The DOI matches its Crossref record; 77 of the 77 references it cites were found in a registry or an index; 24 of the 58 sources its claims cite were checked for support.
The benefit of generating errors during learning
Open accessAbstract
Testing has been found to be a powerful learning tool, but educators might be reluctant to make full use of its benefits for fear that any errors made would be harmful to learning. We asked whether testing could be beneficial to memory even during novel learning, when nearly all responses were errors, and where errors were unlikely to be related to either cues or targets. In 4 experiments, participants learned definitions for unfamiliar English words, or translations for foreign vocabulary, by generating a response and being given corrective feedback, by reading the word and its definition or translation, or by selecting from a choice of definitions or translations followed by feedback. In a final test of all words, generating errors followed by feedback led to significantly better memory for the correct definition or translation than either reading or making incorrect choices, suggesting that the benefits of generation are not restricted to correctly generated items. Even when information to be learned is novel, errorful generation may play a powerful role in potentiating encoding of corrective feedback. Experiments 2A, 2B, and 3 revealed, via metacognitive judgments of learning, that participants are strikingly unaware of this benefit, judging errorful generation to be a less effective encoding method than reading or incorrect choosing, when in fact it was better. Predictions reflected participants' subjective experience during learning. If subjective difficulty leads to more effort at encoding, this could at least partly explain the errorful generation advantage.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
24 of 58 cited sources checked · 16 supported · 1 flagged · accuracy not yet measured
Every source the paper cites for this claim was checked and supports it.
In the paper
“Kornell et al. found that the test condition led to better final test performance than the read-only condition.”
§ The benefit of generating errors during learning > Unsuccessful retrieval
What the sources say
- Supportedcites Kornell et al., 2009 source
The Results and Discussion (p. 6) report significantly higher final-test cued-recall accuracy in the test condition than in the read-only condition in Experiment 4. The same pattern is reported for other experiments on pp. 6–7. (LLM, conf=0.94)
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“On the other hand, there is also evidence that generating responses can be beneficial even when many errors are produced, as long as corrective feedback is given (e.g., Kornell, Hays & Bjork, 2009).”
§ The benefit of generating errors during learning > Abstract
What the sources say
- Supportedcites Kornell, Hays & Bjork, 2009 source
The General Discussion (p. 7) reports that unsuccessful retrieval attempts accompanied by feedback enhanced learning. In Results and Discussion (p. 6), 77% of responses were errors of commission, yet final recall was significantly higher after retrieval attempts with feedback than after studying the answers for the same amount of time. The Concluding Comment (p. 8) explicitly says feedback is critical. (LLM, conf=0.94)
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“However, there is evidence that feedback may enhance the benefits of tests (Butler & Roediger, 2008;Pashler, Cepeda, Wixted & Rohrer, 2005).”
§ The benefit of generating errors during learning > Abstract
What the sources say
- Supportedcites Butler & Roediger, 2008 source
In “The Acquisition of Misinformation,” the paper reports that feedback enhanced the observed testing effect: both immediate and delayed feedback produced large gains in correct responses compared with no feedback. This directly supports the citing sentence’s hedged claim that feedback may enhance the benefits of tests. (LLM, conf=0.94)
- Supportedcites Pashler, Cepeda, Wixted & Rohrer, 2005 source
The Results and Discussion section (page not provided) reports that correct-answer feedback after an incorrect test response improved final-test performance and increased one-week retention compared with no feedback. This supports the citing sentence’s qualified claim that feedback may enhance the benefits of tests. (LLM, conf=0.92)
Strength:hedged· not yet calibratedCausal:conditional· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Even in the undetermined condition, where participants nearly always produced an error, generating led to better final test performance than reading.”
§ The benefit of generating errors during learning > Unsuccessful retrieval
What the sources say
- Supportedcites Kane & Anderson, 1978 source
The General Discussion reports that subjects in the Undetermined Sentence Completion group were almost always wrong when supplying the missing word, yet recalled more on the test one week later than the Reading Only control group. This supports the claimed advantage of generating over reading in that condition. (LLM, conf=0.93)
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes- Supported
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“Moreover, it has been found that the harder the test, and the greater the effort required for retrieval, the greater the benefit to subsequent memory (e.g., Carpenter & DeLosh, 2006;Pyc & Rawson, 2009).”
§ The benefit of generating errors during learning > Abstract
What the sources say
- Supportedcites Carpenter & DeLosh, 2006 source
The General Discussion reports that items retrieved with fewer letter cues were retained better on a later test, a pattern replicated in Experiment 3. It interprets reduced cue support as requiring more elaborative retrieval processing and concludes that test benefits were greatest under those conditions. (LLM, conf=0.91)
- Not checked — no full text of the sourcecites Pyc & Rawson, 2009 source
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Furthermore, it is difficult to explain Kane and Anderson's (1978) findings by Grimaldi and Karpicke's account, since in their undetermined condition the targets were unrelated to the most obvious, incorrect, completions so it is highly unlikely that they would have been retrieved along with them.”
§ The benefit of generating errors during learning > Errorful generation
What the sources say
- Partly supportedcites Kane and Anderson's (1978) sourceExperimental — accuracy not yet measured
Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.
Load-bearingStrength:hedged· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes- Partly supported
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“There is also evidence that greater attention is given to feedback that does not match expectations (e.g., Butterfield & Metcalfe, 2001;Fazio & Marsh, 2009)”
§ The benefit of generating errors during learning > Errorful generation
What the sources say
- Not checked — no full text of the sourcecites Butterfield & Metcalfe, 2001 source
- Supportedcites Fazio & Marsh, 2009 source
The General Discussion states that unexpected feedback leads to greater effort to encode it and describes the present studies as showing a focus on that feedback. The experiments found better memory for feedback that did not match participants’ expectations, supporting the claim about greater attention. (LLM, conf=0.94)
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesThe cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Kornell et al. found no advantage of generating over reading when total trial time was equated.”
§ The benefit of generating errors during learning > Errorful generation
What the sources say
- Partly supportedcites Kornell et al., 2009 sourceExperimental — accuracy not yet measured
Experiment 2 held total trial time constant and found no significant difference in final recall for fictional items (Method and Results and Discussion, p. 4). But the claim presents that result as general: Experiments 4 and 6 also equated trial time and found significantly better recall after retrieval attempts than after reading (Results and Discussion, pp. 6–7). (LLM, conf=0.88)
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes- Partly supported
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“However, it has also been found that ease of processing, or encoding fluency, influences JOLs (e.g.Castel, McCabe, & Roediger, 2007;Koriat, 2008;Hertzog, Dunlosky, Robinson & Kidder, 2003;Schwartz, Benjamin & Bjork, 1997).”
§ The benefit of generating errors during learning > Errorful generation
What the sources say
- Supportedcites Castel, McCabe, & Roediger, 2007 source
The General Discussion (page not supplied) says that fluency plays a critical role in judgments of learning (JOLs) and that fluent identical pairs receive inflated JOLs. Experiment 2 also identifies ease of processing as a proposed reason for their high JOLs. (LLM, conf=0.93)
- Supportedcites Koriat, 2008 source
The Results and Discussion state that ease of learning affects JOLs and that processing fluency makes a unique contribution to them. The paper also identifies study time and trials to acquisition as indicators of encoding fluency, supporting the paraphrase. (LLM, conf=0.94)
- Not checked — no full text of the sourcecites Hertzog, Dunlosky, Robinson & Kidder, 2003 source
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“In both Grimaldi and Karpicke's (2012) and Huelser and Metcalfe's (2012) studies, participants generated guesses which were highly related to the cues in both the related and unrelated conditions (e.g., sleep for the cue pillow).”
§ The benefit of generating errors during learning > Errorful generation
What the sources say
- Partly supportedcites Grimaldi & Karpicke, 2012 sourceExperimental — accuracy not yet measured
The Experiment 1 procedure says participants guessed a target for both related and unrelated cue–target pairs. The Experiment 2 Materials section says the lure for each cue was its most frequently produced response, and Experiment 2 describes a lure as related to its cue. These passages support cue-related guesses in both pair types, but do not establish that participants’ guesses generally were highly related to the cues. (LLM, conf=0.81)
- Supportedcites Huelser & Metcalfe, 2012 source
The Latent semantic analysis section reports that participants generated errors related to the cue in both the related- and unrelated-materials conditions. It gives similar cue–error relatedness values for the two conditions in Experiment 1 (.28 and .25), and reports the same pattern in Experiment 2. The provided passages do not give page numbers. (LLM, conf=0.94)
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes- Partly supported
- It may state two separate claims as one. 75
- Probably not a substantial claim. 36
- Not yet classified. 5
- The quote only partly states this claim. 3
Some cited sources support this claim and some don’t — read each source’s verdict below.
In the paper
“Our findings are inconsistent with these proposals.”
§ The benefit of generating errors during learning > General Discussion
What the sources say
- Disputedcites Grimaldi and Karpicke (2012) sourceExperimental — accuracy not yet measured
The Results and discussion for Experiment 1 say the findings are consistent with a search-set account in which related candidates are activated during retrieval, and that the memory benefit occurred for related cue–target pairs but not unrelated pairs. The paper instead finds general error-correction and additional-cue theories difficult to reconcile with the results. It does not say the correct answer had to be already known during generation. (LLM, conf=0.85)
Checked by a reasoning model — may be wrong.
- Partly supportedcites Huelser & Metcalfe, 2012 sourceExperimental — accuracy not yet measured
The Learning phase performance section reports a memory benefit after incorrect guesses, with correctly guessed trials excluded; this supports the challenge to an account requiring the correct answer to be produced during generation. But the Abstract and Experiment 2 discussion say the benefit occurred only for semantically related cue–target pairs, not unrelated pairs. Those findings do not support saying the paper is inconsistent with a proposal requiring an existing cue–target relationship. (LLM, conf=0.86)
- Not checked — no full text of the sourcecites Hays et al., 2013 source
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- Disputed
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“The prevailing view is that a benefit of errorful generation only occurs when there is a pre-existing semantic association between cue and target.”
