Applying Learning Science Research to the
Design and Use of Educational Technology for
Promoting Learning about Complex Systems
Faculty of Education and Social Work,
A complex system is a system composed of interconnected parts that as a whole exhibit one or more properties (behavior among the possible properties) not obvious from the properties of the individual parts (Weaver, 1948). Complex systems exist in various subjects (e.g. biology, physics and society) within education settings and they are believed to be difficult for learners to learn. Many learning scientists have been studying the difficulties involved in understanding concepts within complex systems as emergence and decentralized control and in helping learners grow their understandings of these difficult concepts (Kolodner, 2006). Journal of the Learning Sciences, one of the official journals of the International Society of the Learning Sciences (ISLS), published a strand of articles on promoting learning about complex systems in the first volume in 2006. This paper will study the guidelines that three articles from the strand provide for the design and use of educational technology in relation to complex systems, and then identify, discuss, compare, and critically evaluate their methodological approaches, quality and applicability.
Guidelines from the Learning Sciences Community
Jacobson & Wilensky (2006) argue that the field of the learning sciences now has a major opportunity to help address the widening gap between current best understandings and analytical tools in the physical and social sciences and the working knowledge of professionals, policymakers, and citizens. They asserted that the learning sciences have been well positioned to contribute in this area and then they proposed five principles for creating environments and tools to help students learn scientific ideas about complex physical and social systems, which are (pp. 20-24):
(1) Experiencing complex systems phenomena
(2) Making the complex systems conceptual framework explicit
(3) Encouraging collaboration, discussion, and reflection
(4) Constructing theories, models, and experiments
(5) Learning trajectories for deep understandings and explorations
Different from Jacobson & Wilensky (2006), Lesh (2006) does not view current learning science theories as being sufficient to provide answers to most questions about the nature of the conceptual systems. In contrast, he believes that the most exciting point about learning science investigations of complex systems is precisely that such research is likely to require a variety of significant paradigm shifts beyond current ways of thinking. Furthermore, he believes that these paradigm shifts should have implications for learning and problem solving related to a wide range of constructs and situations where relationships to systemic understandings are far less obvious than in the case of complex systems. To support these claims, he suggests distinguishing among three kinds of complex systems (p. 46):
(1) “Real life” systems (or simulations of such systems) that occur (or are created) in everyday situations
(2) Conceptual systems that humans develop to design, model, or make sense of the preceding “real life” systems
(3) Models that researchers develop to describe and explain students’ modeling abilities
In other words, layers of complexity of complex systems are different. Systems such as double pendulum can be referred to as being “simply complex” to contrast them with the kind of “deeply complex” systems that abound in “real life” systems where the “agents” within the system are often living organisms or ecosystems that are not characterized by simple, linearly combined, or concatenated hardware- or software-based rules. Therefore, to have adequate explanatory power, the design of models of complex systems should be based on at least three assumptions and principles (p. 49):
(1) Useful models of complex systems’ most important properties cannot be derived from a list of simple functional rules.
(2) Knowledge about useful models of complex systems tends to be both situated and distributed.
(3) Useful models of complex systems meaningfully capture and illuminate some properties of the world.
Hmelo-Silver & Azevedo (2006) point out that one of the major issues affecting students’ ability to learn about complex systems is their cognitive, metacognitive, and self-regulatory processes. Understanding and reasoning about complex systems places a huge burden on working memory resources and is often counterintuitive. So they believe that studying the use of metacognitive processes in understanding complex systems is critical to understanding how we can facilitate learning about complex systems, as learners must engage in monitoring multiple activities during such learning—their emerging understanding, the aspects of their learning context and also their conceptual growth.
In addition, though agree with Jacobson and Wilensky’s (2006) claim about the importance of experience with complex systems, Hmelo-Silver & Azevedo (2006) argue that discovery alone is not sufficient and computer-based learning environment should provide students with embedded scaffolds to guide their exploration and experience, but should not provide all students with all students the same scaffolding from embedded scaffolds.
