Cognitive load theory and aging: effects of worked examples on training efficiency

Cognitive load theory (CLT) is aimed at developing training material that efficiently makes use of the available cognitive processing capacity and stimulates the learner's ability to use acquired knowledge and skills in new situations. It is claimed that CLT-based training formats meet the cognitive abilities of elderly learners particularly well. That is, cognitive aging brings about several declines of working memory, which impede the acquisition of complex cognitive skills. By making an optimal use of the ‘remaining’ cognitive resources, learning can be enhanced. For that purpose, CLT provides a promising range of training formats that have proven their effectiveness relative to conventional formats in young adults. This article presents an experimental study (N=54) aimed at the efficiency of worked examples as a substitute for conventional practice problems in training both elderly and young adults. According to CLT, studying worked examples is a more efficient means of training complex skills than solving conventional problems. As predicted, the results show that — with respect to the elderly — the efficiency of studying worked examples is higher than the efficiency of solving conventional problems in that less training time and cognitive load leads to a comparable level of performance.     

Process-Oriented Worked Examples: Improving Transfer Performance Through Enhanced Understanding

The research on worked examples has shown thatfor novices, studying worked examples is oftena more effective and efficient way of learningthan solving conventional problems. Thistheoretical paper argues that addingprocess-oriented information to worked examplescan further enhance transfer performance,especially for complex cognitive skills withmultiple possible solution paths.Process-oriented information refers to theprincipled (``why'') and strategic (``how'')information that experts use when solvingproblems. From a cognitive load perspective,studying the expert's ``why'' and ``how''information can be seen as constituting agermane cognitive load, which can fosterstudents' understanding of the principles of adomain and the rationale behind the selectedoperators, and their knowledge about howexperts select a strategy, respectively. Issueswith regard to the design, implementation, andassessment of effects of process-orientedworked examples are discussed, as well as thequestions they raise for future research.     

The acquisition of problem-solving skills in mathematics: How animations can aid understanding of structural problem features and solution procedures

In this paper the augmentation of worked examples with animations for teaching problem-solving skills in mathematics is advocated as an effective instructional method. First, in a cognitive task analysis different knowledge prerequisites are identified for solving mathematical word problems. Second, it is argued that so called hybrid animations would be most effective for acquiring these prerequisites, because they show the continuous transition from a concrete, but superficial problem representation to a more abstract, mathematical problem model that forms a basis for solving a problem. An experiment was conducted, where N = 32 pupils from a German high school studied either only text-based worked examples explaining different problem categories from the domain of algebra or worked examples augmented with hybrid animations. Learners with hybrid animations showed superior problem-solving performance for problems of different transfer distance relative to those in the text-only condition.     

Effect of worked examples and Cognitive Tutor training on constructing equations

Algebra students studied either static-table, static-graphics, or interactive-graphics instructional worked examples that alternated with Algebra Cognitive Tutor practice problems. A control group did not study worked examples but solved both the instructional and practice problems on the Cognitive Tutor (CT). Students in the control group requested fewer hints and made fewer errors on the CT practice problems but required more learning time on the instructional examples. There was no difference among the four groups in constructing equations on a paper-and-pencil posttest or on a delayed test that included training and transfer problems. However, students who studied worked examples with a table were best at identifying the meaning of the equation components. The concept of transfer-appropriate processing (the overlap between instructional task and assessment task) aided our interpretation of the findings. Although the CT had a short-term effect on reducing errors and hint requests on CT practice problems, the worked examples were as effective on delayed paper-and-pencil tests. The subsequent construction of a new module for the Animation Tutor (Reed and Hoffman, Animation Tutor: Mixtures. Instructional software, 2011) used both the interactive-graphics and static-table worked examples to take advantage of the complementary strengths of different representations (Ainsworth, Learn Instr 16:183–198, 2006).     

Training self-regulated learning skills with video modeling examples: Do task-selection skills transfer?

