Construction and interference in learning from multiple representation

This paper presents an integrated view of learning from verbal and pictorial representations. Learning from these representations is considered as a task oriented process of constructing multiple mental representations. Construction of these representations includes information selection and information organisation, parsing of symbol structures, mapping of analog structures as well as model construction and model inspection. Based on this theoretical view an experiment was conducted to analyse the effects of different kinds of multiple external representations on the structure of mental models. Sixty university students were randomly assigned to one of the three experimental conditions. The text-only group learned the subject matter with a hypertext, whereas the other two groups learned the subject matter with a hypermedium including this hypertext and different kinds of graphics. The findings indicate that the structure of graphics affects the structure of the mental model. They also indicate that presenting graphics is not always beneficial for the acquisition of knowledge. Whereas task-appropriate graphics may support learning, task-inappropriate graphics may interfere with mental model construction.     

The planning illusion: does active planning of a learning route support learning as well as learners think it does?

Is actively planning one’s learning route through a learning domain beneficial for learning? Moreover, can learners accurately judge the extent to which planning has been beneficial for them? This study examined the effects of active planning on learning. Participants received a tool in which they created a learning route themselves before accessing learning material and, for comparison, also worked with a tool in which the route was planned automatically. Eighty-three participants participated in learning sessions with both tools over two topics in statistics. Both tools were found to influence the learning process. However, results indicate inconsistency between participants’ judgement of the relative effectiveness of the two types of tools and their actual learning outcomes. Although participants thought they had learned more when they actively created a plan themselves, knowledge tests showed no differences on performance between actively planning a learning route or receiving it ready-made.     

The effect of a human agent’s external regulation upon college students’ hypermedia learning

In this study we examined the effectiveness of self-regulated learning (SRL) and externally regulated learning (ERL) on college students’ learning about a science topic with hypermedia during a 40-min session. A total of 82 college students with little knowledge of the topic were randomly assigned either to the SRL or ERL condition. Students in the SRL condition regulated their own learning, while students in the ERL condition had access to a human tutor who facilitated their self-regulated learning. We converged product (pretest–posttest declarative knowledge and qualitative shifts in participants’ mental models) with process (think-aloud) data to examine the effectiveness of SRL versus ERL. Analysis of the declarative knowledge measures showed that the ERL condition group mean was statistically significantly higher than the group mean for the SRL condition on the labeling and flow diagram tasks. There were no statistically significant differences between groups on the matching task, but both groups showed statistically significant increases in performance. Further analyses showed that the odds of being in a higher mental model posttest group were decreased by 65% for the SRL group as compared to the ERL group. In terms of SRL behavior, participants in the SRL condition engaged in more use of selecting new information sources, re-reading, summarizing, free searching, and enacting control over the context of their learning. In comparison, the ERL participants engaged in more activation of prior knowledge, utilization of feeling of knowing and judgment of learning, monitoring their progress toward goals, drawing, hypothesizing, coordination of information sources, and expressing task difficulty. An earlier version of this paper was presented at the international conference of Artificial Intelligence in Education (AI-Ed 2007), Los Angeles, LA (July, 2007).     

Domain expertise and the effectiveness of dynamic simulator interfaces in the acquisition of procedural motor skills

Previous research into the effectiveness of dynamic versus static instructional design paradigms has reported divergent findings. Dynamic instructions have been shown to be more effective in teaching novel procedural skills. In contrast, the apparent benefit of dynamic over static instructions has been attributed in other studies to the cognitive capabilities and previous knowledge of the learner. Can the benefit of dynamic instruction persist in learners with domain expertise on learning novel tasks? In this paper, we report the result of an experiment that shows that irrespective of the learner's previous knowledge, dynamic instructions retain a significant effectiveness over statics for teaching intra‐domain novel task performance. Twenty‐four participants with domain expertise were divided into three independent groups to perform a procedural motor task following treatment with different training interfaces. After controlling for spatial abilities and excluding previous specific‐to‐task knowledge, we observe that participants that trained with interfaces containing dynamic content recorded better task performance measures than others using non‐dynamic interfaces. This suggest that within the context of motor skill acquisition, dynamic instructional interfaces can yield significant increases to post‐learning task performance measures, which is independent of the learner's cognitive capabilities or previous knowledge.     

