six top journals in the field of machine learning published from 1988 to 2017. An integrated approach combining the author-topic (AT) model with communities using through the directed affiliations (CoDA) method is explored to identify the research interest of the communities in a scientific collaboration network. The top rank communities are identified using the crank network community prioritization method. Finally, the similarity and dissimilarity of research interest in communities across decades are uncovered using the cosine similarity. The experimental results demonstrate the effectiveness and efficacy of the proposed technique. This study may be helpful for upcoming researchers to explore the research trends and topics in machine learning research communities.

     

Rotational perspective and learning procedural tasks from dynamic media

There have been conflicting accounts regarding the effectiveness of animations for learning. Procedural motor learning represents one of the few areas in which animations have consistently shown to be facilitative. Some have suggested that this benefit is related to activation of the mirror neuron system (MNS), with higher activation leading to better performance. This study examines this explanation, and observed the effects of instructional media (animation vs. static), as a function of viewing perspective (face-to-face vs. over-the-shoulder) on understanding a procedural motor task (knot tying). Results indicate that performance was significantly improved with animations over static images, however this appeared to be most pronounced in situations which matched the learners' own perspective (i.e., over-the-shoulder). These findings have implications for the design of instructional media for procedural motor tasks and provide confirmation of the assertion that appropriate activation of the perceptual system can be leveraged to increase performance.     

An emerging – Social and emerging computing enabled philosophical paradigm for collaborative learning systems: Toward high effective next generation learning systems for the knowledge society

The vast evolution of Social Computing in the last years and the tremendous improvement of novel technologies including cloud computing, open source technologies, recommender systems, personalized knowledge management systems, Big Data Systems, and Open Educational Resources approaches set a challenging context for the establishment of novel high effective approaches to Collaborative learning in both Business and Academia.This editorial provides an overview of a magnificent top quality research collection of articles related to the New Generation Collaborative Learning Systems. It is an opportunity for a scientific debate for the enabling technologies and the required adjustments in Academic Programs and Executives Training programs worldwide. It is a bold contribution to a new philosophical paradigm for the need to promote flexible, open, collaborative learning beyond time, personality, and place constraints. It seems that the old fashioned classroom based learning has to be enriched or in some cases replaced by technological learning innovations fostering collaboration between learners.Another important contribution of this special issue is the in depth discussion of a variety of requirements for next generation learning systems. This can be extremely useful for researchers interested on future research on the domain. Two more special issues on prestigious journals have been confirmed on similar topics for the next year in order to provide a continuity on this fascinating research domain that is directly linked to the vision of the Knowledge Society.     

The interaction of instructional representation and aptitude for performing statistical quality control tasks

Due to the fast pace of changing technologies and increasing job requirements for cognitive skills, workers are forced to update their cognitive skills continually. Consequently, there is a demand in current manufacturing environments for a training program that can improve cognitive task performance of workers. Instructional design is a key step in the design of training programs. Research in Aptitude-Treatment Interaction (ATI) and Instructional Systems Development (ISD) is reviewed in this study, and an integrative approach for instructional design is derived. Results from this study indicate that individual difference in prior achievement is a significant determinant of performance on statistical quality control (SQC) tasks. Also, with the same familiarity level of domain concepts, subjects with higher prior achievement perform better than those with lower prior achievement when more instructional support in procedural instructional representation is presented. However, this evidence is not found with less instructional support in procedural instructional representation. It is concluded that instructional design in SQC or related task domains must account for the individual difference in prior achievement and consider the instructional representation to facilitate the learning process and enhance task performance. © 1997 John Wiley & Sons, Inc.     

