Jeroen van Merriënboer’s doctoral work was on teaching computer programming.1 He championed a style of teaching centered on “completion tasks.” van Merrienboer would present students with a partially-complete program. Then, he tasked his students with filling in the missing pieces to make the program functional. At first, the gaps were quite small, but with time he would leave larger and larger sections of the program for his students to complete.
van Merriënboer found that these completion tasks were often very effective, more effective than worked examples. Why?
“…students will often skip over the examples, not study them at all, or only start searching for examples that fit in with their solution when they experience serious difficulties in solving a programming problem. … [In completion problems] students are required to study the examples carefully because there is a direct, natural bond between examples and practice.”
Studying worked examples could help students learn to write software, but they had to be properly motivated to do so. Just presenting students with an explanation was not enough, in practice.
While van Merriënboer’s early work does not mention CLT, he soon came to embrace Sweller’s theory. In 1994 van Merriënboer published another study (“Variability of Worked Examples and Transfer of Geometrical Problem-Solving Skills: A Cognitive-Load Approach”).2 This new paper was firmly within the CLT framework, and its findings supported Sweller’s ideas about extraneous cognitive load. At the same time, the paper challenged the idea that, when it comes to cognitive load, the lower the better.
In his study, van Merriënboer had heeded Sweller’s warnings and taken care not to unduly burden his students. Following Sweller, again, he had tasked his students with studying a series of worked examples. The use of worked examples had ensured that participants in the study had mental fuel to spare. What do you do with that spare mental fuel, though? In Sweller’s earlier experiments, that leftover capacity had been ignored. van Merriënboer, in contrast, realized that if “bad” cognitive load was reduced, there was an opportunity to increase students’ load in a more productive way. Rather than discarding this spare capacity, he could reinvest it into learning.
For some of his students, van Merriënboer increased the cognitive load by increasing the variability of their worked examples. In the low-variability condition, each worked example was followed by another that was identical except for the numbers. If one example showed how to find the distance between two points, the next one did too. In the high-variability condition, the second problem was changed. “Find the X-Coordinate of P2 given the distance between P1 and P2 and all the other coordinates.” While the mathematics of this second problem overlaps significantly with the first, the problem was entirely new.
On the one hand, the high-variability learning activity was significantly harder for students. It took them longer to finish the activity, and the students reported a higher degree of mental effort. At the same time, these students significantly outperformed their low-variability counterparts in a follow-up test. Unlike the effort that Sweller had studied, this cognitive load seemed to be good for learning. Along with his results showing the advantages completion tasks had over worked examples, van Merriënboer had pushed CLT into new territory.
van Merriënboer’s research didn’t contradict any of Sweller’s results, but they were challenging to the direction of his work. Their was summarized several years later in a joint paper by Sweller and van Merriënboer3:
“Until now, cognitive load theory research almost exclusively has studied instructional designs intended to decrease extraneous cognitive load. Recently, some studies have been conducted in which [cognitive load] was increased for processes considered to be directly relevant to schema construction.”
Why hadn’t Sweller come up with completion tasks or high-variability examples in his own work? van Merriënboer and Sweller came from different traditions of research. Sweller’s work involved “basic” learning, in the sense that the learning he studied involved acquiring isolated skills in a laboratory setting. As van Merriënboer’s doctoral work shows, he came from a world of “complex” learning. The learning that van Merriënboer studied was the acquisition of competence in an entire domain of inter-related skills (computer programming) and took place in a classroom, not a laboratory.
CLT had been created out of Sweller’s work with basic learning. For acquiring these sorts of skills, a more limited instructional toolkit was sufficient. There is usually just one skill being taught — it’s unsurprising that Sweller hadn’t introduced high-variability conditions. Sweller was also working with highly-motivated participants in a laboratory, not students in a classroom. He didn’t need to worry about them not being properly motivated to study his worked examples with care. This issue came up for van Merriënboer, though, because he was working with students in a classroom. van Merriënboer was taking CLT into newer, more complex learning.
In 1996, Sweller spent a sabbatical at van Merriënboer’s university and the two tried to bring their approaches together. In a later reflection on this collaboration, van Merriënboer (characteristically) suggested that their work was difficult, but in a good way4:
John and I encountered many problems in bringing cognitive load theory and models for complex learning together, because they are rooted in very different traditions. But problems are there to be solved and we always have a lot of fun doing so.
Their collaboration resulted in an article, “Cognitive Architecture and Instructional Design.”5
In their joint work, they introduced a new type of cognitive load, which they called germane load. Germane load was — like the effort introduced by completion tasks or high-variability — an addition mental burden that was good for learning. It’s opposite was extraneous learning, that load which was bad for learning. Following van Merriënboer’s lead, the goal for CLT was no longer merely to reduce extraneous load, but to then use that newly available mental capacity to good effect. “Learners’ attention must be withdrawn from processes not relevant to learning and directed toward processes that are relevant to learning,” they wrote.
The collaboration with van Merriënboer finds Sweller working with a larger palette of learning factors. Echoing van Merriënboer’s earlier work, their joint paper points out the benefits completion tasks can have over worked examples:
A lack of training with genuine problem-solving tasks may have negative effects on learners’ motivation. A heavy use of worked examples can provide learners with stereotyped solution patterns that may inhibit the generation of new, creative solutions to problems…For this reason, goal-free problems and completion problems…may offer a good alternative to an excessive use of worked examples.
Before this collaboration, Sweller had not written rarely written about student motivation.6 Likewise, he had not been concerned with long-term learning issues resulting from “stereotyped” solution patterns. CLT was moving into complex learning, and it was changing in the process.
Sources: van Merriënboer and Complex Learning
- van Merriënboer, 1990. ↩
- Paas & van Merrienboer, 1994. ↩
- Sweller, van Merrienboer & Paas, 1998. ↩
- http://archive.sciencewatch.com/dr/erf/2009/09augerf/09augerfMerr/ ↩
- Sweller, van Merrienboer & Paas, 1998. ↩
- Sweller & Cooper, 1985 discussed the motivational consequences of only using worked examples as opposed to alternating examples with problems. “It was assumed that motivation, while reading a worked example, would be increased by the knowledge that a similar problem would need to be solved immediately afterwards,” they write. ↩
"Not a Theory of Everything" 