The benefit of generating errors during learning: What is the locus of the effect?

“hispid) and asked how curious they were to learn their meaning [45].”

“Learners also appear to be unaware of this “errorful generation effect,” as measured by item-level JOLs, aggregate JOLs, and questionnaires assessing learners’ beliefs (Huelser and Metcalfe 2012; Potts and Shanks 2014; Potts et al. 2019; Yang et al. 2017; Zawadzka and Hanczakowski 2019).”

“When errors are not semantically associated with targets, generating errors benefits memory for targets only through increased attention, motivation, and/or positive emotions (e.g., Clark et al., 2021; Potts et al., 2019).”

“…illustrated in Table 1, studies involving paired or triple associate words (e.g., Kornell et al., 2009; Pan et al., 2019; Yan et al., 2014; see also Potts et al., 2019), trivia facts (e.g., Kor-nell, 2014), and text passages (e.g., Kliegl et al., 2022; Little & Bjork, 2016) have all shown memory…”

“However, this potential errorless learning advantage may not persist if inference attempts are followed by definitions because conceptual errors facilitate retrieval (Cyr &Anderson, 2015) and encourage deeper engagement with definitions (Potts et al., 2018).”

“This might be especially likely given recent evidence to suggest that errorful generation increases participants’ curiosity (Potts et al., 2019) and motivation (in the present experiments) to learn the true answers.”

“It may also be that generating errors enhances encoding of their subsequent correction (e.g., Hays et al., 2013; Kornell et al., 2009; Potts et al., 2019), such that correcting one’s deliberate errors has been found to produce better learning than leaving them uncorrected (Wong and Lim, 2022b).”

“There is also evidence that stimulating curiosity for information by guessing it before it is learned improves recall of the information (Berlyne, 1954; Brod et al., 2018; Kang et al., 2009; Potts et al., 2019).”

“A guessing or inferencing task can also pique learners’ curiosity, which will then make them attend more closely to the correct meaning when it is offered as feedback (e.g. Hays et al., 2013; Potts et al., 2019; Potts & Shanks, 2014; Zawadzka & Hanczakowski, 2019).”

“Critically, however, attentional accounts typically assume that pretesting enhances primarily attentive encoding of the response word itself, but does not strengthen the association between stimulus and response word (Potts et al., 2019; Seabrooke et al., 2021).”

“…that prequestions make learners curious to learn the correct answers, and this increased curiosity directs attention to the learning material during the study phase (Berlyne, 1954a, b, 1962; Bull & Dizney, 1973; Patrick, 1976; see also Metcalfe & Finn, 2011; Potts et al., 2019; Wade & Kidd, 2019).”

“This chimes with the findings of Seabrooke et al. (2019), who were the first to demonstrate that guessing might not be the best strategy for foreign vocabulary acquisition.”

“For instance, recent research suggests that when participants are unable to provide a correct answer at test, their motivation to learn the correct answer when it is subsequently presented as feedback may be particularly high (e.g., Kang et al., 2009; Potts and Shanks, 2019).”

“Thus, there may be enhanced encoding of corrective feedback relative to other aspects of learning trials (Potts et al., 2019), even when the feedback is not necessarily surprising (cf.”

“Recent work by Potts et al. (2019) on novel vocabulary learning has shown better learning success when participants had to guess the meaning of a novel word first than when they were presented with form-meaning pairs from the start.”

“Overall, we provide additional support for the enhanced processing of feedback as a key consequence of error generation [24].”

“While some have theorised that such states might explain the pre-testing e ff ect (e.g., Potts et al., 2019; Potts & Shanks, 2014), they have done so only with respect to memory of the corrective feedback itself and have not explored whether there is a wider bene fi t to learning after a guess.”

“Moreover, investigations into the antecedents of retroactive curiosity have mostly been limited to one type of prediction error, when individuals’ confident response is incorrect (negative prediction error) (e.g., Potts et al., 2019).”

“First, it has been proposed that what drives the benefits of guessing is curiosity, by which making a guess makes participants more eager to learn the correct answer, resulting in greater attention towards feedback and hence its better encoding (Potts et al., 2018).”

“Priming curiosity through drawing participants’ attention to the unknown can increase engagement retention of material (Potts et al., 2019; Ruan et al., 2018).”

“This pattern is consistent with previous results (Potts et al., 2019; Potts & Shanks, 2014; Seabrooke, Hollins, et al., 2019; Seabrooke, Mitchell, et al., 2019; Seabrooke, Mitchell, Wills, & Hollins, 2021; Seabrooke, Mitchell, Wills, Inkster, et al., 2021).”

“In a later study by Potts, Davies and Shanks (2019), participants either attempted to retrieve the translations of L2 words immediately before or after receiving corrective feedback.”

“One possibility is that participants attend more to the feedback when it contradicts a response participants thought was correct and therefore encode it more effectively (Potts et al., 2019).”

“This account assumes that pretesting boosts the encoding of the target item itself, without strengthening the link between the cue and the target items (e.g., Potts et al., 2019; Seabrooke et al., 2021).”

“The second mechanism that can account for our findings suggests that the errorful generation of language forms, especially in the absence of prior knowledge, is particularly beneficial for learning, as has previously been shown for vocabulary learning (Potts & Shanks, 2014; Potts et al., 2019).”

“Yet , a growing body of evidence shows that incorrect guessing can improve memory for second-language vocabulary items compared to errorless learning (Potts & Shanks, 2014; Potts et al., 2019; Seabrooke et al., 2019).”

“As discussed, most studies investigating the errorful generation effect have utilised a cued-recall test in the context of a mixed-list design (Metcalfe & Xu, 2018; Pan & Rivers, 2023; Potts et al., 2019; Potts & Shanks, 2014; Seabrooke et al., 2019a; Seabrooke et al., 2019b; Yang et al., 2017).”

“Recent research indeed has yielded evi - dence that both types of processes may play a central role for the pretesting effect (e.g., Bartl et al., 2024 ; Huelser & Metcalfe, 2012 ; Potts et al., 2019 ; Sana & Carpenter, 2023 ).”

“A study by Potts et al. (2019) demonstrated that individuals learning pairs of rare English words and their definitions gave higher curiosity ratings when they had to guess a definition and when ratings were solicited after, and not prior to, the act of generation.”

“…showing that participants rated their curiosity about the correct response higher if the rating was given after, rather than before, a guessing attempt had been made during pretesting, suggesting that pretests may indeed raise participants’ curiosity about the correct response (Potts et al., 2019).”

“As an alternative hypothesis, Potts et al. (2019) proposed an information gap, defined as the difference between what we know and want to know.”

“Evidence from Potts et al. (2019) suggests a beneficial effect of ‘guessing’ on vocabulary retrieval.”

“…target information, possibly due to guessing increasing curiosity about the upcoming answer, causing the participant to allocate greater attentional resources to process that information (e.g., Potts & Shanks, 2014; Potts et al., 2019; Sana & Carpenter, 2023; Zawadzka & Hanczakowski, 2019).”

“Previous work has shown that learners corrected knowledge based on self-reported motivation (Seabrooke et al., 2019), perceived usefulness of the to-be-learned information (Iwaki et al., 2013), and reported curiosity to learn the answers (Potts et al., 2019).”

“Pretesting, also known as errorful generation, aligns with prediction-based learning: by making knowledge gaps salient, it can heighten attention to key arguments and trigger coactivation of relevant prior knowledge, facilitating integration of new content 13,19,28–31 .”

“The multiple-choice findings, however, better align with prior results involving purely verbal materials (e.g., Hol-lins et al., 2023; Potts & Shanks, 2014; Potts et al., 2019).”