Self-assessment and task-selection skills are crucial in self-regulated learning situations in which students can choose their own tasks. Prior research suggested that training with video modeling examples, in which another person (the model) demonstrates and explains the cyclical process of problem-solving task performance, self-assessment, and task-selection, is effective for improving adolescents’ problem-solving posttest performance after self-regulated learning. In these examples, the models used a specific task-selection algorithm in which perceived mental effort and self-assessed performance scores were combined to determine the complexity and support level of the next task, selected from a task database. In the present study we aimed to replicate prior findings and to investigate whether transfer of task-selection skills would be facilitated even more by a more general, heuristic task-selection training than the task-specific algorithm. Transfer of task-selection skills was assessed by having students select a new task in another domain for a fictitious peer student. Results showed that both heuristic and algorithmic training of self-assessment and task-selection skills improved problem-solving posttest performance after a self-regulated learning phase, as well as transfer of task-selection skills. Heuristic training was not more effective for transfer than algorithmic training. These findings show that example-based self-assessment and task-selection training can be an effective and relatively easy to implement method for improving students’ self-regulated learning outcomes. Importantly, our data suggest that the effect on task-selection skills may transfer beyond the trained tasks, although future research should establish whether this also applies when trained students perform novel tasks themselves.     

Completion problems can reduce the illusions of understanding in a computer-based learning environment on genetics

Inaccurate judgments of task difficulty and invested mental effort may negatively affect how accurate students monitor their own performance. When students are not able to accurately monitor their own performance, they cannot control their learning effectively (e.g., allocate adequate mental effort and study time). Although students' judgments of task difficulty and invested mental effort are closely related to their study behaviors, it is still an open question how the accuracy of these judgments can be improved in learning from problem solving. The present study focused on the impact of three types of instructional support on the accuracy of students' judgments of difficulty and invested mental effort in relation to their performance while learning genetics in a computer-based environment. Sixty-seven university students with different prior knowledge received either incomplete worked-out examples, completion problems, or conventional problems. Results indicated that lower prior knowledge students performed better with completion problems, while higher prior knowledge students performed better with conventional problems. Incomplete worked-out examples resulted in an overestimation of performance, that is, an illusion of understanding, whereas completion and conventional problems showed neither over- nor underestimation. The findings suggest that completion problems can be used to avoid students' misjudgments of their competencies.     

Effects of problem solving after worked example study on secondary school children’s monitoring accuracy

Monitoring accuracy, measured by judgements of learning (JOLs), has generally been found to be low to moderate, with students often displaying overconfidence, and JOLs of problem solving are no exception. Recently, primary school children’s overconfidence was shown to diminish when they practised problem solving after studying worked examples. The current study aimed to extend this research by investigating whether practising problem solving after worked example study would also improve JOL accuracy in secondary education. Adolescents of 14–15 years old (N = 143) were randomly assigned to one of five conditions that differed in timing of JOLs, whether practice problems were provided, and timing of the practice problems provided: (1) worked examples – JOL, (2) worked examples – delay – JOL, (3) worked examples – practice problems – JOL, (4) worked examples – practice problems – delay – JOL or (5) worked examples – delay – practice problems – JOLs. Results showed that practice problems improved absolute accuracy of JOLs as well as regulation accuracy. No differences in final test performance were found.     

Training higher education teachers for instructional design of competency-based education: Product-oriented versus process-oriented worked examples

Teachers involved in the development of competency-based higher education (CBE) are expected to fulfil a new role of instructional designer. As a consequence, they are confronted with the problem to translate abstract new curriculum principles into concrete learning tasks. Recent studies have shown that teachers can be trained to apply an instructional systems design methodology. After this training the teachers were able to design better learning tasks for CBE in comparison with their experienced-based design efforts. In order to optimize the training, this study compares an experimental condition with process-oriented worked examples with a conventional training condition with emphasis on product-oriented worked examples. After the training, the participants—25 higher education teachers—had to apply the ISD methodology to two design problems. The quality of the resulting design materials, as rated by experts, was higher in the product-oriented worked examples condition than in the process-oriented worked examples condition. The significance of this finding for training approaches to design methodology for CBE is discussed.     

Redirecting learners' attention during training: effects on cognitive load,transfer test performance and training efficiency

Cognitive load theory provides guidelines for improving the training of complex cognitive skills and their transfer to new situations. One guideline states that extraneous cognitive load that is irrelevant to the construction of cognitive schemata should be minimised. Experiment 1 (N=26) compares completion problems, conventional problems, and a learner-controlled condition in which learners may choose between problem formats. Completion problems decrease cognitive load during training and have a zero or positive effect on transfer performance. A second guideline states that germane cognitive load that is directly relevant to schema construction should be optimised. In Experiment 2 (N=69) practice schedules of either high or low contextual interference are compared (HCI and LCI). HCI increases cognitive load during training and shows a trend towards higher transfer performance. Experiment 3 (N=87) combines both guidelines in a factorial experiment with the factors problem format (completion vs. conventional) and contextual interference (HCI vs. LCI). It is hypothesised that redirecting attention from extraneous to germane processes will improve training efficiency, i.e. positively affect the balance between cognitive load during training and transfer test performance. In support of this hypothesis, it is found that the completion-HCI group shows highest training efficiency. But transfer test performance for this group is disappointing. The results are discussed in relation to the operationalisation of HCI in combination with completion problems.     