- Also stated in: § Unsuccessful retrieval · plain
“The prevailing view, then, is that errors can only be beneficial when there is a preexisting semantic relationship between cue and target.”
§ The benefit of generating errors during learning > Abstract
What the sources say
- Partly supportedcites Grimaldi & Karpicke, 2012 sourceExperimental — accuracy not yet measured
The General discussion (page unavailable) reports that retrieval attempts enhanced learning for related cue–target words but not unrelated ones across three experiments. It does not establish this as the prevailing view in the literature; the Experiment 1 discussion describes other theories that predict benefits for both related and unrelated pairs. (LLM, conf=0.85)
- Partly supportedcites Huelser & Metcalfe, 2012 sourceExperimental — accuracy not yet measured
The Results report that error generation improved retention for semantically related cue–target pairs but not for completely unrelated pairs. However, the introductory passage says the conditions under which errors help learning were not yet known, and the Abstract presents relatedness as the authors’ hypothesis rather than an established prevailing view. Our copy provides no page numbers for these passages. (LLM, conf=0.84)
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim.
- Also stated in: § Unsuccessful retrieval · plain
Every source the paper cites for this claim was checked and supports it.
In the paper
“Two subsequent studies using the same weak-associate paradigm (Grimaldi & Karpicke, 2012;Huelser & Metcalfe, 2012) came to the conclusion that a benefit of generating errors could only be observed when there was such a pre-existing semantic association between cue and target and not when the cue was unrelated to the target.”
§ The benefit of generating errors during learning > Unsuccessful retrieval
What the sources say
- Supportedcites Grimaldi & Karpicke, 2012 source
The Results and discussion section reports that attempting retrieval improved final recall only for related cue–target pairs, not unrelated pairs. The General discussion says this pattern occurred in all three experiments. The paper describes the initial retrieval attempts as producing frequent errors. (LLM, conf=0.92)
- Supportedcites Huelser & Metcalfe, 2012 source
The Results discussion states that generating errors enhanced retention for semantically related materials but not for completely unrelated materials. The General discussion also reports that Grimaldi and Karpicke found benefits only for semantically related items. (LLM, conf=0.94)
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“in Huelser and Metcalfe's (2012) study it was significantly more common for an error made at study to be repeated at test in the unrelated than the related condition, whether the design was between subjects (Experiment 1) or within subjects (Experiment 2).”
§ The benefit of generating errors during learning > Errorful generation
What the sources say
- Supportedcites Huelser and Metcalfe (2012) source
The Results report that in Experiment 1, initially incorrect responses appeared on the final test more often for unrelated than related materials (20% versus 5%, p < .001). The Experiment 2 error-persistence passage reports the same pattern (15% versus 7%, p < .01). The Abstract identifies Experiment 1 as between-participants and Experiment 2 as within-participants. (LLM, conf=0.95)
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Therefore, Grimaldi and Karpicke's proposal that errorful generation is only beneficial for related items (because they are covertly retrieved at study and thus benefit from a classic testing effect) and not for unrelated items (because they are not retrieved at study), may not be the only possible explanation for their findings.”
§ The benefit of generating errors during learning > Errorful generation
What the sources say
- Partly supportedcites Grimaldi and Karpicke's (2012) sourceExperimental — accuracy not yet measured
The General discussion reports that retrieval attempts enhanced learning of related targets but not unrelated targets, and explains this through activation of related candidates in a search set during the retrieval attempt. It does not propose that targets are covertly retrieved at study or that they benefit from a classic testing effect; that part of the attributed explanation is unsupported. (LLM, conf=0.88)
Not yet classified.
- Partly supported
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“Of course, we have not measured effort or attention directly but this proposal is consistent with other research suggesting that ease of processing leads to higher JOLs”
§ The benefit of generating errors during learning > Metacognition and errorful generation
What the sources say
- Supportedcites Castel et al., 2007 source
Experiment 2 links perceived ease of processing with higher judgments of learning (JOLs) for identical pairs, and the General Discussion says their fluency contributes to overestimated recall. The citing sentence’s statement that the authors did not measure effort or attention refers to their own study, not to this source. (LLM, conf=0.91)
- Supportedcites Koriat, 2008 source
The paper’s Experiment 5 discussion (page unavailable) states that easily learned items elicit higher judgments of learning (JOLs), supporting the point attributed to this source. The remark about not measuring effort or attention describes the citing authors’ own study, not Koriat’s. (LLM, conf=0.92)
- Not checked — no full text of the sourcecites Hertzog et al., 2003 source
Strength:hedged· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one. The quote only partly states this claim.
- Supported
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Experiments 2A, 2B, and 3 revealed, via metacognitive judgments of learning, that participants are strikingly unaware of this benefit, judging errorful generation to be a less effective encoding method than reading or incorrect choosing when in fact it was better.”
§ The benefit of generating errors during learning > Abstract
Load-bearingStrength:boosted· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“and an advantage for generating errors over reading, t(23) = 3.36, p = .003,d = .28.”
§ The benefit of generating errors during learning > Final test performance: Experiment 2B (Euskara words).
Not yet classified.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“final test performance differed between study methods, F(2,58) = 9.85, p < .001,p 2 = .25”
§ The benefit of generating errors during learning > Final test performance: Experiment 2A (English words).
Not yet classified.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“and also revealed a benefit of errorful generation over incorrect choosing.”
§ The benefit of generating errors during learning > Final test performance: Experiment 2B (Euskara words).
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesThe quote only partly states this claim.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Aggregate JOLs showed a largely similar pattern (Figure 3A).”
§ The benefit of generating errors during learning > Judgments of learning: Experiment 2A (English)
Strength:plain· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“generating nearly always produced an error, and the correct definition was available for much less timejust 7 s, compared with 17 s for Read trials.”
§ The benefit of generating errors during learning > Results
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Participants spent longer studying definitions for Read items (M = 8.02 s, SD = 4.35) than for both Generate items (M = 6.68,SD = 4.14), t(23) = 3.75, p = .001,d = .32,and Choice items (M = 5.35, SD = 3.90), t(23) = 9.25, p < .001,d = .65,and longer for Generate than for Choice items, t(23) = 4.41, p < .001,d = .33.”
§ The benefit of generating errors during learning > Study time
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“and there was no difference between Read and Choice JOLs, t(23) = .28,p = .783,d = .07.”
§ The benefit of generating errors during learning > Judgments of learning: Experiment 2B (Euskara)
Not yet classified.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“These high average JOLs given to Choice items were largely due to much higher JOLs for Choice items guessed correctly at study, while JOLs for items incorrect at study were no higher than JOLs for Read items”
§ The benefit of generating errors during learning > Discussion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In Experiment 2A the stimuli were obscure English words which were likely to be largely unknown to participants pre-experimentally, yet they showed the same benefit of generating over reading.”
§ The benefit of generating errors during learning > Discussion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- Rests on an assumption: In Experiment 2A, the obscure English words were likely to be largely unknown to participants before the experiment.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Participants' JOLs, however, showed a very different pattern, with the lowest item JOLs consistently being given to Generate items and the highest to Choice items.”
§ The benefit of generating errors during learning > General Discussion
Strength:plain· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In all four experiments reported here, we observed a benefit of errorful generation over reading when participants learned definitions of previously unfamiliar English words or translations of novel foreign vocabulary, even though the correct answer was displayed for a much shorter time in the Generate than in the Read condition.”
§ The benefit of generating errors during learning > General Discussion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Our second experiment replicated the benefit of generating over reading for unusual English words that we observed in Experiment 1 and showed that it extends to the learning of foreign vocabulary.”
§ The benefit of generating errors during learning > Discussion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Participants in Experiment 2B had no prior knowledge of Euskara, yet they showed better final test performance for items for which they had generated an incorrect guess than for items they had studied in the Read condition.”
§ The benefit of generating errors during learning > Discussion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- Rests on an assumption: Participants in Experiment 2B had no prior knowledge of Euskara.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Twenty-seven participants (17 in the SP group, 10 in the EP group) showed a benefit of generating over reading, with only five (2 in the SP group, 3 in the EP group) showing the opposite pattern.”
§ The benefit of generating errors during learning > Final test performance
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The EP group gave higher aggregate JOLs to the Read condition (M = 40.3,SD = 4.3) than to the Choice (M = 34.3,SD = 5.2) or Generate (M = 29.8,SD = 3.6) conditions but the difference between the study methods fell short of significance, F(2,30) = 2.78, p = .078,p 2 = .16.”
§ The benefit of generating errors during learning > Final test performance
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Finally, JOLs for Choice items guessed incorrectly at study were significantly lower than for Read items, t(23) = 2.38, p = .026,though with the exclusion of two participants who performed above chance at study (see footnote 2), this difference was no longer significant, t(21) = 1.98, p = .061”
§ The benefit of generating errors during learning > Judgments of learning: Experiment 2B (Euskara)
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For the SP group, the difference was significant, F(2, 46) = 5.80, p = .006,p 2 = .20.Aggregate JOLs for the Choice condition (M = 40.8,SD = 21.1) were higher than JOLs for the Generate condition (M = 26.8,SD = 17.6), t(23) = 3.46, p = .002,d = .72,and for the Read condition (M = 28.7,SD = 15.9),t(23) = 2.32, p = .030,d = .65,with no difference between Read and Generate aggregate JOLs, t(23) =.49,”
§ The benefit of generating errors during learning > Final test performance
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For Generate items, participants predicted poorest test performance while achieving highest performance.”
§ The benefit of generating errors during learning > Discussion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Participants' aggregate JOLs were also highest in the Choice condition, though memory performance showed no advantage of choosing over either reading or generating.”