Task-dependent construction of mental models as a basis for conceptual change

Learning can be seen as a task-oriented process which often requires the reorganization of existing knowledge, usually referred to as conceptual change. This paper describes a theoretical framework for the analysis of conceptual change that considers conceptual knowledge as a generative cognitive tool for the creation of more specific mental representations — propositional symbolic structures and analog mental models. According to this view, conceptual change is based on a task-oriented interaction between these different kinds of mental representations. The assumption is made that it is possible to foster conceptual change by presenting to students well-defined tasks that stimulate the construction of elaborated mental models as well as an intensive interaction between these models and the corresponding propositional representations. In order to test this assumption an empirical study was conducted, in which subjects had to express their prior knowledge about a complex subject matter from the field of geography (time differences on the earth), which contained various conceptual deficits. The subjects were then randomly assigned to different groups who received the same learning material but had to solve different learning tasks requiring differently structured mental models. Afterwards, the subjects were asked to express their knowledge about the subject matter again and were tested for understanding with a comprehension test. The results support the view that a task-oriented interaction between propositional structures and mental models can help learners to evaluate the consistency of their conceptual knowledge. Accordingly, conceptual deficits result in the formation of mental models with an inadequate structure. Such deficits can be detected if the respective model is used in a sufficiently variable way, whereas they can remain unnoticed if it is used in a limited manner.     

Effectiveness of holistic mental model confrontation in driving conceptual change

Prior research on conceptual change has identified multiple kinds of misconceptions at different levels of representational complexity including false beliefs, flawed mental models, and incorrect ontological categories. We hypothesized that conceptual change of a mental model requires change in the system of relations between the features of the prior model. To test this hypothesis, we compared instruction aimed at revising knowledge at the mental model level called holistic confrontation - in which the learner compares and contrasts a diagram of his or her flawed mental model to an expert model - to instruction aimed at revising knowledge at the false belief level - in which the learner is prompted to self-explain the expert model alone. We found evidence that participants who engaged in holistic confrontation were more likely to acquire a correct mental model, and a deeper understanding of the systems of relations in the model than those who were prompted to self-explain the expert model. The results are discussed in terms of their implications for science instruction.     

Evaluating mental models in mathematics: a comparison of methods

Cognitive scientists investigate mental models (how humans organize and structure knowledge in their minds) so as to understand human understanding of and interactions with the world. Cognitive and mental model research is concerned with internal conceptual systems that are not easily or directly observable. The goal of this research was to investigate the use of Evaluation of Mental Models (EMM) to assess the mental models of individuals and groups in solving complex problems and to compare novices and experts models as bases for providing feedback to learners. This study tested a qualified web-based assessment tool kit, Highly Interactive Model-based Assessment Tools and Technologies (HIMATT), in an as yet untested domain—mathematics. In this study, university students and their mathematics instructors used two tools in HIMATT, Dynamic Evaluation of Enhanced Problem Solving (DEEP) and Text-Model Inspection Trace of Concepts and Relations (T-MITOCAR). The research questions include: Do novice participants exhibit common patterns of thoughts when they conceptualize complex mathematical problems? Do novices conceptualize complex mathematical problems differently from experts? What differences in DEEP and T-MITOCAR patterns and responses exist according to the measures of HIMATT? Findings suggest that EMM and HIMATT could effectively support formative assessment in a complex mathematical domain. Finally, this study confirms a common assumption of cognitive scientists that the tool being used could affect the tool user’s understanding of the problem being solved. In this case, while DEEP and T-MITOCAR led to somewhat different expert models, both tools prove useful in support of formative assessment.     

The reverse modality effect: Examining student learning from interactive computer‐based instruction

The purpose of this study was to explore the effects of modality on learning from multimedia instruction. This study utilized a factorial between‐subject design to examine the effects of modality on student learning outcomes, study patterns and mental effort. An interactive computer‐presented diagram was developed to teach the places of articulation in human speech. A total of 151 undergraduate students at a large southwestern university in USA participated in this study. Participants were randomly assigned to one of two modality conditions (ie, written text and spoken text). Data were obtained through surveys, student logs and knowledge tests. Findings revealed a reverse modality effect, wherein participants who studied with written text outperformed those who studied with spoken text.     