In terms of the use of educational technology, they notice that, although educators have become inundated with computer-based learning environments that may have the potential to facilitate students’ learning about complex systems, unfortunately teachers are not being trained how to support students’ learning with such technology-based environments. It is especially important for teachers to have the skills needed to support learning of complex systems from simulation and modeling software.
Among the three articles, Jacobson & Wilensky (2006) and Hmelo-Silver & Azevedo (2006) discussed issues of methodological approaches regarding the research on the guidelines and principles that they proposed. Both the discussions aim to bridge the gap between the studies on complex systems in multiple disciplines and the learning sciences research through the use of data from complex systems in real world for educational purposes. However, they focused on different aspects of the use and the data.
Jacobson & Wilensky (2006) emphasize the way that we use the data and the validity of the data, and claim that the gap can be bridged by research on the use of conceptual and methodological disciplinary toolkit such as computational modeling of systems of learning and education where validated data from real world have been embedded. Such computational modeling approach allows dramatically enhanced capabilities to investigate complex and dynamical systems and has been widely implemented in scientific practice. Once researchers have demonstrated a valid model for a particular system compared to available data, it is then possible to run “computational experiments” in which what-if scenarios about the behavior of the system may be explored to understand a system under different conditions. And such model can be developed for educational purposes under the design guidelines and principles of educational technology.
Differently, Hmelo-Silver & Azevedo (2006) emphasize the theoretical and empirical base synthesized by studying the data collected as well as the variety of the types and sources of data. They argue that we need to amalgamate diverging theoretical frameworks from multidisciplinary-approach-based research in complex systems and also need to collect multiple data source, such as those from laboratory and classroom experimentation, and use mixed methods to triangulate between qualitative data and quantitative data. They believe that the amalgamated framework and the different types of data are necessary for us to analyze and understand the complexities in learning about complex systems, which may be beneficial for corresponding design and use of educational technology.
It is evident that the two articles are promoting different directions for bridging the gap. While Jacobson & Wilensky (2006) base their exploration on complex systems perspectives and consider how to introduce computation modeling into learning sciences research, Hmelo-Silver & Azevedo (2006) start the research from the complexity of learning per se and promote educational-setting-based studies in relation to the field of complex systems. As Hmelo-Silver & Azevedo (2006) have noticed that the learning sciences is at an early stage of understanding how people think and learn about complex systems, Jacobson & Wilensky’s (2006) methodological approach—from complex systems perspectives to learning sciences—seems relatively more easy and efficient because computational modeling has been somewhat more mature in science practice and it can be immediately applied in educational settings once related guidelines from learning sciences research have been integrated into the process of its design and use.
Quality and Applicability
Though all of the three articles provide educational technology designers and users with the guidelines from different perspectives, they are still conveying the same information from the learning sciences community—it is not easy to design and use educational technology for promoting learning about complex systems, and they acknowledge that further research is needed as many current problems have not been solved. These problems can be sorted into three catalogues—cognition, computational tools, and teachers’ instruction.
For cognition, the three articles have addressed how difficult for students to change the way that they have been used to think, but only they have gone through the process of conceptual change can they better understand complex systems. Jacobson & Wilensky (2006) encourage further research about how to provide students with opportunities to experience complex systems and how to make complex systems conceptual framework explicit, which indicates that they intend to solve the problem out of the learners. Lesh (2006) criticizes the mechanistic information processing metaphors and claims that educational technology design should not reduce complex systems into a list of simple functional rules but does not claim what should be done to solve the problem. Hmelo-Silver & Azevedo (2006), focusing the cognitive challenge per se, suggest studying the use of metacognitive processes in understanding complex systems, which is critical to understand how to facilitate learning about complex systems.