§ The benefit of generating errors during learning > Discussion
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“JOLs for Choice items correct at study were significantly higher than JOLs for Choice items incorrect at study F(1,38) = 57.6,p < .001,p 2 = .60,with no difference between the groups, F(1,38) = 1.83, p = .185,p 2 = .05,and no interaction, F(1,38) = .02,p = .901,p 2 = 0 (Figure 3C).”
§ The benefit of generating errors during learning > Final test performance
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“There was no difference between the groups, F(1,38) = .37,p = .545,p 2 = .01,but there was an interaction between Study method and Group, F(2,76) = 3.81, p = .027,p 2 = .09.”
§ The benefit of generating errors during learning > Final test performance
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“we compared this with the average similarity of the targets to all of the other targets (M =.105, SD = .045),t(59) = 1.81, p = .076.”
§ The benefit of generating errors during learning > Discussion
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Participants' JOLs, however, showed a very different pattern from their actual test performance, with the lowest item JOLs being given to Generate items and the highest to Choice items.”
§ The benefit of generating errors during learning > Discussion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Choice JOLs were higher than Generate JOLs, t(39) = 3.25, p = .002,d = .55,with no difference between Generate and Read, t(39) = 1.78, p = .083,d = .32,or between Read and Choice, t(39) = 1.27, p = .211,d = .25.”
§ The benefit of generating errors during learning > Final test performance
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“JOLs for Choice items incorrect at study were higher than JOLs for Generate items, t(23) = 2.22, p = .037,but again this difference disappeared (t(21) = 1.89, p = .072)”
§ The benefit of generating errors during learning > Judgments of learning: Experiment 2B (Euskara)
Strength:plain· not yet calibratedn = 2Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Just as in Experiment 2, errorful generation produced better final test performance than either reading or incorrect choosing but participants failed to predict this benefit, giving the lowest JOLs to Generate items.”
§ The benefit of generating errors during learning > Final test performance
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Participants spent the same amount of time studying items they would later get right at test (M = 7.58 s, SD = 3.77, for Read; M = 6.32,SD = 4.26 for Generate; M = 5.24, SD = 3.86 for Choice) as they did studying items they would later get wrong (M = 8.31, SD = 5.52, for Read; M = 6.13,SD = 3.54 for Generate; M = 4.81, SD = 3.50 for Choice).”
§ The benefit of generating errors during learning > Additional information
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“but now there was no difference between JOLs for Read items and those for Choice items incorrect at study, t(39) = .11,p = .910,d =”
§ The benefit of generating errors during learning > Additional information
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“This was also the case in Exp 3, for the SP group: There was no difference between the three study methods, F(2,46) = .43,p = .651.”
§ The benefit of generating errors during learning > Additional information
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“When, in Experiment 3, participants had the option, at final test, of selecting their own generated responses and their own chosen responses, they were more likely to pick a definition they had incorrectly selected at study in the Choice condition than one they had incorrectly generated in the Generate condition, which may have contributed to the advantage for generating over choosing.”
§ The benefit of generating errors during learning > General Discussion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Incorrect guessing in the Choice condition, however, was less effective, producing equivalent memory performance to reading.”
§ The benefit of generating errors during learning > General Discussion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Somewhat surprisingly, and unlike in Experiments 1 and 3, participants in Experiments 2A and 2B selected the correct definition/translation at study in the Choice condition at a rate higher than chance.”
§ The benefit of generating errors during learning > Conclusion
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“and did not differ between conditions, F(2,18) = .35,p = .710.”
§ The benefit of generating errors during learning > Additional information
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“and longer for Generate than for Choice JOLs (M = 3.40s, SD = 1.6), t(23) = 2.74, p = .012.”
§ The benefit of generating errors during learning > Additional information
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Mean response times were 3.51s (SD = 1.5) for the Read condition; 3.69s (SD = 2.0) for the Generate condition; and 3.59s (SD = 3.1) for the Choice condition.”
§ The benefit of generating errors during learning > Additional information
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“However, when participants made an error at test in the Choice condition, they chose their own original error at a rate no higher than chance, suggesting that incorrect choosing caused no particular detriment to memory.”
§ The benefit of generating errors during learning > General Discussion
Strength:hedged· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“and no interaction, F(2,40) = 1.46, p = .245.”
§ The benefit of generating errors during learning > Additional information
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“However, Choice JOLs for items incorrect at study were often higher (Exp 2B and Exp 3) than JOLs for Generate items, even though both involved making an error at study, suggesting that the making of an error was not the only factor affecting JOLs.”
§ The benefit of generating errors during learning > Metacognition and errorful generation
Load-bearingStrength:hedged· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Although incorrect choosing led to lower test performance than errorful generation, it never led to worse performance than reading.”
§ The benefit of generating errors during learning > General Discussion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“A 2 (Accuracy: correct vs. incorrect) x 3 (Study method) ANOVA showed a main effect of Study method, F(2,40) = 30.68,p < .001”
§ The benefit of generating errors during learning > Additional information
Strength:plain· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Overall these results demonstrate that, within each condition, participants showed some ability to predict their true likelihood of recalling each item, although this was not statistically significant in Experiment 2B.”
§ The benefit of generating errors during learning > Conclusion
Strength:boosted· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“but no effect of Accuracy, F(1,20) = .02,p = .884”
§ The benefit of generating errors during learning > Additional information
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The mean time to make JOLs was 2.54s (SD = .8)in the Read condition; 2.57s (SD = .6)in the Generate condition; and 2.49s (SD = .5)in the Choice condition.”
§ The benefit of generating errors during learning > Additional information
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In Experiment 2A there was no difference in JOL response times between the three study conditions, F(1.29, 37.37) = .22,p = .704(Greenhouse-Geisser correction applied).”
§ The benefit of generating errors during learning > Additional information
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“higher confidence ratings (what Dunlosky et al. called”
§ The benefit of generating errors during learning > Additional information
What the sources say
- Not checked — no full text of the sourcecites Dunlosky, Serra, Matvey and Rawson (2005) source
Strength:plain· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“In particular, Choice JOLs for items correct at study were always higher than JOLs for Read items, even when participants gave similar aggregate JOLs to Choice and Read items (as in Exp 2B and Exp 3), suggesting that success at study for Choice items gave participants high confidence that they would also remember those items at final test.”
§ The benefit of generating errors during learning > Metacognition and errorful generation
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- Rests on an assumption: JOLs measure participants' beliefs or predictions about their future memory performance.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In Experiment 1, the number of correct generations was largely driven by the responses of one participant who generated correct guesses for over half the items in the Generate condition and three quarters in the Choice condition.”
§ The benefit of generating errors during learning > Conclusion
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In a more educationally relevant learning scenario, we found that generating errors could be helpful to memory even during the learning of novel material but that participants were strikingly unaware of this benefit.”
§ The benefit of generating errors during learning > Conclusion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- Rests on an assumption: JOLs measure participants' beliefs or predictions about their future memory performance.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“As can be seen in Table 3, Choice items correct at study were indeed better remembered than Read items, in line with participants' JOLs.”
§ The benefit of generating errors during learning > Metacognition and errorful generation
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Participants spent longer making JOLs for Generate items (M = 3.66s, SD = 1.6) than they did for Read items (M = 3.27s, SD = 1.4), t(23) = 2.84, p = .009”
§ The benefit of generating errors during learning > Additional information
Not yet classified.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In conclusion, generating errors benefitted vocabulary learning even when the items to be learned had no pre-existing associations, but participants did not predict this benefit.”
§ The benefit of generating errors during learning > Conclusion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- Rests on an assumption: Participants had no pre-existing associations with the novel words or foreign vocabulary before the experiment.
- Rests on an assumption: JOLs measure participants' beliefs or predictions about their future memory performance.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Making incorrect choices was neither harmful nor helpful to memory compared with reading.”
§ The benefit of generating errors during learning > General Discussion
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In Experiment 2B, however, there was a significant difference in times to make JOLs between the three conditions, F(2,46) = 5.85, p = .005).”
§ The benefit of generating errors during learning > Additional information
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“If this is the case, this could limit the usefulness of testing in situations where errors are likely to be made, but this view is based on just a handful of recent studies, all of which have used artificial tasks and materials which are rather different from those likely to be encountered during real world learning, and it remains to be seen whether an errorful generation benefit could occur in a”
§ The benefit of generating errors during learning > Abstract
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim.
In the paper
“Forlano and Hoffman (1937) found that “telling” was better than “guessing” when schoolgirls learned Hebrew-English word pairs.”
§ The benefit of generating errors during learning > Errorful generation
What the sources say
- Not checked — no full text of the sourcecites Forlano and Hoffman (1937) source
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“Berlyne, Carey, Lazare, Parlow, and Tiberius (1968) obtained similar findings for intentional learning of Turkish-English word pairs in adults.However, they did find an advantage of generating under incidental learning conditions”
§ The benefit of generating errors during learning > Errorful generation
What the sources say
- Not checked — no full text of the sourcecites Berlyne, Carey, Lazare, Parlow, and Tiberius (1968) source
Strength:plain· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Thus, although there was no overall detriment to the Choice condition by comparison with the Read condition, there was some evidence that errors made at study interfered with final test performance.”
§ The benefit of generating errors during learning > Results
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“Each item appeared equally often in each condition across participants.”
§ The benefit of generating errors during learning > Materials.
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“With repeated testing, superior performance has even been observed (McDaniel, Wildman, & Anderson, 2012).”
§ The benefit of generating errors during learning > Errorful generation
What the sources say
- Not checked — no full text of the sourcecites McDaniel, Wildman, & Anderson, 2012 source
Strength:plain· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“However, any negative effect of interference seems to have been offset by a positive effect of selecting a definition from a choice of two, perhaps because this involved deeper processing than passively reading the word and its definition.”
