Retrieval Practice Improves Learning, But Will it Help ALL of My Students? More Experimental Evidence
Cover image by Prawny from Pixabay
By Megan Sumeracki
In the first year that our blog was created—2016—I wrote a piece titled, Retrieval Practice Improves Learning, But Will it Help ALL of My Students? In this piece, I covered an experiment conducted by Pooja Agarwal and colleagues (1) about the benefits of retrieval practice for students with different working memory capacities. The main question was whether retrieval practice would benefit students with a wide range of working memory capacity—or the ability to hold information in mind and engage in attentional control. It is reasonable to ask whether retrieval practice, given the needed balance of success and difficulty, might benefit students with higher working memory ability more—a rich get richer situation. If retrieval practice only benefits certain types of students, then it would be important to know that! Thankfully, that was not what they found.
The main findings were that retrieval practice led to greater performance on a learning assessment than repeated “studying” or reading, and this was true whether the students had lower or higher working memory capacities (1). In fact, when the learning assessment was completed one week after learning, they students with lower working memory capacity actually benefited more from retrieval practice than those with higher working memory capacity. The authors argued that retrieval practice was beneficial for students with a range of working memory ability, not just students with higher working memory capacity.
In today’s post, I cover another research article asking whether retrieval practice is equally beneficial for students with different skills. Bruna Fernanda Tolentino Moreira and colleagues (2) conducted two experiments, one each with 4th and 6th grade children, investigating whether retrieval practice was equally beneficial for students with different visual word decoding skills. The authors note that children’s visual word decoding abilities tend to differ dramatically, even when looking at children in the same age range. If retrieval practice is less effective, not effective, or harmful for learning for some children, it is important to know this!
The Experiments
Experiments 1 and 2 were very similar. Experiment 1 was conducted with 6th graders, and experiment 2 was conducted with 4th graders. There were a couple of other differences: In Experiment 2 a few additional measures were included. The children also practiced retrieval and restudied twice rather than just once in, and only completed one final assessment test instead of two. However, the methods were largely the same, and the main findings were also the same across experiments.
Experimental Retrieval Practice Procedure:
Children first participated in a typical retrieval practice experimental procedure. The children first read a text passage about the sun twice—once by themselves and once along with the experimenter. Then, after a brief distraction task (solving math problems), the children practiced retrieval for some of the information in the text and reread other information (the retrieval practice and restudying conditions, respectively). During retrieval practice the children were given a cued-recall test; they saw sentences from the text passage with key words removed, and they were asked to remember the word. During restudy, the word was presented in bold font within the sentence.
For example, when the words Hydrogen and photosphere were assigned to the retrieval practice condition, they looked like this:
The Sun is mostly made of a gas called ________ and a gas called Helium
The superficial layer of the sun is called ________
When the words Hydrogen and photosphere were assigned to the restudy condition, they looked like this:
The Sun is mostly made of a gas called Hydrogen and a gas called Helium
The superficial layer of the sun is called photosphere
One week later, the children completed final assessment tests (cued recall and multiple-choice in Experiment 1, and multiple-choice in Experiment 2).
Visual Word Decoding Ability
A few days after the final learning assessment was administered, the children performed a reading aloud task to assess visual word decoding ability. The children read Brazilian Portuguese nouns out loud (the experiment was conducted in Brazil) that were classified as regular, irregular, or pseudowords that were pronounceable but not true Brazilian Portuguese words. The children were instructed to read the words out loud, and they were scored on the number of words they pronounced correctly, and how quickly they responded.
Other Measures
In Experiment 2, the authors added other measures to assess cognitive ability. They attempted to measure reading comprehension, vocabulary skills, an estimated IQ, and working memory capacity.
The Main Findings
In Experiment 1 with 6th graders, retrieval practice led to greater performance on the final learning assessment than restudying. Importantly, the benefit was similar across children with different word decoding abilities.
In Experiment 2 with 4th graders, the same was true! The size of the retrieval practice benefit, compared to restudying, was the same across children with different word decoding abilities.
Word decoding ability was positively correlated with final assessment test performance overall. When children had greater word decoding ability, their performance on the final assessment test was higher overall. Still, across word decoding ability, the size of the retrieval practice effects were comparable. This means that those with higher word decoding abilities generally did better in both conditions than those with lower word decoding abilities, but retrieval practice led to better performance compared to restudying across students. In other words, retrieval practice did not increase or decrease the performance gap between students with higher or lower verbal decoding ability. Retrieval practice seemed to work equally well.
In Experiment 2, the other measures included, such as reading comprehension, vocabulary skills, intelligence (estimated IQ), and working memory, all had no relationship to the size of the retrieval practice effect.
Taken together, the main conclusion from this research paper was that retrieval practice does seem to be effective for students with varying abilities!
References
(1) Agarwal, P. K., Finley, J. R., Rose, N. S., & Roediger, H. L. (2017). Benefits from retrieval practice are greater for students with lower working memory capacity. Memory, 25(6), 764-771. https://doi.org/10.1080/09658211.2016.1220579
(2) Moreira, B. F. T., Pinto, T. S. da S., Justi, F. R. R., & Jaeger, A. (2019). Retrieval practice improves learning in children with diverse visual word recognition skills. Memory, 27(10), 1423-1437. https://doi.org/10.1080/09658211.2019.1668017
