Academic journal article Educational Technology & Society

Error-Based Simulation for Error-Awareness in Learning Mechanics: An Evaluation

Academic journal article Educational Technology & Society

Error-Based Simulation for Error-Awareness in Learning Mechanics: An Evaluation

Article excerpt


Natural science explains and predicts natural phenomena. One of the most important purposes of elementary science education is to make students able to explain and predict natural phenomena with scientific concepts. It is, however, often difficult for students to connect abstract scientific concepts to concrete natural phenomena. As well, disconnection of them causes several serious misconceptions (Driver, Guesne, & Tiberghien, 1985; Osborne & Freyberg, 1985). Therefore, supporting students' comprehension of this connection is a very important issue in elementary science education.

Scientific experiment or demonstration is a popular teaching method to connect them. First, a phenomenon is shown to the students, and then it is explained with scientific concepts that are the targets of teaching. Simulation-based learning environments (SLE) have been investigated to assist such "learning from experiments" and have been found useful for the introduction or acquisition of scientific concepts (Towne, de Jong, & Spada, 1993). When a student makes an error in prediction, showing the correct simulation could be useful for correcting that error. The difference between the correct simulation generated by the SLE and the phenomenon predicted by the student makes the student aware of the error. However, students sometimes have wrong concepts for explaining correct phenomena. In such cases, phenomena generated by SLE aren't useful because the phenomena are the same as those they connected to the wrong concepts. For example, in elementary mechanics, students often answer that gravity is the only force acting on a block on a table while they predict that the block will stay at rest on the table.

Error-based simulation (EBS) addresses this problem by generating a phenomenon using students' erroneous ideas to help them become aware of errors when they know some correct phenomena connected to their wrong concepts (Hirashima, Horiguchi, Kashihara, & Toyoda, 1998). In the above example, EBS generates an unnatural phenomenon where the block sinks into the table because gravity, which isn't cancelled by any other force, causes a downward motion of the block. EBS is a generally useful method to generate counterexamples to students' misconceptions or erroneous answers, and we have developed several prototype systems not only for mechanics but also for drawing (Matsuda et al., 2003) and English composition (Kunichika, Hirashima, & Takeuchi, 2006).

EBS has an important feature that other systems don't have that can bring out the pedagogical merit in simulationbased learning. Even when a model is incalculable (e.g., because the constraints specified by the student contradict each other), our approach involves creating an EBS by relaxing some of the student's model constraint(s). In the above example, the constraint "rigid objects (i.e., the block and table) never overlap" is violated in the simulation. Thus, EBS works as a counterexample, demonstrating that "if students' ideas were correct, the fundamental constraint in the real world would be violated." Because of this feature, shown the next section in detail, one can design a learning environment with EBS that has the advantages of discovery learning (Crews, Biswas, Goldman, & Bransford, 1997; Loh, Reiser, Radinsky, Edelson, Gomez, & Marshall, 2000) and directed learning (Klahr, 2009).

So far, preliminary experiments in which small number of university students learned with the above systems and teachers evaluated the functions of these systems suggested the possibility of EBS (Kunichika et al., 2006; Matsuda et al., 2003). Additionally, a preliminary test in which small number of junior high school students learned elementary mechanics with EBS suggested that EBS promoted their conceptual understanding (Horiguchi, Imai, Toumoto, & Hirashima, 2007). Though the test was conducted in a practical situation, the result wasn't statistically validated because no control group was made for comparison. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.