§ The benefit of generating errors during learning > Results
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Participants' JOLs, then, were strikingly inaccurate: the Generate condition produced the highest recall scores but the lowest JOLs.”
§ The benefit of generating errors during learning > Judgments of learning: Experiment 2A (English)
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“Each subset was matched for average number of letters and syllables per word, and each subset contained the same number of nouns, verbs and adjectives.”
§ The benefit of generating errors during learning > Materials.
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
In the paper
“People typically believe that studying is more effective than testing for previously studied material, even though the converse is true (e.g., Roediger & Karpicke, 2006b).”
§ The benefit of generating errors during learning > Errorful generation
What the sources say
- Not matchedcites Roediger & Karpicke, 2006b
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
- Not matched
In the paper
“Each lure was matched with the true definition or translation for number of syllables and for approximate word frequency (Kucera & Francis, 1967).”
§ The benefit of generating errors during learning > Materials.
What the sources say
- Not checked — no full text of the sourcecites Kucera & Francis, 1967
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“The assumption of sphericity was not met, χ 2 (2) = 9.63, p = .008,so the Huynh-Feldt correction was applied.”
§ The benefit of generating errors during learning > Judgments of learning: Experiment 2A (English)
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“total trial time was equated for each of the three conditions.”
§ The benefit of generating errors during learning > Errorful generation
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Just as for the English version of the task, generating produced the highest final test scores but the lowest JOLs.”
§ The benefit of generating errors during learning > Judgments of learning: Experiment 2B (Euskara)
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“In Experiment 3 we equated familiarity of the lures between the three conditions.”
§ The benefit of generating errors during learning > Experiment 3
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Generating errors (which happens on almost every trial) leads to less fluency and lower JOLs, while reading leads to intermediate fluency and JOLs.”
§ The benefit of generating errors during learning > Judgments of learning: Experiment 2A (English)
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“If metacognitive beliefs influence study decisions, such as the decision as to how much effort needs to be applied to encoding a given item, participants may make more effort to learn definitions for items subjectively experienced as more difficult and this could be one way that even errorful generation potentiates subsequent encoding.”
§ The benefit of generating errors during learning > Discussion
Strength:hedged· not yet calibratedCausal:conditional· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“These results again suggest that participants were strongly influenced by their success or failure at study, and particularly by the fortuitous selection of a correct choice.”
§ The benefit of generating errors during learning > Judgments of learning: Experiment 2B (Euskara)
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Even when they made errors, participants believed they would learn better by the Choice than the Generate method, though in reality the reverse was true.”
§ The benefit of generating errors during learning > Discussion
Load-bearingStrength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- Rests on an assumption: JOLs measure participants' beliefs or predictions about their future memory performance.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Furthermore, the benefit of generating over reading in the SP group suggests that the superiority of the Generate condition was not due to displaced rehearsal of Generate items during study of Read items, since allowing participants to choose how long to study should also eliminate any need for displaced rehearsal.”
§ The benefit of generating errors during learning > Discussion
Load-bearingStrength:hedged· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Although they spent significantly more time studying Choice items that were incorrect at study than correct ones, they spent even longer on Generate items, suggesting they believed the Choice items would be more easily learned even when they had made an error.”
§ The benefit of generating errors during learning > Discussion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- Rests on an assumption: JOLs measure participants' beliefs or predictions about their future memory performance.
- Rests on an assumption: Study time reflects the amount of effort or attention participants devoted to encoding.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Our proposal is that higher JOLs in the Read and Choice conditions reflect an illusion of knowing which leads to less effort being applied to encoding, or to a less efficient encoding strategy, than for Generate items, where initial responses are always incorrect.”
§ The benefit of generating errors during learning > Discussion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“It is interesting that the advantage for generating over reading was observed even though participants spent longer on Read than Generate items, suggesting that there was something about the process of generating a guess which enabled more efficient processing of the correct answer.”
§ The benefit of generating errors during learning > Discussion
Strength:hedged· not yet calibratedCausal:direct· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“Although JOLs showed some ability to predict final test scores (Figure 4A), these data should be interpreted with caution since they may be affected by item selection effects.”
§ The benefit of generating errors during learning > Judgments of learning: Experiment 2A (English)
Strength:hedged· not yet calibratedCausal:correlational· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“Against this interpretation, however, study times for Read items include time spent processing the cue word, whereas in the Generate and Choice conditions this has already taken place before timing begins, a factor which may at least partly account for the longer study times for Read items.”
§ The benefit of generating errors during learning > Discussion
Strength:hedged· not yet calibratedCausal:conditional· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
In the paper
“This is consistent with other research showing that greater effort during study leads both to better recall and to lower JOLs (Zaromb, Karpicke, & Roediger, 2010).”
§ The benefit of generating errors during learning > Discussion
What the sources say
- Not checked — no full text of the sourcecites Zaromb, Karpicke, & Roediger, 2010 source
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesProbably not a substantial claim. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Therefore even when an incorrect choice is made, the answer, when it appears, is already familiar and may, as a result, be processed more fluently, leading to higher JOLs and shorter study times.”
§ The benefit of generating errors during learning > Discussion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Making an incorrect choice also leads to low fluency and JOLs, but correct choice -despite being fortuitous -leads to much greater fluency and JOLs.”
§ The benefit of generating errors during learning > Judgments of learning: Experiment 2A (English)
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Together with the finding that JOLs were also higher for Choice items participants got right at study than for those they got wrong, these results suggest that participants' experience of success in the Choice condition led them to perceive correctly guessed items as easier to learn, both devoting less time to their study and giving them higher JOLs.”
§ The benefit of generating errors during learning > Study time
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
In the paper
“The Read condition may even give rise to an illusion of knowing, or”
§ The benefit of generating errors during learning > Discussion
What the sources say
- Not checked — no full text of the sourcecites Fischhoff, 1977 source
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Generating errors therefore leads both to lower JOLs and to higher final test performance.”
§ The benefit of generating errors during learning > Discussion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Thus generation can benefit memory both by strengthening existing associations and by potentiating the encoding of feedback.”
§ The benefit of generating errors during learning > General Discussion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“This is consistent with our proposal that fluency at encoding influences participants' metacognitive judgments which in turn affect the degree of effort or attention applied to the processing of corrective feedback.”
§ The benefit of generating errors during learning > Metacognition and errorful generation
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In this case, the pattern of results we observed could be due to a combination of factors.”
§ The benefit of generating errors during learning > General Discussion
Strength:hedged· not yet calibratedCausal:conditional· not yet calibratedTests✓ Asserted: passes✗ Checkable: fails✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“In the case of one participant this left only one usable item so this participant's data were excluded.”
§ The benefit of generating errors during learning > Additional information
Strength:plain· not yet calibratedSingle caseTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“If participants apply more effort, or more effective encoding strategies, to items they find difficult to learn this could account, at least in part, for the surprising benefit of errorful generation.”
§ The benefit of generating errors during learning > Conclusion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✗ Based: failsProbably not a substantial claim. The paper may give no reason why it holds.
In the paper
“these results should be treated with caution since item selection artefacts make it impossible to draw any conclusions about them.”
§ The benefit of generating errors during learning > Additional information
What the sources say
- Not checked — no full text of the sourcecites Nelson, 1993 source
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A claim is checked only against the cited paper’s full text, never its abstract, and none is available for its sources.
In the paper
“This perception that Generate items were harder to learn than items studied under the other two methods may have led participants to apply more effort to the learning of corrective feedback for Generate items, consistent with research showing that studying difficult items requires greater cognitive effort which in turn enhances memory”
§ The benefit of generating errors during learning > Metacognition and errorful generation
What the sources say
Strength:hedged· not yet calibratedCausal:conditional· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“We recomputed all the subsequent analyses with these participants excluded and none of the conclusions was changed, except (as noted in the relevant Results section) for two comparisons involving JOLs for Choice items in Experiment 2B, which come into line with the findings of Experiment 2A when these two participants are excluded.”
§ The benefit of generating errors during learning > Conclusion
Strength:plain· not yet calibratedn = 2Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In our study, aggregate JOLs given at the end of the study phase may also have reflected participants' expectations rather than their actual experience, and these too showed that participants expected memory to be poorest for items studied in the Generate condition.”
§ The benefit of generating errors during learning > Metacognition and errorful generation
Strength:hedged· not yet calibratedCausal:conditional· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passesIt may state two separate claims as one.
- Rests on an assumption: JOLs measure participants' beliefs or predictions about their future memory performance.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“These data were not captured for the EP group.”
§ The benefit of generating errors during learning > Additional information
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“We reran the analysis with this participant excluded and the results were unaffected.”
§ The benefit of generating errors during learning > Conclusion
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesThe quote only partly states this claim.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Our finding that JOLs were heavily influenced by success at study is theoretically interesting because it suggests that not only are people's perceptions of their learning strongly affected by fluency of processing at encoding but that this fluency is itself influenced by the production of either a correct or incorrect answer immediately prior to encoding of corrective feedback.”
§ The benefit of generating errors during learning > Metacognition and errorful generation
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“A possible explanation for our findings is that participants' curiosity about the correct answer was aroused to a greater extent in the Generate than in the Choice condition, perhaps because searching for an answer in the Generate condition involves more active engagement and effort than simply selecting one in the Choice condition or perhaps because, in the Choice condition, the participant knows”
§ The benefit of generating errors during learning > General Discussion
Strength:hedged· not yet calibratedCausal:conditional· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
In the paper
“In our study, where participants had to learn completely novel material, they would have known that their generated guesses were almost certain to be wrong but they may still have been highly curious to learn the correct answer, and the discrepancy between their own response and the actual answer may have induced a similar sense of surprise and led to greater attention being applied to encoding”
§ The benefit of generating errors during learning > General Discussion
What the sources say
- Not matchedcites Rescorla & Wagner, 1972
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- Rests on an assumption: Participants knew their generated guesses were almost certain to be wrong.