As for computational tools, all of the three articles agree that we should use it to assist the learning about complex systems, but what the authors are worrying about is that computational tools may not provide students with the true representation of complex systems in real world and this may lead to misunderstanding and/or difficulty of understanding of real complex systems. Jacobson & Wilensky (2006), on the one hand, suggest introducing computational modeling approach from science practice domain to educational settings, and on the other hand turn the research focus to explore students’ learning during constructing and revising their own models. These two approaches can avoid the embarrassment in pursuing “real complex systems” in the educational technology context and are also much applicable. Lesh (2006) suggests that we differentiate three kinds of models of different degrees of complexity. Knowledge about the “simply complex” ones may be well distributed by computational tools while the “deeply complex” ones may not. However, the assumptions and principles that he proposes about useful models of complex systems tend to be hardly in relation to the solution of the “reality” issue of computational modeling. Hmelo-Silver & Azevedo’s (2006) solution is similar to Jacobson & Wilensky’s (2006) avoiding the embarrassment. But their focus is on the design of computational tools’ content and functions. They argue that computational tools should provide students with different kinds of embedded scaffolds to support their various learning needs. Nevertheless, they are still not sure how to realize the claim and then they argue that more research is necessary to understand when, how, and why to scaffold learning about complex systems in the context of computer-based learning environment.
As for teachers’ instruction, all of the three articles have discussed the challenge for teachers to teach about complex systems, especially to teach those “non-everyday” or impossible-to-directly-experience phenomena. Jacobson & Wilensky (2006) claim that teachers should organise collaboration, discussion and reflection in learning environments in which students come to experience and to construct their understandings about complex systems context. In addition, Jacobson & Wilensky (2006) also encourage investigating whether the complex-systems-knowledge-and-methodologies integrated curriculum fosters learning trajectories for deep understanding and explorations that students can apply not only in the domain of that curriculum but also other areas. But they have not provided implementation details though such claim may be still applicable but difficult for many teachers. Both Lesh (2006) and Hmelo-Silver & Azevedo (2006) focus on teachers’ preparation before class rather than the instructional activities in class and the investigation after class. While Hmelo-Silver & Azevedo (2006) still emphasize preparing teachers to be skillful of supporting learning about complex systems in class, Lesh (2006) identifies the importance for teachers to model students’ modeling abilities which may result in teachers’ deep understanding of how students’ learn about complex systems. And he also argues that researchers should help develop useful models to describe and explain students’ modeling abilities under three assumptions and principles that he proposed. Among the three articles, Lesh’s (2006) solution tends to be the most radical. Once the work of modeling students’ modeling abilities has been done by further research, it would be much easier for teachers to process their teaching about complex systems. Therefore, though three articles’ claims about the teachers’ instruction issue have not been followed by detailed guidelines for the application, those proposed by Lesh (2006) are relatively more reasonable comparing to the others.
There are strengths and weaknesses in the three articles regarding to solving the problems in relation to cognition, computational tools and teachers’ instruction. While Hmelo-Silver & Azevedo’s (2006) and Jacobson & Wilensky’s (2006) solutions are respectively the most applicable and of highest quality in terms of dealing with the issues about cognition and computational tools, the applicability of Lesh’s (2006) suggestion for teachers’ instruction issue still need to be examined in the future, but it is probably the best way to prepare teachers for the teaching about complex systems.
This paper has summarized the guidelines that three articles from the Complex Systems Strand of Journal of the Learning Sciences provide for the design and use of educational technologies for promoting learning about complex systems, and discuss, compare and evaluate their methodological issues, quality and applicability. All of three articles provide us with valuable discussion about technology-based learning about complex systems, but their emphases and scopes of the applicability are different. Therefore, we should apply the most appropriate guidelines for the corresponding issues—bridging the gap between complex systems research and learning sciences research, students’ cognition, computational tools and teachers’ instruction when we design and use educational technology for promoting learning about complex system.
Hmelo-Silver, C. E., & Azevedo, R. (2006). Understanding Complex Systems: Some Core Challenges. Journal of the Learning Sciences, 15(1), 53-61.
Jacobson, M. J., & Wilensky, U. (2006). Complex Systems in Education: Scientific and Educational Importance and Implications for the Learning Sciences. Journal of the Learning Sciences, 15(1), 11-34.
Kolodner, J. L. (2006). A Note From the Editor. Journal of the Learning Sciences, 15(1), 1-2.
Lesh, R. (2006). Modeling Students Modeling Abilities: The Teaching and Learning of Complex Systems in Education. Journal of the Learning Sciences, 15(1), 45-52.
Weaver, W. (1948). Science and Complexity. American Scientist, 36, 536.