Featured article: Can we learn from errors? Retrieval facilitates the correction of false memories for pragmatic inferences; Maraver, Lapa, Garcia-Marques, Carneiro & Raposo. PLoS One 2022; Faculty of Psychology, University of Lisbon, Portugal; https://pubmed.ncbi.nlm.nih.gov/35917361/
Given the overwhelming data that supports the efficacy of the ‘testing effect’ (AKA ‘retrieval practice’), no one doubts its power at improving long-term retention of learning. Concerns remain, though, as to whether submitting an incorrect answer to a test item causes the learner to remember the incorrect answer rather than the correct response.
A group of researchers at the University of Lisbon recently tested two important research questions: (1) to what degree does corrective feedback improve error correction among learners and (2) does this the degree of error correction vary among students taught by retrieval practice (testing!) versus passive studying. To answer these questions, they randomized 120 undergraduate students via a 2×2 factorial design into 4 groups:
1. Retrieval practice – corrective feedback
2. Retrieval practice – no corrective feedback
3. Restudy – corrective feedback
4. Restudy – no corrective feedback
During an encoding phase, all students were instructed to read 30 sentences presented in random order for 4.5 seconds each.
In the intermediate (intervention) phase, the retrieval practice students were presented in random order the 30 sentences with a critical fragment missing (e.g. The baby _____ all night) for 10.5 seconds and were given 20 seconds to provide the missing word. The restudy students were presented the full sentence for 10.5 seconds and asked to simply to read the text. Those in the corrective feedback group were presented the complete sentence for 4.5 seconds after submitting their answer; to facilitate their later analyses, the researchers presented an incorrect full sentence 50% of the time. Those in the ‘no corrective feedback’ group performed an easy math problem (e.g. 23-7=) for 4.5 seconds.
In the assessment phase, all students were presented the 30 sentences in random order with a critical fragment missing (e.g. The baby _____ all night) and were given 60 seconds to provide the missing word.
Upon analyzing their data from the final assessment, the researchers made several important discoveries. First, as expected, retrieval practice improved performance compared to restudy (again, the testing effect in action). Second, accurate recall was higher when participants received feedback than no feedback, regardless of the learning condition. Third and most surprisingly, corrective feedback was more effective when it was presented after retrieval practice rather than after restudying practices. I find this result to be astounding! The authors hypothesize that retrieval practice (testing) induces ‘memory malleability’ that increases the efficiency by which new related (correct) information can be incorporated into longer-term memory, in contrast to restudy of the material. The researchers state “Even when (an answer) is wrong, retrieval does more than correcting wrong information, and it through the effortful engagement in intentional explicit retrieval … (that) memory (is) in a sufficient malleable state to allow the incorporation of feedback.” Not only did the researchers convincingly show that corrective feedback is an important tool to improve learning, but they also showed that corrective feedback is most effective when presented in the context of retrieval practice (testing).
Based on data from several earlier research studies, we at Qstream have always felt that corrective feedback after answering a question (including an explanation as to why the correct answer was correct) was an important component to Qstream’s educational efficacy. Little did we know that providing the feedback in the context of Qstream’s testing methodology increased the platform’s educational efficacy even further.