Cognitive load and mental rotation: structuring orthographic projection for learning and problem solving

Cognitive load theory was used to generate a series of three experiments to investigate the effects of various worked example formats on learning orthographic projection. Experiments 1 and 2 investigated the benefits of presenting problems, conventional worked examples incorporating the final 2-D and 3-D representations only, and modified worked examples with several intermediate stages of rotation between the 2-D and 3-D representations. Modified worked examples proved superior to conventional worked examples without intermediate stages while conventional worked examples were, in turn, superior to problems. Experiment 3 investigated the consequences of varying the number and location of intermediate stages in the rotation trajectory and found three stages to be superior to one. A single intermediate stage was superior when nearer the 2-D than the 3-D end of the trajectory. It was concluded that (a) orthographic projection is learned best using worked examples with several intermediate stages and that (b) a linear relation between angle of rotation and problem difficulty did not hold for orthographic projection material. Cognitive load theory could be used to suggest the ideal location of the intermediate stages.     

Coordinating principles and examples through analogy and self-explanation

Research on expertise suggests that a critical aspect of expert understanding is knowledge of the relations between domain principles and problem features. We investigated two instructional pathways hypothesized to facilitate students’ learning of these relations when studying worked examples. The first path is through self-explaining how worked examples instantiate domain principles and the second is through analogical comparison of worked examples. We compared both of these pathways to a third instructional path where students read worked examples and solved practice problems. Students in an introductory physics class were randomly assigned to one of three worked example conditions (reading, self-explanation, or analogy) when learning about rotational kinematics and then completed a set of problem solving and conceptual tests that measured near, intermediate, and far transfer. Students in the reading and self-explanation groups performed better than the analogy group on near transfer problems solved during the learning activities. However, this problem solving advantage was short lived as all three groups performed similarly on two intermediate transfer problems given at test. On the far transfer test, the self-explanation and analogy groups performed better than the reading group. These results are consistent with the idea that self-explanation and analogical comparison can facilitate conceptual learning without decrements to problem solving skills relative to a more traditional type of instruction in a classroom setting.     

External and Internal Instantiation of Abstract Information Facilitates Transfer in Insight Problem Solving

Four studies were conducted to examine how instantiation of abstract statements facilitates solutions to insight problems. The first two experiments explored the effects of external instantiation (adding a concrete example to the abstract source information) on transfer. Providing an example along with the abstract source statement fostered transfer of analogous solutions. Moreover, adding a complete concrete instantiation to an abstract statement produced greater transfer than adding only specific items to illustrate the abstract terms. The second two experiments examined the effects of internal instantiation (encouraging learners to generate their own concrete examples of the abstract information) on transfer. This factor, too, facilitated problem solving. Generating examples to instantiate the abstract statement yielded even better transfer than simply informing participants of the relationship between the abstract statements and the target problems. These findings suggest that both external and internal instantiation facilitate transfer by promoting mental processes associated with implementing the source analogs.     

The consequences of fading instructional guidance on delayed performance: the case of financial services training

Empirical studies within a cognitive load framework have determined that for novice learners, worked examples provide appropriate levels of instructional guidance. As learners advance in specific subject domains, worked examples should be gradually replaced by practice problems with limited guidance. This study compared performance, both immediately post‐instruction and delayed, following instruction under different conditions: using example–problem pairs, using gradual fading of worked examples, and using pure problem‐solving. The study was conducted with employees of a financial services company in a classroom environment. Results indicated that the fading condition consistently outperformed the example–problem and problem‐solving conditions, and the advantage of this condition was enhanced, with statistically significant differences in performance, in delayed and transfer posttest performance.     

Completion of partially worked-out examples as a generation strategy for improving monitoring accuracy

Students’ Judgments of Learning (JOLs) are often inaccurate: students often overestimate their future test performance. Because of the consequences that JOL inaccuracy can have for regulating study activities, an important question is how JOL accuracy can be improved. When learning texts, JOL accuracy has been shown to improve through ‘generation strategies’, such as generating keywords, summaries, or concept maps. This study investigated whether JOL accuracy can also be improved by means of a generation strategy (i.e., completing blank steps in the examples) when learning to solve problems through worked example study. Secondary education students of 14–15 years old (cf. USA 9th grade) either studied worked examples or completed partially worked examples and gave JOLs. It was found that completion of worked examples resulted in underestimation of future test performance. It seems that completing partially worked-out examples made students less confident about future performance than studying fully worked examples. However, this did not lead to better regulation of study.     