Effects of planning on task load,knowledge, and tool preference: a comparison of two tools

Self-regulated learners are expected to plan their own learning. Because planning is a complex task, it is not self-evident that all learners can perform this task successfully. In this study, we examined the effects of two planning support tools on the quality of created plans, planning behavior, task load, and acquired knowledge. Sixty-five participants each worked with two versions of a planning tool. In one version, learning plans were actively constructed by the learners themselves; the other version provided learners with an adaptable computer-generated plan. The results indicated that the quality of learner-created plans was lower than computer-generated plans. Furthermore, participants reported a higher task load when they constructed the plans by themselves. However, participants gained more structural knowledge about the learning domain when they actively created plans. There was not an apparent preference for one of the tools if participants were to create a plan for someone else. However, if they were to use the plan for their own learning, participants preferred to actively create their own plans.     

Learning through intermediate problems in creating cognitive models

Cognitive modelling is one of the representative research methods in cognitive science. It is believed that creating cognitive models promotes learners’ meta-cognitive activities such as self-monitoring and reflecting on their own cognitive processing. Preceding studies have confirmed that such meta-cognitive activities actually promote learning effects. However, there are some difficulties in bringing about learning by creating cognitive models in an educational context. To overcome the difficulties, we propose an innovative learning design, ‘learning through intermediate problems’ and also developed a web-based production system called DoCoPro that can be used anywhere and anytime in an environment connected to the Internet. We performed three introductory cognitive science classes in which the participants learned cognitive modelling and constructed running computer models using our system. In the first and second classes, the participants were required to construct production system models that solve pulley problems. They also posed their original pulley problems that their own models were subsequently able to solve. These generated problems were distributed to the other members. The participants were able to find incompleteness in their cognitive models, revise them to remove the incompleteness, and improve their models while solving the given problems. The participants, by successfully creating sophisticated models, acquired a deeper knowledge of the learning domain. The class practices confirmed the utility of ‘learning through intermediate problems’ when constructing an educational environment for learning creating cognitive models. In the third class, the participants constructed cognitive models solving addition and subtraction problems using DoCoPro. The cognitive processing underlying such problem solving is automated, therefore it may be difficult to verbalize and externalize such cognitive processes. The post-questionnaire showed evidence that the participants actually performed meta-cognitive activities while monitoring their own internal information processing.     

Rapid Assessment of Learners’ Proficiency: A cognitive load approach

To tailor instruction to levels of learner proficiency in a domain in adaptive learning environments, rapid on‐line methods of assessing learner knowledge and skills are required. Cognitive studies indicate that major factors in expert performance are organised structures in the person’s knowledge base. Measuring different levels of acquisition of such knowledge structures should be a major purpose of cognitive diagnostic assessment. Traditional tests may not provide reliable evidence for diagnostic purposes and are not suitable for dynamic learning environments. This paper discusses an alternative diagnostic approach based on monitoring immediate traces of students’ knowledge structures in working memory. Results of an experimental study applying this method to assessment of user reading skills are described. The study demonstrated a sufficiently high level of correlation (.66) between obtained measures and traditional test scores, and a substantial increase in reliability, with testing time reduced by a factor of 3.8.     

Evaluating the effects of analogy enriched text on the learning of science: The importance of learning indexes

Research has shown that differences in the prior knowledge of the participants and in the learning indexes adopted can explain why some studies show positive learning effects of analogy enriched text while others do not. In the present studies, these two factors were combined into one through the construction of a learning index that measured incremental positive changes in the participants' prior knowledge after reading an analogy enriched or no analogy text. A second learning index was also used to evaluate whether the participants created well-formed conceptual models after reading the science text. These learning indexes were used in two studies in which the effects of analogy enriched versus no analogy text were compared on the learning of the scientific explanations of the day/night cycle and of the seasons. The participants were 3rd and 5th graders in the first study and 6th graders and college students in the other. Although only few of the participants learned the correct scientific explanation, those who read the analogy enriched text produced more incremental positive changes in their pretest explanations at posttest and delayed test and created more well-formed conceptual models close to the scientific one than those who read the no analogy text. They also recalled more information and created fewer invalid inferences in their recalls. The results indicate that analogies can be used without reservation to facilitate the learning of science and have broader implications about how to evaluate the learning of science in general.     