- Not matched
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“This analysis yielded a similar pattern to that of the subset.”
§ The benefit of generating errors during learning > Additional information
Strength:plain· not yet calibratedn = 107Tests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“However, participants do not realise that the extra effort involved in processing Generate feedback will lead to better memory, so they still give low JOLs to these items.”
§ The benefit of generating errors during learning > Metacognition and errorful generation
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passesIt may state two separate claims as one.
- Rests on an assumption: JOLs measure participants' beliefs or predictions about their future memory performance.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“The assumption of sphericity was not met, χ 2 (2) = 11.10,p = .004,so the Huynh-Feldt correction was applied”
§ The benefit of generating errors during learning > Additional information
Strength:plain· not yet calibratedTests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passesProbably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
The paper relies on this without support.
Participants in Experiment 2B had no prior knowledge of Euskara.
“Participants in Experiment 2B had no prior knowledge of Euskara”
The paper relies on this without support.
In Experiment 2A, the obscure English words were likely to be largely unknown to participants before the experiment.
“In Experiment 2A the stimuli were obscure English words which were likely to be largely unknown to participants pre-experimentally”
The paper relies on this without support.
Participants had no pre-existing associations with the novel words or foreign vocabulary before the experiment.
“the stimulus materials were previously unknown to participants so there could be no pre-existing association to activate and reinforce”
The paper relies on this without support.
JOLs measure participants' beliefs or predictions about their future memory performance.
“Even when they made errors, participants believed they would learn better by the Choice than the Generate method, though in reality the reverse was true.”
The paper relies on this without support.
Study time reflects the amount of effort or attention participants devoted to encoding.
“The amount of time participants in the SP group allocated to studying correct definitions also reflected this misconception.”
The paper relies on this without support.
Test lures were equally familiar across the three study conditions.
“even though test lures were equally familiar across the three study conditions”
The paper relies on this without support.
Participants knew their generated guesses were almost certain to be wrong.
“they would have known that their generated guesses were almost certain to be wrong”
Statistics that don’t add up
Each reported p-value recomputed from its test statistic and degrees of freedom.
p-value fits an unstated one-tailed test: t(39) = 3.49, p < .001
MinorThe paper reports p < .001 here; recomputing from the test statistic and its degrees of freedom gives p = .001 (two-tailed). The reported value is what a one-tailed test gives, but the paper does not say near this result that a one-tailed test was used.
Reported p < .001 · recomputed p = .001
“t(39) = 3.49, p < .001”
p-value fits an unstated one-tailed test: t (23) = 3.64, p < .001
MinorThe paper reports p < .001 here; recomputing from the test statistic and its degrees of freedom gives p = .001 (two-tailed). The reported value is what a one-tailed test gives, but the paper does not say near this result that a one-tailed test was used.
Reported p < .001 · recomputed p = .001
“t (23) = 3.64, p < .001”
Inconsistent p-value: t(23) = 2.21, p = .046
MinorThe paper reports p = .046 here; recomputing from the test statistic and its degrees of freedom gives p = .037 (two-tailed). The check cannot tell whether the statistic or the p-value is the misprint.
Reported p = .046 · recomputed p = .037
“t(23) = 2.21, p = .046”
p-value fits an unstated one-tailed test: t(39) = 3.24, p < .002
MinorThe paper reports p < .002 here; recomputing from the test statistic and its degrees of freedom gives p = .002 (two-tailed). The reported value is what a one-tailed test gives, but the paper does not say near this result that a one-tailed test was used.
Reported p < .002 · recomputed p = .002
“t(39) = 3.24, p < .002”
Inconsistent p-value: t(105) = 2.45, p = .015
MinorThe paper reports p = .015 here; recomputing from the test statistic and its degrees of freedom gives p = .016 (two-tailed). The check cannot tell whether the statistic or the p-value is the misprint.
Reported p = .015 · recomputed p = .016
“t(105) = 2.45, p = .015”
Statistics recomputed
Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.
- Doesn’t add upt(23) = 2.21reported p = .046 · recomputed p = .037
“t(23) = 2.21, p = .046”
- Doesn’t add upt(39) = 3.49reported p < .001 · recomputed p = .001
“t(39) = 3.49, p < .001”
- Doesn’t add upt(23) = 3.64reported p < .001 · recomputed p = .001
“t (23) = 3.64, p < .001”
- Doesn’t add upt(39) = 3.24reported p < .002 · recomputed p = .002
“t(39) = 3.24, p < .002”
- Doesn’t add upt(105) = 2.45reported p = .015 · recomputed p = .016
“t(105) = 2.45, p = .015”
- Consistentt(23) = 1.6reported p = .133 · recomputed p = .123
“t(23) = 1.6, p = .133”
- ConsistentF(2, 46) = 7.62reported p = .001 · recomputed p = .001
“F(2,46) = 7.62, p = .001”
- Consistentt(23) = 3.65reported p = .001 · recomputed p = .001
“t(23) = 3.65, p = .001”
- Consistentt(23) = 2.8reported p = .01 · recomputed p = .010
“t(23) = 2.80, p = .010”
- Consistentt(23) = 0.77reported p = .447 · recomputed p = .449
“t(23) = .77,p = .447”
- ConsistentF(2, 46) = 3.33reported p = .044 · recomputed p = .045
“F (2,46) = 3.33, p = .044”
- Consistentt(23) = 2.4reported p = .025 · recomputed p = .025
“t(23) = 2.40, p = .025”
- Consistentt(23) = 1.91reported p = .069 · recomputed p = .069
“t(23) = 1.91, p = .069”
- Consistentt(23) = 0.01reported p = .99 · recomputed p = .992
“t(23) = .01,p = .990”
- Consistentt(15) = 3.44reported p = .004 · recomputed p = .004
“t(15) = 3.44, p = .004”
- Consistentt(17) = 0.78reported p = .447 · recomputed p = .446
“t(17) = .78,p = .447”
- Consistentt(15) = 5.31reported p < .001 · recomputed p < .001
“t(15) = 5.31, p < .001”
- Consistentt(29) = 2.76reported p = .01 · recomputed p = .010
“t(29) = 2.76, p = .010”
- ConsistentF(2, 58) = 9.85reported p < .001 · recomputed p < .001
“F(2,58) = 9.85, p < .001”
- Consistentt(29) = 4.27reported p < .001 · recomputed p < .001
“t(29) = 4.27, p < .001”
- Consistentt(29) = 3.62reported p = .001 · recomputed p = .001
“t(29) = 3.62, p = .001”
- Consistentt(29) = 0.54reported p = .596 · recomputed p = .593
“t (29) = .54,p = .596”
- ConsistentF(2, 58) = 10.01reported p < .001 · recomputed p < .001
“F(2,58) = 10.01,p < .001”
- Consistentt(29) = 4.28reported p < .001 · recomputed p < .001
“t(29) = 4.28, p < .001”
- Consistentt(29) = 3.91reported p = .001 · recomputed p < .001
“t(29) = 3.91, p = .001”
- Consistentt(29) = 0.9reported p = .373 · recomputed p = .376
“t(29) = .90,p = .373”
- Consistentt(23) = 1.38reported p = .18 · recomputed p = .181
“t(23) = 1.38, p = .180”
- ConsistentF(2, 46) = 6.05reported p = .005 · recomputed p = .005
“F(2,46) = 6.05, p = .005”
- Consistentt(23) = 3.36reported p = .003 · recomputed p = .003
“t(23) = 3.36, p = .003”
- Consistentt(23) = 1.84reported p = .079 · recomputed p = .079
“t (23) = 1.84, p = .079”
- Consistentt(23) = 1.68reported p = .106 · recomputed p = .106
“t(23) = 1.68, p = .106”
- ConsistentF(2, 46) = 6.81reported p = .003 · recomputed p = .003
“F(2,46) = 6.81, p = .003”
- Consistentt(23) = 2.77reported p = .011 · recomputed p = .011
“t(23) = 2.77, p = .011”
- Consistentt(23) = 0.14reported p = .892 · recomputed p = .890
“t(23) = .14,p = .892”
- Consistentt(16) = 2.59reported p = .02 · recomputed p = .020
“t(16) = 2.59, p = .020”
- ConsistentF(2, 58) = 20.73reported p < .001 · recomputed p < .001
“F(2,58) = 20.73,p < .001”
- Consistentt(29) = 3.71reported p = .001 · recomputed p < .001
“t(29) = 3.71, p = .001”
- Consistentt(29) = 6.37reported p < .001 · recomputed p < .001
“t(29) = 6.37, p < .001”
- Consistentt(29) = 2.59reported p = .015 · recomputed p = .015
“t(29) = 2.59, p = .015”
- Consistentχ²(2) = 9.63reported p = .008 · recomputed p = .008
“χ 2 (2) = 9.63, p = .008”
- ConsistentF(1.62, 46.99) = 7.78reported p = .002 · recomputed p = .002
“F(1.62,46.99) = 7.78, p = .002”
- Consistentt(29) = 2.58reported p = .015 · recomputed p = .015
“t(29) = 2.58, p = .015”
- Consistentt(29) = 3.49reported p = .002 · recomputed p = .002
“t(29) = 3.49, p = .002”
- Consistentt(29) = 0.85reported p = .4 · recomputed p = .402
“t(29) = .85,p = .400”
- Consistentt(29) = 7.02reported p < .001 · recomputed p < .001
“t(29) = 7.02, p < .001”
