Learning Versus Performance
Author: Nicholas C. Soderstrom and Robert A. Bjork
"…instruction should endeavour to facilitate learning, which refers to the relatively permanent changes in behaviour or knowledge that support long-term retention and transfer. Paradoxically, however, such learning needs to be distinguished from performance, which refers to the temporary fluctuations in behaviour or knowledge that can be observed and measured during or immediately after the acquisition process."
An overview
This article provides an overview of the research into the difference between long-term learning and short-term performance. It is split into several sections, each of which I will describe briefly below. It is important to note that most of the sections look at "motor learning" (practical skills and movements) and "verbal learning" (knowledge and academic skills) separately, so will be of interest to all subjects.
1. Foundational Studies
This section is about research that is much older and focuses on the idea that learning can take place, even if performance is not shown. Within this section the authors talk about three separate ideas.
The first is "latent learning" which "is defined as learning that occurs in the absence of any obvious reinforcement". One example is given of a person riding a bus everyday who would learn the route, but would only show this if one day they were forced to drive to work. Another example from the research is about rats that are given regular treats, those with no treats, and those given none to start with, but then given treats from day 11. The delayed group quickly catch up with the group given regular treats by the next day, showing they had been learning from the start but not performing as there was no incentive.
The second is "overlearning ". Overlearning is the idea that once you have mastered something, you no longer show improvement in performance, but despite this, by continuing to practice you can still improve the long term retention, or learning. The example here is given of a violinist continuing to practice even after they have mastered the piece, and the benefits this had on long term mastery of the piece.
The third is "fatigue". Fatigue works in a similar way to overlearning, but the research "showed that when fatigue limited or eliminated gains in performance, learning nonetheless occurred, as revealed by a subsequent test on the task after the fatigue had dissipated".
"…long-lasting learning could occur while training or acquisition performance provided no indication that learning was actually taking place…"
2. Distribution of Practice
This section focuses on the way in which a curriculum or programme of study is built, and in particular looked at the two opposing ways of approaching this - massing practice or distributing practice. In massed practice you practice a particular skill in a block, then practice the next skill in a block. In distributed practice you practice the first skill, then the second, then return to the first, and so on.
The main example given is on participants practicing a skill to improve their time. One group completed 18 trials of the three patterns in a massed way, that is doing 18 trials of pattern 1, then 18 trials of pattern 2, then 18 trials of pattern 3. The other group had the 8 trials of each pattern randomly allocated on each go. Importantly, both groups had 18 trials of each pattern. The research showed that the group who did the massed practice did better immediately (their performance was better), but that on a retention test done 10 days later those in the distributed group outperformed the massed group (they had learned more).
The research throughout this section shows the same results: although distributing practice can appear to be less effective since performance is lower, it has a significant impact on the learning of the students, and so is a desirable thing to aspire to.
"…long-term learning profits from distributing (spacing) the practice of to-be-learned skills or information with time or other intervening activities…"
3. Variability of Practice
Variability refers to students practicing skills or knowledge in a way that introduces small changes or related tasks when practicing. The opposite would be to practice a single skill only, and then move to the related skills.
In this section one example relates to basketball practice, where one group is made to practice shooting from a given a fixed distance of 12 feet from the hoop, and the other group practiced at distances of 8 feet and 15 feet . Both groups completed the same total number of throws, and the retention test administered 72 hours later was done at a fixed distance of 12 feet (the same as the fixed group, but not a distance the varied group used). Surprisingly, the group from the varied practice did better on the retention test, averaging less errors (misses).
A similar experiment showed the same to be true for students who studied a list of words in a single fixed room, and another group of students who studied the list in two different rooms. When the test was done in a third room that none of the students had seen before, the group who practiced in different rooms did better.
Much like the examples from distributing practice, variability of practice generally leads to worse immediate performances, but better long-term learning.
"…variable practice might not appear to be effective during practice - specifically, more performance errors would likely be induced … - but would facilitate long-term learning…"
4. Retrieval Practice
Retrieval practice refers to the idea of practicing retrieving information from our memory. This is an active process, and is best seen in the form of tests. The article states "…retrieval processes triggered by testing actually changes the retrieved information in important ways…". That is, the act of (successfully) retrieving information in a test actually helps us learn that information.
The example given here is about two groups of students, one of whom studies three times before taking a final test, while the other group studied once, took two tests and then the final test. In the final test the group who studied more performed better, however, on retention tests 2 days later the group who had done the tests instead of extra studying did better, showing they had learned more.
Many studies have shown the same results, and what is most surprising is that even if the tested participants are given no feedback, they still learned more and did better on the retention tests. This suggests that if they had been given feedback to improve their knowledge after each test, they would have done even better.
"…practicing retrieval…may appear unhelpful during acquisition and on immediate memory tests, but it provides substantial benefits in preserving or stabilizing long-term memory…"
5. Metacognition
"…there appears to be a lack of understanding on the part of instructors and learners alike that performance during acquisition is a highly imperfect index of long-term learning…"
This section focuses on the natural inclination of both teachers and students to favour those methods which promote short-term performance, but are actually worse for long-term learning. Students were asked which methods they used for studying and most said rereading material, despite the huge body of evidence for the so called testing effect. Those that did test themselves did it to highlight gaps in their knowledge, not as a learning task in itself, as suggested by retrieval practice.
Moreover, students who were taught in a "distributed-practice group reported being … less satisfied with their training because they felt they were falling behind the massed-practice group" despite actually doing better in the retention test, and showing more learning.
It also mentions the concept that in general "learners tend to be overconfident in predicting their own learning", and "…learners often mistakenly conflate short-term performance with long-term learning".
My Takeaways
One thing that struck me was the nature of performance vs learning, and that most of us as teachers (I know I do anyway) rely on performance measures to judge learning, and according to this review, this is a big flaw. Indeed the whole section on metacognition highlights that both teachers and students do this regularly, and that students often go even further and prefer the teaching methods that promote short-term performance as opposed to long-term learning.
I have already started to incorporate ideas of retrieval practice into my classes in several ways. With my 5B group I am using an exam question on a topic that we have done previously (not the current topic) in each lesson as the starter. This is aiming to continue to get students to retrieve knowledge and hence develop their learning. Similarly, with my S3 class I am now doing weekly review quizzes on a given topic that we have already finished. I have tried to explain in brief the ideas of retrieval practice to my S3 students as well, to make them more aware of why I am doing this, and address the metacognition issue.
With regards to variability of practice and distribution of practice, I need to think about how I (or we as a department) can incorporate more of this into my teaching. I certainly don’t show one type of question at a time and get students to practice that one skill for 10 minutes, so I guess I already am using the premise of variability. And distribution is sort of built in to our schemes of work as we repeat topics, but I interpret this more as teaching topics interleaved with each other.
This has been another fascinating article to read, and I think the ideas of short-term performance versus long-term learning are ones that need to be addressed, both as teachers and also directly with students, so as to displace their misconceptions over what constitutes good learning.