Evaluation of the effectiveness of 3D vascular stereoscopic models in anatomy instruction for first year medical students

The head and neck region is one of the most complex areas featured in the medical gross anatomy curriculum. The effectiveness of using three‐dimensional (3D) models to teach anatomy is a topic of much discussion in medical education research. However, the use of 3D stereoscopic models of the head and neck circulation in anatomy education has not been previously studied in detail. This study investigated whether 3D stereoscopic models created from computed tomographic angiography (CTA) data were efficacious teaching tools for the head and neck vascular anatomy. The test subjects were first year medical students at the University of Mississippi Medical Center. The assessment tools included: anatomy knowledge tests (prelearning session knowledge test and postlearning session knowledge test), mental rotation tests (spatial ability; presession MRT and postsession MRT), and a satisfaction survey. Results were analyzed using a Wilcoxon rank‐sum test and linear regression analysis. A total of 39 first year medical students participated in the study. The results indicated that all students who were exposed to the stereoscopic 3D vascular models in 3D learning sessions increased their ability to correctly identify the head and neck vascular anatomy. Most importantly, for students with low‐spatial ability, 3D learning sessions improved postsession knowledge scores to a level comparable to that demonstrated by students with high‐spatial ability indicating that the use of 3D stereoscopic models may be particularly valuable to these students with low‐spatial ability. Anat Sci Educ 10: 34–45. © 2016 American Association of Anatomists.     

An Exploration of Secondary Students’ Mental States When Learning About Acids and Bases

This study explored factors of students’ mental states, including emotion, intention, internal mental representation, and external mental representation, which can affect their learning performance. In evaluating students’ mental states during the science learning process and the relationship between mental states and learning achievement, valid, reliable, and scalable measures of students’ mental states and learning achievement are needed. This paper presents the development of the Mental State Conceptual Learning Inventory (MSCLI) to identify students’ mental states before and after learning about acids and bases. This instrument is time efficient and convenient and can be administered to large student samples so that teachers and researchers can gain profound insights into their students’ learning of acids and bases in science class. The results of this study indicate that students’ mental states are highly correlated with their achievement. As a whole, low-achieving students tended to have negative emotions and low intentions, were not good at internal visualization, and were unable to interpret graphics and draw pictures. In contrast, high-achieving students had positive emotions and intentions when learning life-related topics about acids and bases, and were good at internal visualization and drawing and interpreting graphics.     

Adaptive Human Scaffolding Facilitates Adolescents’ Self-regulated Learning with Hypermedia

This study examines the effectiveness of three scaffolding conditions on adolescents’ learning about the circulatory system with a hypermedia learning environment. One hundred and eleven adolescents (n = 111) were randomly assigned to one of three scaffolding conditions (adaptive scaffolding (AS), fixed scaffolding (FS), or no scaffolding (NS)) and were trained to use a hypermedia environment to learn about the circulatory system. Pretest and posttest data were collected to measure the qualitative changes in students’ mental models of the topic and quantitative changes in their declarative knowledge. Verbal protocols were collected during the 40-minute learning task to examine how each condition affected the way in which students regulated their learning. Findings revealed that learners in both the AS and NS conditions gained significantly more declarative knowledge than did those in the FS condition. Also, the AS condition was associated with shift in learners’ mental models significantly more than the other conditions. Associated with these significant shifts in their mental models, learners in the AS condition regulated their learning by planning and activating prior knowledge, monitoring their cognitive activities and their progress toward learning goals, using several effective strategies, and engaging in adaptive help-seeking. By contrast, those in the NS condition used fewer effective strategies, while those in the FS regulated their learning by using several regulatory processes which seemed to impede their learning. Implications for the design of scaffolds for fostering students’ self-regulated learning with hypermedia are presented.     