- Consistentt(29) = 6.62reported p < .001 · recomputed p < .001
“t(29) = 6.62, p < .001”
- Consistentt(29) = 0.77reported p = .45 · recomputed p = .448
“t(29) = .77,p = .450”
- Consistentt(29) = 1.74reported p = .093 · recomputed p = .092
“t(29) = 1.74, p = .093”
- ConsistentF(2, 46) = 12.6reported p < .001 · recomputed p < .001
“F(2,46) = 12.60, p < .001”
- Consistentt(23) = 3.69reported p = .001 · recomputed p = .001
“t(23) = 3.69, p = .001”
- Consistentt(23) = 4.31reported p < .001 · recomputed p < .001
“t(23) = 4.31, p < .001”
- Consistentt(23) = 1.38reported p = .181 · recomputed p = .181
“t(23) = 1.38, p = .181”
- ConsistentF(2, 46) = 5.09reported p = .01 · recomputed p = .010
“F(2,46) = 5.09, p = .010”
- Consistentt(23) = 3.21reported p = .004 · recomputed p = .004
“t(23) = 3.21, p = .004”
- Consistentt(23) = 0.28reported p = .783 · recomputed p = .782
“t(23) = .28,p = .783”
- Consistentt(23) = 2.73reported p = .012 · recomputed p = .012
“t(23) = 2.73, p = .012”
- Consistentt(23) = 5.52reported p < .001 · recomputed p < .001
“t(23) = 5.52, p < .001”
- Consistentt(23) = 4.36reported p < .001 · recomputed p < .001
“t(23) = 4.36, p < .001”
- Consistentt(23) = 2.38reported p = .026 · recomputed p = .026
“t(23) = 2.38, p = .026”
- Consistentt(21) = 1.98reported p = .061 · recomputed p = .061
“t(21) = 1.98, p = .061”
- Consistentt(23) = 2.22reported p = .037 · recomputed p = .037
“t(23) = 2.22, p = .037”
- Consistentt(21) = 1.89reported p = .072 · recomputed p = .073
“t(21) = 1.89, p = .072”
- Consistentt(59) = 1.81reported p = .076 · recomputed p = .075
“t(59) = 1.81, p = .076”
- Consistentt(39) = 1.73reported p = .091 · recomputed p = .092
“t(39) = 1.73, p = .091”
- ConsistentF(2, 76) = 9.17reported p < .001 · recomputed p < .001
“F(2,76) = 9.17, p < .001”
- ConsistentF(1, 38) = 0.27reported p = .605 · recomputed p = .606
“F(1,38) = .27,p = .605”
- ConsistentF(2, 76) = 1.57reported p = .216 · recomputed p = .215
“F(2,76) = 1.57, p = .216”
- Consistentt(39) = 5.22reported p < .001 · recomputed p < .001
“t(39) = 5.22, p < .001”
- Consistentt(39) = 2.49reported p = .017 · recomputed p = .017
“t(39) = 2.49, p = .017”
- Consistentt(39) = 1.75reported p = .089 · recomputed p = .088
“t(39) = 1.75, p = .089”
- ConsistentF(1.7, 64.44) = 7.35reported p = .002 · recomputed p = .002
“F(1.70, 64.44) = 7.35, p = .002”
- ConsistentF(1, 38) = 0.19reported p = .662 · recomputed p = .665
“F(1,38) = .19,p = .662”
- ConsistentF(1.7, 64.44) = 1.53reported p = .224 · recomputed p = .224
“F(1.70, 64.44) = 1.53, p = .224”
- Consistentt(39) = 1.36reported p = .181 · recomputed p = .182
“t(39) = 1.36, p = .181”
- Consistentt(39) = 2.08reported p = .044 · recomputed p = .044
“t(39) = 2.08, p = .044”
- Consistentt(39) = 2.21reported p = .033 · recomputed p = .033
“t(39) = 2.21, p = .033”
- Consistentt(39) = 0.27reported p = .786 · recomputed p = .789
“t(39) = .27,p = .786”
- ConsistentF(1, 38) = 0.53reported p = .469 · recomputed p = .471
“F(1,38) = .53,p = .469”
- ConsistentF(1.62, 61.54) = 3.78reported p = .037 · recomputed p = .037
“F(1.62, 61.54) = 3.78, p = .037”
- Consistentt(23) = 3.04reported p = .006 · recomputed p = .006
“t(23) = 3.04, p = .006”
- Consistentt(23) = 1.34reported p = .193 · recomputed p = .193
“t(23) = 1.34, p = .193”
- Consistentt(15) = 0.95reported p = .355 · recomputed p = .357
“t(15) = .95,p = .355”
- Consistentt(15) = 4.25reported p = .001 · recomputed p < .001
“t(15) = 4.25, p = .001”
- Consistentt(15) = 2.66reported p = .018 · recomputed p = .018
“t (15) = 2.66, p = .018”
- ConsistentF(1, 38) = 57.6reported p < .001 · recomputed p < .001
“F(1,38) = 57.6,p < .001”
- ConsistentF(1, 38) = 1.83reported p = .185 · recomputed p = .184
“F(1,38) = 1.83, p = .185”
- ConsistentF(1, 38) = 0.02reported p = .901 · recomputed p = .888
“F(1,38) = .02,p = .901”
- Consistentt(39) = 7.83reported p < .001 · recomputed p < .001
“t(39) = 7.83, p < .001”
- Consistentt(39) = 1.06reported p = .296 · recomputed p = .296
“t(39) = 1.06, p = .296”
- ConsistentF(2, 76) = 4.1reported p = .02 · recomputed p = .020
“F(2,76) = 4.10, p = .020”
- Consistentt(39) = 3.25reported p = .002 · recomputed p = .002
“t(39) = 3.25, p = .002”
- Consistentt(39) = 1.78reported p = .083 · recomputed p = .083
“t(39) = 1.78, p = .083”
- Consistentt(39) = 1.27reported p = .211 · recomputed p = .212
“t(39) = 1.27, p = .211”
- ConsistentF(1, 38) = 0.37reported p = .545 · recomputed p = .547
“F(1,38) = .37,p = .545”
- ConsistentF(2, 76) = 3.81reported p = .027 · recomputed p = .026
“F(2,76) = 3.81, p = .027”
- ConsistentF(2, 30) = 2.78reported p = .078 · recomputed p = .078
“F(2,30) = 2.78, p = .078”
- ConsistentF(2, 46) = 5.8reported p = .006 · recomputed p = .006
“F(2, 46) = 5.80, p = .006”
- Consistentt(23) = 3.46reported p = .002 · recomputed p = .002
“t(23) = 3.46, p = .002”
- Consistentt(23) = 2.32reported p = .03 · recomputed p = .030
“t(23) = 2.32, p = .030”
- Consistentt(23) = 0.49reported p = .632 · recomputed p = .629
“t(23) =.49, p = .632”
- ConsistentF(2, 46) = 35.44reported p < .001 · recomputed p < .001
“F(2, 46) = 35.44,p < .001”
- Consistentt(23) = 3.75reported p = .001 · recomputed p = .001
“t(23) = 3.75, p = .001”
- Consistentt(23) = 9.25reported p < .001 · recomputed p < .001
“t(23) = 9.25, p < .001”
- Consistentt(23) = 4.41reported p < .001 · recomputed p < .001
“t(23) = 4.41, p < .001”
- Consistentt(23) = 3.13reported p = .005 · recomputed p = .005
“t(23) = 3.13, p = .005”
- Consistentt(37) = 2.44reported p = .02 · recomputed p = .020
“t(37) = 2.44, p = .020”
- ConsistentF(2, 70) = 2.67reported p = .076 · recomputed p = .076
“F (2,70) = 2.67, p = .076”
- Consistentχ²(2) = 11.1reported p = .004 · recomputed p = .004
“χ 2 (2) = 11.10,p = .004”
- ConsistentF(1.69, 64.22) = 24.43reported p < .001 · recomputed p < .001
“F(1.69, 64.22) = 24.43,p < .001”
- Consistentt(39) = 6.95reported p < .001 · recomputed p < .001
“t(39) = 6.95, p < .001”
- Consistentt(39) = 4.75reported p < .001 · recomputed p < .001
“t(39) = 4.75, p < .001”
- ConsistentF(1.69, 64.22) = 3.76reported p = .035 · recomputed p = .035
“F(1.69, 64.22) = 3.76, p = .035”
- Consistentt(23) = 7.71reported p < .001 · recomputed p < .001
“t(23) = 7.71, p < .001”
- Consistentt(23) = 3.92reported p = .001 · recomputed p < .001
“t(23) = 3.92, p = .001”
- Consistentt(15) = 2.72reported p = .016 · recomputed p = .016
“t(15) = 2.72, p = .016”
- Consistentt(15) = 0.39reported p = .701 · recomputed p = .702
“t(15) = .39,p = .701”
- ConsistentF(1.62, 61.54) = 6.84reported p = .004 · recomputed p = .004
“F(1.62, 61.54) = 6.84, p = .004”
- Consistentt(39) = 3.05reported p = .004 · recomputed p = .004
“t(39) = 3.05, p = .004”
- Consistentt(39) = 0.11reported p = .91 · recomputed p = .913
“t(39) = .11,p = .910”
- ConsistentF(1.29, 37.37) = 0.22reported p = .704 · recomputed p = .704
“F(1.29, 37.37) = .22,p = .704”
- ConsistentF(2, 46) = 0.43reported p = .651 · recomputed p = .653
“F(2,46) = .43,p = .651”
- ConsistentF(2, 46) = 5.85reported p = .005 · recomputed p = .005
“F(2,46) = 5.85, p = .005”
- Consistentt(23) = 2.84reported p = .009 · recomputed p = .009
“t(23) = 2.84, p = .009”
- Consistentt(23) = 2.74reported p = .012 · recomputed p = .012
“t(23) = 2.74, p = .012”
- Consistentt(24) = 4.43reported p < .001 · recomputed p < .001
“t(24) = 4.43, p < .001”
- Consistentt(19) = 3.5reported p = .002 · recomputed p = .002
“t(19) = 3.50, p = .002”
- Consistentt(24) = 2.96reported p = .007 · recomputed p = .007
“t(24) = 2.96, p = .007”
- ConsistentF(2, 36) = 0.112reported p = .894 · recomputed p = .894
“F (2,36) = .112,p = .894”
- Consistentt(16) = 0.69reported p = .501 · recomputed p = .500
“t(16) = .69,p = .501”
- Consistentt(10) = 1.28reported p = .23 · recomputed p = .229
“t(10) = 1.28, p = .230”
- Consistentt(16) = 0.99reported p = .336 · recomputed p = .337
“t(16) = .99,p = .336”
- ConsistentF(2, 18) = 0.35reported p = .71 · recomputed p = .709
“F(2,18) = .35,p = .710”
- Consistentt(38) = 7.36reported p < .001 · recomputed p < .001
“t(38) = 7.36, p < .001”
- ConsistentF(2, 208) = 7.36reported p = .001 · recomputed p < .001
“F(2, 208) = 7.36, p = .001”
- ConsistentF(1, 104) = 2.72reported p = .102 · recomputed p = .102
“F(1,104) = 2.72, p = .102”
- ConsistentF(2, 208) = 2.32reported p = .101 · recomputed p = .101
“F(2,208) = 2.32, p = .101”
- Consistentt(105) = 4.18reported p < .001 · recomputed p < .001
“t(105) = 4.18, p < .001”
- Consistentt(105) = 1.2reported p = .234 · recomputed p = .233
“t(105) = 1.20, p = .234”
- ConsistentF(2, 40) = 30.68reported p < .001 · recomputed p < .001
“F(2,40) = 30.68,p < .001”
- ConsistentF(1, 20) = 0.02reported p = .884 · recomputed p = .889
“F(1,20) = .02,p = .884”
- ConsistentF(2, 40) = 1.46reported p = .245 · recomputed p = .244
“F(2,40) = 1.46, p = .245”
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Self-check results appear here once an independent check has confirmed them.
