During the learning phase in Experiment 1, undergraduate students watched video clips about science topics and then answered application questions from their memories. The experiment had two simple conditions: In the “same” condition, for each concept the students answered questions requiring them to apply the concept to the same example three times. In the “variable” condition, for each concept the students answered questions requiring them to apply the concept to three different examples about that same concept. The students got feedback after they answered each question.
Two days later, the students took a final test. The final test had all new application questions (i.e., none of the participants had seen these before).
The results: students did better on the final application test if they answered variable application questions during retrieval practice than if they answered the same question. These findings are in line with what we see with research on concrete examples and transfer as well—students are better able to transfer what they have learned to novel examples if they were given multiple, varied concrete examples (see this post).
The primary finding across the remaining experiments:
In the remaining experiments, the researchers replicated the main finding from Experiment 1. That is, practice retrieving and applying concepts to multiple, varied examples led to better transfer to novel examples later compared to applying the concept to one example multiple times. This effect was consistent and relatively large across all four experiments.
For those interested in the ways in which the later experiments were modified, below is a brief summary:
In Experiment 2, the researchers manipulated whether labels were present with each question to help the students make the link between the questions students answered about the same concept. Including the label slightly improved application performance for those students answering varied application questions during learning.
In Experiment 3, they added restudy control conditions. In these conditions, students studied each example instead of answering application questions about the example. They either studied the same example three times, or they studied three different examples. Overall, both retrieval practice conditions led to greater performance than the restudying (variable retrieval = 68%; same retrieval = 53%; variable restudying = 51%; same restudying = 38%). The variability manipulation, same examples vs. variable examples, showed a similar pattern in the restudy conditions as in the retrieval conditions—restudying variable examples led to greater application performance than restudying the same example three times. (For those interested in the statistics, both main effects were statistically significant, and the interaction was not statistically significant.)
Finally, in Experiment 4 the researchers basically replicated Experiment 3, but added another component to the final assessment test. After answering the application questions to measure the students’ ability to transfer to novel examples, the researchers gave the students descriptions of the concepts and let them try all the questions again. This allowed the researchers to see whether lower performance was due to an inability to remember what the concept was, an inability to appropriately apply the concept to the new situation, or both. The results showed the same pattern as Experiment 3 on the first part of the application test. On the second part, the effects largely went away, suggesting that variable retrieval practice improves application of knowledge by improving retention. However, when students restudied the same examples, their score was the lowest. Thus, retrieval practice and engaging with varied examples likely also promotes a deeper understanding.
Practicing retrieval by applying the information to multiple, varied examples improves students ability to transfer what they are learning and apply those concepts to novel situations.
Sometimes in education we have a sense of how students will use the concepts they are learning in the future. When that is the case, we absolutely should give them some practice that is similar to those likely future scenarios. However, so often we are not sure exactly how the students might use the concepts they are learning. The applications could be endless! This research, and other literature as well, suggests that giving students practice applying to a wide variety of scenarios will help them generalize, and be better able to apply to a completely novel situation later.
References:
(1) Butler, A. C., Black-Maier, A. C., Raley, N. D., & Marsh, E. J. (2017). Retrieving and applying knowledge to different examples promotes transfer of learning. Journal of Experimental Psychology: Applied, 23(4), 433-446. https://doi.org/10.1037/xap0000142