Knowledge presented in concept maps: correlations with conventional cognitive knowledge tests

Our study focuses on the correlation of concept map (CMap) structures and learning success tested with short answer tests, taking into particular account the complexity of the subject matter. Novice sixth grade students created CMaps about two subject matters of varying difficulty. The correlation of the complexity of CMaps with the post-test was small but highly significant in both subject matters. The complexity of the CMaps correlated with the long-term knowledge in the difficult subject matter but not in the context of the easy one. Furthermore, the high number of technical errors makes it close to impossible to estimate students’ knowledge. In summary, CMaps do not provide an adequate alternative to conventional short answer knowledge tests, but together with them they may offer a better comprehension of a student’s knowledge structure and aid in the preparation of further instruction tailored to individual needs.     

Knowledge based systems for teaching and learning maths

The main task of an educational system is to generate instructional situations which induce students to learn knowledge and problem solving abilities as applied to a cognitive domain.
To this end, an instructional dialogue must be seen as a process aimed at making student and teacher learn from each other: the student has to learn the subject matter from the teacher, while the teacher has in turn to learn from the student how to regulate instructional interaction. The process relies on different kinds of expert knowledge: experience of the domain, about the actual student, and of teaching methods and objectives. This means that an instructional dialogue cannot be realised without an explicit representation of all the kinds of knowledge involved.
Research has established that Knowledge Based Educational Systems (KBESs) can be the keystone in building effective learning tools, but because of the difficulty in realising a system of this kind, most existing systems are prototypes and are intended only as workbenches for the computational analysis of educational processes.
This paper analyses the main problems which underlie the realisation of such systems, with reference to research into knowledge based systems intended for use with teaching/learning mathematics.     

Process-oriented instruction in learning and thinking strategies

Process-oriented instruction is defined as instruction aimed at teaching thinking strategies and domain-specific knowledge in coherence. This new conception of instruction is derived from psychological research on the way students learn and on the interplay between self-regulation and external regulation of learning. In the research reported here the learning effects of a process-oriented instructional program for university students were empirically studied. The instructional program consisted of a diagnostic learning style instrument, a learning guide and tutorials. The results showed that the majority of students reported significant general, knowledge, insight and application learning effects. The program effects were typified more by integrating and making usable metacognitive knowledge already present, than by increasing knowledge about new subjects. Evidence for transfer effects was obtained because participants in the program scored better than non-participants on two exams of another course. The learning effects were higher than the effects of an preliminary version of the program implemented with students from an open university. These results support the importance of the process-oriented instructional model. The linking of a thorough diagnosis of personal learning styles to individually tailored instructional measures, turned out to be a powerful way to activate students to reflect on their learning and to develop their mental models of learning.     

Research in Science Education at the University of Glasgow

The purpose of this study was to examine how prior knowledge of cellular transport influenced how high school students in the USA viewed and interpreted graphic representations of this topic. The participants were Advanced Placement Biology students (n = 65); each participant had previously taken a biology course in high school. After assessing prior knowledge using the Diffusion and Osmosis Diagnostic Test, two graphical representations of cellular transport processes were selected for analysis. Three different methods of data collection—eye tracking, interviews, and questionnaires—were used to investigate differences in perceived salient features of the graphics, interpretations of the graphics, and processing difficulty experienced while attending to and interpreting the graphics. The results from the eye tracking data, interviews, and instructional representation questionnaires were triangulated and revealed differences in how high and low prior knowledge students attended to and interpreted particle differences, concentration gradient, the role of adenosine triphosphate , endocytosis and exocytosis, and text labels and captions. Without adequate domain knowledge, low prior knowledge students focused on the surface features of the graphics (ex. differences in particle colour) to build an understanding of the concepts represented. On the other hand, with more abundant and better‐organised domain knowledge, high prior knowledge students were more likely to attend to the thematically relevant content in the graphics, which enhanced their understanding. The findings of this study offer a more complete understanding of how differentially prepared learners view and interpret graphics and have the potential to inform instructional design.