Testing the model-observer similarity hypothesis with text-based worked examples

Example-based learning is a very effective and efficient instructional strategy for novices. It can be implemented using text-based worked examples that provide a written demonstration of how to perform a task, or (video) modelling examples in which an instructor (the ‘model’) provides a demonstration. The model-observer similarity (MOS) hypothesis predicts that the effectiveness of modelling examples partly depends on the degree to which learners perceive the models to be similar to them. It is an open question, however, whether perceived similarity with the person who created the example, would also affect learning from text-based worked examples. Therefore, two experiments were conducted to investigate whether MOS would also play a role in learning from worked examples. In Experiment 1 (N = 147), students were led to believe via pictures and a short story that the worked examples were created by a male or female peer student. Males showed higher performance and confidence, but no effects of MOS on learning were found. In Experiment 2 (N = 130), students were led to believe that a peer student or a teacher created the examples. Again, no effects of MOS were found. These findings suggest that the perceived origin of text-based worked examples is not important for learning.     

Differential effects of problem-solving demands on individual and collaborative learning outcomes

The effectiveness and efficiency of individual versus collaborative learning was investigated as a function of instructional format among 140 high school students in the domain of biology. The instructional format either emphasized worked examples, which needed to be studied or the equivalent problems, which needed to be solved. Because problem solving imposes a higher cognitive load for novices than does studying worked examples it was hypothesized that learning by solving problems would lead to better learning outcomes (effectiveness) and be more efficient for collaborative learners, whereas learning by studying worked examples would lead to better learning outcomes and be more efficient for individual learners. The results supported these crossover interaction hypothesis. Consequences of the findings for the design of individual and collaborative learning environments are discussed.     

Effects of worked examples on step performance in solving complex problems

The instructional effect of worked examples has been investigated in many research studies. However, most of them evaluated the overall performance of the participants in solving post-intervention problems, rather than individual step performance in multi-step problems. The two experiments reported in this article investigated the relations between using worked examples and individual step performance in solving isomorphic problems. In Experiment 1, the effect of worked examples was found for overall performance for novice learners, whereas this effect was gradually reduced from Step 1 (the most difficult one) at which the effect was the strongest, to Step 3 (the easiest one) at which the effect was the weakest or even disappeared. In Experiment 2, relatively more knowledgeable participants learned the same sets of materials, and no effect of worked examples was found for either overall performance or individual step performance. Learner levels of expertise and levels of element interactivity were used to explain the results.     

M.A.S.S.: A Model for Producing Transfer

Transfer is the application in the workplace of the knowledge, skills and attitudes learned in training. With transfer, trainers hope to link training to increased job performance. However, training alone will not produce transfer. To affect job performance as a result of training, trainers must intentionally promote transfer using a variety of strategies based on known principles of human performance technology. The MASS model, presented in this paper, brings together four of these principles. According to the MASS model, trainers who promote transfer (and who thereby become performance technologists) 1) Motivate trainees to learn and use the training material; 2) increase trainees' Awareness of when to use new skills and ideas; 3) enable trainees to master and to apply Skills; and 4) give trainees psychological and physical Support on the job. When performance technologists follow the MASS model, they can expect to produce trainees who apply at work what they have been taught in training. Use of the model is illustrated with two examples.     

Transfer of visual identity matching-to-sample in two california sea lions (zalophus californianus)

In order to assess the abilities of two California sea lions to generalize an identity concept, both animals were taught a two-choice, visual matching-to-sample task. We hypothesized that initial identity-matching problems would be learned as conditional (if...then) discriminations but that an identity concept would emerge after training numerous exemplars of identity matching. After training with 15 two-stimulus identity matching-to-sample problems, transfer tests consisting of 15 novel problems were given to the animals. Pass-fail criteria were defined in terms of performance on Trial 1 of each test problem, performance on test trials compared with baseline trials, and performance on four-trial problem blocks. One sea lion passed on the second transfer test and the other passed on the third; both demonstrated successful generalization of an identity concept by all criteria used. A second experiment consisted of presentation of stimuli previously learned in a different context (arbitrary matching-to-sample). Both subjects immediately applied an identity concept to accurately solve these new problems. These tests conclusively demonstrate transfer of an identity matching rule in California sea lions.