An instructional strategy planning model to improve learning and cognition

Developments in cognitive science (e.g., psychology, artificial intelligence, and expert systems) provide a framework to propose an instructional strategy planning model that links knowledge acquisition and employment with specific instructional strategies. The model presented in this article identifies unique computer-based instructional strategies to improve both learning and cognition. Using a meta-learning theory, the storage memory system components of declarative, procedural, and conceptual knowledge are respectively linked to drill and practice, tutorial, and task-oriented simulation strategies. Likewise, for the retrieval memory system components (i.e., differentiation, integration, and creation), the instructional strategies include problem-oriented simulations and self-directed experiences. Each of the instructional strategies are composed of variables and conditions that have been empirically tested and shown to improve specific forms of knowledge acquisition and employment.     

稀疏分层概率自组织图实例迁移学习方法

针对基于实例的迁移学习在关联多源异构领域数据时遇到的数据颗粒度不匹配问题,以单领域分层概率自组织图(HiPSOG)聚类方法为基础,提出一种具有迁移学习能力的稀疏化非监督分层概率自组织图(TSHiPSOG)方法。首先,在源领域和目标领域分别基于概率混合多变量高斯分布生成分层自组织模型以便在多领域中分别提取不同粒度的表示向量,并用稀疏图方法通过概率准则控制模型增长;其次,利用最大信息系数(MIC),在具有富信息的源领域中寻找与目标领域表示向量最相似的表示向量,并利用这些源领域表示向量的类别标签细化目标领域数据分类;最后,在国际通用分类数据集20新闻组数据集和垃圾邮件检测数据集上进行了实验,结果表明算法可以利用源领域的有用信息辅助目标领域的分类问题,并使分类准确率最高提高约15.26%和9.05%;对比其他经典迁移学习方法,通过稀疏分层可以挖掘不同颗粒度的表示向量,分类准确率最高提高约4.48%和4.13%。     

Adaptive educational hypermedia accommodating learning styles: A content analysis of publications from 2000 to 2011

Implementing instructional interventions to accommodate learner differences has received considerable attention. Among these individual difference variables, the empirical evidence regarding the pedagogical value of learning styles has been questioned, but the research on the issue continues. Recent developments in Web-based implementations have led scholars to reconsider the learning style research in adaptive systems. The current study involved a content analysis of recent studies on adaptive educational hypermedia (AEH) which addressed learning styles. After an extensive search on electronic databases, seventy studies were selected and exposed to a document analysis. Study features were classified under several themes such as the research purposes, methodology, features of adaptive interventions and student modeling, and findings. The analysis revealed that the majority of studies proposed a framework or model for adaptivity whereas few studies addressed the effectiveness of learning style-based AEH. Scales were used for learning style identification more than automatic student modeling. One third of the studies provided a framework without empirical evaluation with students. Findings on concrete learning outcomes were not strong enough; however, several studies revealed that suggested models influenced student satisfaction and success. Current trends, potential research gaps and implications were discussed.     

Explaining authors’ contribution to pivotal artifacts during mass collaboration in the Wikipedia’s knowledge base

This article discusses the relevance of large-scale mass collaboration for computer-supported collaborative learning (CSCL) research, adhering to a theoretical perspective that views collective knowledge both as substance and as participatory activity. In an empirical study using the German Wikipedia as a data source, we explored collective knowledge as manifested in the structure of artifacts that were created through the collaborative activity of authors with different levels of contribution experience. Wikipedia’s interconnected articles were considered at the macro level as a network and analyzed using a network analysis approach. The focus of this investigation was the relation between the authors’ experience and their contribution to two types of articles: central pivotal articles within the artifact network of a single knowledge domain and boundary-crossing pivotal articles within the artifact network of two adjacent knowledge domains. Both types of pivotal articles were identified by measuring the network position of artifacts based on network analysis indices of topological centrality. The results showed that authors with specialized contribution experience in one domain predominantly contributed to central pivotal articles within that domain. Authors with generalized contribution experience in two domains predominantly contributed to boundary-crossing pivotal articles between the knowledge domains. Moreover, article experience (i.e., the number of articles in both domains an author had contributed to) was positively related to the contribution to both types of pivotal articles, regardless of whether an author had specialized or generalized domain experience. We discuss the implications of our findings for future studies in the field of CSCL.