Effects of Problem Solving and Example-Based Learning on Cognitive Load and Performance Author:Yi-Hung Huang
Research Article
The 12-year Basic Education Curriculum guidelines emphasize the importance of discovery activities in the curriculum.Therefore, methods of implementing discovery-based learning in the classroom and achieving effective teaching goals have become a major topic for current teachers. Alfieri et al. (2011) conducted a meta-analysis of 164 studies and reported that allowing students to only perform pure discovery activities may be detrimental to learning; however, assisted discovery learning, including feedback, worked examples, and self-explanation prompts, can benefit learning. Studies on assisted discovery learning, have provided empirical results regarding the integration of problem solving and example learning. These studies have indicated that learners who cooperate in solving problems before receiving direct teaching or seeing examples perform better than those who first receive direct teaching or examples and then solve problems. However, the findings of these studies only demonstrate that solving problems cooperatively and then studying examples is a more favorable approach than studying the examples individually and then practicing solving the problems. With inconsistent experimental conditions (i.e., cooperative vs. individual), distinguishing the effects of the variables of the timing of problem solving and the type of example learning on teaching effectiveness is impossible. Sweller and Paas (2017) also indicated that most of these studies did not control variables appropriately and that they changed multiple variables simultaneously. Therefore, the experimental designs of these studies were fundamentally flawed and could not isolate the effects of specific variables. Although some studies on the integration of problem solving and example learning have analyzed the learning process from the perspective of cognitive load theory, this is only a theoretical discussion. Studies have not conducted actual measurements of the cognitive load perceived by the learner during the learning process (Kapur & Bielaczyc, 2012; Loibl & Rummel, 2014). In addition, for studies in which the experimental participants were elementary school students, only one item was used to evaluate cognitive load, such as “How much effort did you use in the process of learning just now?”; in general, this cognitive load could not be accurately measured (Huang & Shie, 2016 ; Wong et al., 2012). Furthermore, if the intrinsic and extraneous cognitive loads perceived by the learner during learning cannot be known, cognitive load theory cannot be used to effectively analyze the learner’s cognition and information processing during learning (Leppink et al., 2013). Therefore, this study excluded the effect of cooperative learning and allowed learners to learn materials individually to avoid confounding variables. The “basic quantity and comparative quantity” unit of mathematics was used for the experiments in this study. The researcher conducted an experiment to study the effects of two factors on learning outcomes and perceived cognitive load: Problem-solving timing (problem solving first or reading an example first) and the type of example (comparative example or problem-solving practice). The multidirectional scale developed by Leppink and van den Heuvel (2015) was used to measure intrinsic and extraneous cognitive loads perceived during learning, and an objective reference value was used for the learning outcome to observe germane cognitive load. Therefore, on the basis of cognitive load theory and the relevant literature, the research questions verified in this study were as follows: 1. Does the timing of problem solving and the type of worked-out example affect the intrinsic, extraneous, and overall cognitive load when students learn to draw line segments, operate line segments, and solve reference and comparison quantity problems? 2. Does the timing of problem solving and the type of worked-out example affect the test scores when students learn to draw line segments, operate line segments, and solve the reference and the comparison quantity problems?
The researcher first used two factors (timing of problem solving and type of worked-out example) to design an experiment involving three exercises: drawing line segments, operating line segments, and solving a reference and comparison quantity problem. Second, 105 sixth-grade children were randomly assigned to four groups for the experiment: reading examples first then comparing examples (group 1), solving problems first then comparing examples (group 2), reading examples first then practicing problem solving (group 3), and solving problems first then practicing problem solving (group 4). After each exercise was completed, the cognitive load scale was used to measure the students’ perceived intrinsic load and extraneous load. Finally, a learning test was used to evaluate the students’ learning performance in drawing line segments, operating line segments, and solving the reference and the comparison quantity problems.
The answers to the first three questions of the cognitive load scale indicated the perceived intrinsic load, and the answers to the last three questions indicated the perceived extraneous load. Factor analysis with the timing of problem solving and type of worked-out example as independent variables and the test scores and cognitive load indicators as the dependent variables was performed. The two-factor multivariate analysis revealed no significant interaction effect between the timing of problem solving and the type of example on the students’ perceived cognitive load or the three test scores. The analysis also indicated that the perceived intrinsic load of the students who solved problems first was significantly higher than that of the students who read examples first when solving the reference and the comparison quantity problems. The overall perceived cognitive load of the students who solved problems first was also significantly higher than that of those who read examples first when operating line segments and solving the reference and comparison quantity problem. No significant difference was observed between the comparative example condition and the practicing problem solving condition in the students’ perceived intrinsic, extraneous, or overall cognitive load No significant difference in test scores was observed between the different problem-solving timing conditions in the three tests; however, the test scores of the students who had seen a comparative example were significantly higher than those of the students who practiced problem solving when drawing line segments.
Four sets of materials were designed to study the effects of problem-solving timing and worked-out examples on learning effectiveness and perceived cognitive load. The researcher used the multidimensional cognitive load scale to successfully measure the students’ perceived intrinsic load and extraneous load. When the experimental conditions were controlled for and the cooperative learning variable was excluded, the experimental results were determined to be consistent with those of previous research regarding worked examples (Kalyuga et al., 2003; Renkl & Atkinson, 2003). The learning performance of group 2 was not higher than that of group 1. Different learning examples may be suitable for learning different types of knowledge. Therefore, on the basis of the findings of this study, the researcher proposes the following strategies for educators. First, because of the nature of problem-solving activities, learners may perceive a higher cognitive load; therefore, learning materials suitable for the level of learners should be carefully designed when implementing a strategy that involves solving problems. Second, the comparison of examples is suitable for learning conceptual knowledge, and practicing of problem solving is suitable for learning procedural knowledge. Finally, when designing teaching materials for comparative examples, educators should add incorrect examples to the materials. The correct and incorrect examples should be compared, and the learner must be asked to explain why the incorrect examples are wrong; doing so can repair the learners’ knowledge and enable them develop the correct concept.
Effects of Problem Solving and Example-Based Learning on Cognitive Load and Performance
下載