Citations
Cites
(1967) · Medical Entomology and Zoology · Cited by 7,532
Rescorla, R. A. (1972) · Medical Entomology and Zoology · Cited by 7,268
An introduction to latent semantic analysis
Landauer, T. K., Foltz, P. W., & Laham, D. (1998) · Discourse Processes · doi:10.1080/01638539809545028 · Cited by 4,880
Test-Enhanced Learning: Taking Memory Tests Improves Long-Term Retention
Roediger, H. L., & Karpicke, J. D. (2006) · Psychological Science · doi:10.1111/j.1467-9280.2006.01693.x · Cited by 2,867
Balota, D. A., Yap, M. J.-M., Hutchison, K. A., Cortese, M. J., Kessler, B., Loftis, B., Neely, J. H., Nelson, D. L., Simpson, G. B., & Rebecca Treiman, A. (2007) · Behavior Research Methods · doi:10.3758/bf03193014 · Cited by 2,758
The MRC Psycholinguistic Database
Coltheart, M. (1981) · The Quarterly Journal of Experimental Psychology Section A · doi:10.1080/14640748108400805 · Cited by 2,506
The power of testing memory: Basic research and implications for educational practice
Roediger, H. L., & Karpicke, J. D. (2006) · Perspectives on Psychological Science · doi:10.1111/j.1745-6916.2006.00012.x · Cited by 2,037
The Critical Importance of Retrieval for Learning
Karpicke, J. D., & Roediger, H. L. (2008) · Science · doi:10.1126/science.1152408 · Cited by 1,791
(1994) · The MIT Press eBooks · doi:10.7551/mitpress/4561.001.0001 · Cited by 1,770
The generation effect: delineation of a phenomenon
Slamecka, N. J., & Graf, P. (1978) · Journal of Experimental Psychology: Human Learning and Memory · doi:10.1037/0278-7393.4.6.592 · Cited by 1,659
Flagged first, then by citations
Research works printed but not linked: 3
- [9]9. Bjork, R. A. (1975). Retrieval as a memory modifier. In R. Solso (Ed.), Information processing and cognition: The Loyola Symposium (pp. 123-144). Hillsdale, NJ: Lawrence Erlbaum Associates.
- [67]67. Richland, l. E., Kornell, N., & Kao, L. S. (2009). The pretesting effect: do unsuccessful tests enhance learning? Journal of Experimental Psychology: Applied, 15(3), 243 -257.
- [71]71. Schwartz, B. L., Benjamin, A.S., & Bjork, R.A. (1997). The inferential and experiential bass of memory. Current Directions in Psychological Science, 6(5), 132 -137.
Cited by
Sorted by citations
Items 101–166 of 166
The role of prediction in second language vocabulary learning
Gambi, C. (2021) · Bilingual processing and acquisition · doi:10.1075/bpa.12.09gam · Cited by 4
Errorful learning of trivia questions and answers: The role of study time
Butowska, E., Hanczakowski, M., & Zawadzka, K. (2024) · Memory & Cognition · doi:10.3758/s13421-024-01608-6 · Cited by 2
“…to guess the response to this question, and only then seeing the correct answer as feedback, benefits memory for this feedback compared with reading the question together with its answer, with no preceding retrieval attempt (e.g., Kornell et al., 2009; Potts & Shanks, 2014; Yan et al., 2014).”
What is the next structure? Guessing enhances L2 syntactic learning in a syntactic priming task
Alzahrani, A. (2023) · Frontiers in Psychology · doi:10.3389/fpsyg.2023.1188344 · Cited by 3
“org Making a guess may have increased the speakers’ desire to know the correct answer and this in turn have led to deeper encoding of that answer (Potts and Shanks, 2014; Potts et al., 2019).”
BackgroundAcquiring an Accurate Mental Model of Human Learning
Pan, S. C., & Bjork, R. A. (2024) · Oxford University Press eBooks · doi:10.1093/oxfordhb/9780190917982.013.69 · Cited by 2
Examining the influence of list composition on the mnemonic benefit of errorful generation
DiMarco, D., Laursen, S. J., Churey, K. R., & Fiacconi, C. M. (2024) · Memory · doi:10.1080/09658211.2024.2413159 · Cited by 2
“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).”
InfluentialHow Do Young Children Respond to Mistakes?
Donaldson, M., Sun, X., Chimgee, A. E., & Hwang, M. (2025) · Journal of Research in Childhood Education · doi:10.1080/02568543.2025.2465582 · Cited by 2
The Effect of Forward Testing as a Function of Test Occasions and Study Material
Sohlberg, R., Olsson, F., & Gander, P. (2021) · Behavioral Sciences · doi:10.3390/bs11090114 · Cited by 2
González, G., Alcalá, J. A., Callejas-Aguilera, J. E., & Rosas, J. M. (2019) · Behavioural Processes · doi:10.1016/j.beproc.2019.103898 · Cited by 3
Performance Expectancies Moderate the Effectiveness of More or Less Generative Activities Over Time
Reinhard, M., Weißgerber, S. C., & Wenzel, K. (2019) · Frontiers in Psychology · doi:10.3389/fpsyg.2019.01623 · Cited by 2
“, Paas, 1992), as well as briefly with the correct answers after the generation, and feedback is important for the effectivity (e.g., Slamecka and Fevreiski, 1983; Pashler et al., 2005; Kang et al., 2007; Metcalfe and Kornell, 2007; Potts and Shanks, 2014; Metcalfe, 2017).”
BackgroundBarenberg, J., & Dutke, S. (2021) · Unterrichtswissenschaft · doi:10.1007/s42010-021-00138-3 · Cited by 5
Use of virtual medical cases as a learning tool in medicine
Bureš, V., Čech, P., Husáková, M., Kokstein, Z., Kříž, P., Krylová, K., Nacházel, T., Ponce, D., Poulová, P., Ryška, A., et al. (2020) · Interactive Learning Environments · doi:10.1080/10494820.2020.1802297 · Cited by 2
“On the other hand, errors in performance occurring during learning are necessary preconditions for e ffi cient learning and an aversion to errors may be counterproductive (Potts & Shanks, 2014).”
BackgroundSchema-based predictive eye movements support sequential memory encodingPreprint
Huang, J., Velarde, I., Ma, W. J., & Baldassano, C. A. (2022) · doi:10.31234/osf.io/gej7q · Cited by 2
The impact of guessing and retrieval strategies for learning phrasal verbs
Strong, B. (2023) · IRAL - International Review of Applied Linguistics in Language Teaching · doi:10.1515/iral-2023-0048 · Cited by 1
Interpolated pretesting can boost memory of related and distinct prose materials
Kliegl, O., & Bäuml, K. T. (2024) · Psychological Research · doi:10.1007/s00426-024-02042-8 · Cited by 2
“In contrast, the core assumption of attention accounts is that a pretest format leads to increased attention to the correct answer feedback presented afterwards, resulting in better recall of the correct response on the final test (e.g., Potts & Shanks, 2014 ).”
Li, M., & Wang, T. (2025) · The Asia-Pacific Education Researcher · doi:10.1007/s40299-025-00976-0 · Cited by 1
“Word Guessing requires learners to infer meanings without context, which some studies argue is less effective for vocabulary acquisition (Potts & Shanks, 2014).”
The pretesting effect under divided attention
Bartl, J., Kliegl, O., & Bäuml, K. T. (2025) · Psychological Research · doi:10.1007/s00426-025-02106-3 · Cited by 2
“One prominent explanation of the pretesting effect is the attentional account, which assumes that Stage 2 of the learning trial is primarily responsible for the pretesting effect (Potts & Shanks, 2014; Zawadzka & Hanczakowski, 2019).”
Yan, V. X., Rea, S. D., & Sana, F. (2025) · Learning and Individual Differences · doi:10.1016/j.lindif.2025.102683 · Cited by 2
The Effects of Structured Input and Output on the Acquisition of the Accusative Case in L2 German
Henry, N. (2025) · Instructed Second Language Acquisition · doi:10.3138/isla-2024-0003 · Cited by 1
Li, X., Bai, H., Feng, X., Liang, Z., Chen, S., & Zhao, Q. (2026) · Learning and Instruction · doi:10.1016/j.learninstruc.2026.102356 · Cited by 1
The Effect of Initial Accuracy on the Learning and Retention of Novel Words
Kuznia, A., & Conwell, E. (2026) · Cognitive Science · doi:10.1111/cogs.70201 · Cited by 1
González, G., Alcalá, J. A., Ogállar, P. M., Rosas, J. M., & Callejas-Aguilera, J. E. (2021) · Behavioural Processes · doi:10.1016/j.beproc.2021.104529 · Cited by 1
How does error correction occur during lexical learning?
Iwaki, N., Takahashi, I., & Kaneko, S. (2024) · Psychological Research · doi:10.1007/s00426-024-01937-w · Cited by 2
“Even when the semantic association between an error and the target is removed or lowered, the benefit of error generation appears robustly (Metcalfe & Huelser, 2020; Potts & Shanks, 2014; Potts et al., 2019).”
Bi, H., Yan, R., & Zare, S. (2025) · Language and Speech · doi:10.1177/00238309251338748 · Cited by 1
“The 2-month interval was determined based on established educational research protocols, which indicate that a substantial time span enables a more accurate assessment of long-term vocabulary retention and learning effects (Potts & Shanks, 2014; Schneider et al., 2002).”
The Pretesting Effect: Exploring the Impact of Feedback and Final Test Timing
Mera, Y., Dianova, N., & Marin-Garcia, E. (2025) · Journal of Cognition · doi:10.5334/joc.455 · Cited by 4
“In pretesting, errors may reflect mere guesses, rather than genuine errors, as there is no prior exposure to the learning materials (Potts & Shanks, 2014).”
Learning and transfer of calorie information
Wohldmann, E. L., & Healy, A. F. (2020) · Applied Cognitive Psychology · doi:10.1002/acp.3727 · Cited by 1
Barenberg, J., & Dutke, S. (2022) · Psychologie in Bildung und Erziehung: vom Wissen zum Handeln · doi:10.1007/978-3-658-36260-7_4 · Cited by 1
Effects of Error Generation on Episodic Memory among Healthy Younger and Older Adults
Cyr, A.-A. (2015) · TSpace (University of Toronto) · Cited by 1
Pan, S. C., & Carpenter, S. K. (2023) · doi:10.31234/osf.io/9rqpm · Cited by 1
Attention and Mimicry in Minimal Groups
Blocker, H. S. (2016) · Digital Commons - DU (University of Denver)
The effect of generating errors on subsequent learning and generalization
Siler, J. (2017) · Illinois Digital Environment for Access to Learning and Scholarship (University of Illinois at Urbana-Champaign)
Memory in Educational Settings
Yan, V. X., & Wang, L. (2018) · doi:10.1093/obo/9780199828340-0230
Hotta, C., Kato, H., & Tajika, H. (2018) · Journal of Human Environmental Studies · doi:10.4189/shes.16.111
“…学習の仕方に対する気づきが得られた。第 3 に、ジグソー 学習後に導入した 24 名での集団のクロストーク段階にお いて、各メンバーの自由な関与が逆効果となり、ABCD の 4 組のうち、1 組において第 1、2 点目共に効果が認め られないことがわかった。 4.2 仮説検証と実践のポイント 本研究の目的は、ジグソー型の学習形態を基に、コミュ ニケーションに関する教材を学習する際に反復検索を取 り入れることで、メタ認知を促す授業を想定した。効果 測定として、第 1 に、どのように学習中に自己の認知活 動への気づきが促されたか、第 2 に実際に授業を通じて…”
InfluentialAssessing pedagogical approaches to foster the comprehension and retention of new L2 Spanish idioms
Tormo, C. U., Deconinck, J., & Stengers, H. (2021) · ELUA Estudios de Lingüística Universidad de Alicante · doi:10.14198/elua2021.35.15 · Cited by 1
“It is commonly believed that immediate corrective feedback will override original memory traces (Kornell, Hays & Bjork, 2009; Potts & Shanks, 2014) – a belief that is widely applied in teaching practice– but there is evidence that incorrect guesses or inferences can lead to erroneous memory traces…”
BackgroundEffects of L1 Single and Multiple-Choice Glosses in L2 Reading on Incidental Vocabulary Learning
Komuro, R. (2020) · ARELE: Annual Review of English Language Education in Japan · doi:10.20581/arele.31.0_129
Pan, S. C., Sana, F., Samani, J., Cooke, J., & Kim, J. (2020) · PsyArXiv (OSF Preprints) · doi:10.31234/osf.io/eqmhu
Burt, J. S., Leggett, J. M. I., Chalmers, K. A., & Boulton, P. A. (2021) · Memory · doi:10.1080/09658211.2021.1988645 · Cited by 1
Retrieval practice transfer effects for multielement event tripletsPreprint
Pickering, J. S., Henderson, L. M., & Horner, A. J. (2021) · PsyArXiv (OSF Preprints) · doi:10.31234/osf.io/54bgy
Strong, B. (2017) · Open Access Victoria University of Wellington | Te Herenga Waka (Figshare) · doi:10.26686/wgtn.17065166
“The learning of novel vocabulary represents a rather more realistic learning scenario for students learning an L2 (e.g., Potts & Shanks, 2014).”
BackgroundLexical inferencing as a generation effect for foreign language vocabulary learning
Dessenberger, S. J., Wang, K., Jordan, E., & Sommers, M. S. (2022) · Memory & Cognition · doi:10.3758/s13421-022-01348-5 · Cited by 1
“With multiple exposures, learners may be able to recognize word forms in each subsequent exposure which could prompt a covert retrieval of the memory of a previous exposure and in turn improve memory retention (Roediger & Karpicke, 2006).”
BackgroundPan, S. C., & Rivers, M. L. (2022) · doi:10.31234/osf.io/qh2ty
Visual Mapping and Guiding Framework for the Educational Success of Twice-Exceptional Learners
Lim, L. (2023) · Advances in educational technologies and instructional design book series · doi:10.4018/978-1-6684-6677-3.ch014
Strong, B. (2023) · SocArXiv (OSF Preprints) · doi:10.31235/osf.io/kqmw8
The effect of initial accuracy on the learning and retention of novel wordsPreprint
Kuznia, A., & Conwell, E. (2023) · doi:10.31234/osf.io/y84au
Briscoe, J., Doherty, J., Burgess, K. V., & Kent, C. D. (2023) · Neuropsychological Rehabilitation · doi:10.1080/09602011.2023.2259017 · Cited by 2
“The word-learning task involved three phases: an encoding phase, a distractor task, and a test phase for 60 Euskara-English word-pairs (from Experiment 2b of Potts & Shanks, 2014, and Experiment 5 of Seabrooke et al., 2019a).”
InfluentialErring on the Side of Caution: Two Failures to Replicate the Derring EffectPreprint
Mera, Y., Modirrousta-Galian, A., Thomas, G., et al. (2024) · doi:10.31234/osf.io/4u526
(2024) · Wiley eBooks · doi:10.1002/9781394308712.ref
Attention and the forward testing effect
Mulligan, N. W., & Buchin, Z. L. (2024) · Memory & Cognition · doi:10.3758/s13421-024-01661-1
“…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).”
Pretesting Promotes Learning but Not Generalisation from Moral Stories in Young ChildrenPreprint
Hsien, F. J., Yu, L., Pan, S. C., & Ding, X. P. (2025) · PsyArXiv (OSF Preprints) · doi:10.31234/osf.io/gdvsb_v1
Foo, Yu, Pan, and Ding (2025) Pretesting Promotes Learning from Moral StoriesPreprint
Hsien, F. J., Yu, L., Pan, S. C., & Ding, X. P. (2025) · PsyArXiv (OSF Preprints) · doi:10.31234/osf.io/gdvsb_v2
Learning from errors: deliberate errors enhance learning
Qiang, X., Ma, X., & Li, T. (2025) · Contemporary Educational Psychology · doi:10.1016/j.cedpsych.2025.102379 · Cited by 2
Effects of Practice Types on the Acquisition of English Phrasal Verbs
Feng, Y., & Yang, M. (2025) · Languages · doi:10.3390/languages10090214 · Cited by 1
Li, A. M., & Boers, F. (2025) · Modern Language Journal · doi:10.1111/modl.70007
“With some exceptions (Warm-ington & Hitch, 2014; Warmington et al., 2013), posttests tend to show better recollection of the meanings of new words after the guessing procedure, even though the guesses were inevitably often wrong (e.g., Potts & Shanks, 2014).”
Testing before learning: Exploring the robustness of the pretesting effect
Mera, Y., Dianova, N., & Marin-Garcia, E. (2025) · Memory & Cognition · doi:10.3758/s13421-025-01813-x · Cited by 4
“For example, in studies by Potts and Shanks (2014) and Seabrooke et al., (2019a, 2019b), participants were asked to estimate their likelihood of remembering word pairs.”
InfluentialTeaching Human Development Using Human Development
Blinkoff, E., Gibbs, H. M., Golinkoff, R. M., & Kathy, H.-P. (2024) · Emerald Group Publishing Limited eBooks · doi:10.1108/979-8-88730-489-220251003
Learning from failure is like learning from unsolicited advice
Fishbach, A. (2025) · Current Opinion in Psychology · doi:10.1016/j.copsyc.2025.102245
Pretesting effects for facts reflect lexical over semantic or structural features
Pan, S. C., & Siauw, F. A. (2026) · Memory & Cognition · doi:10.3758/s13421-025-01823-9
“…the context of pretesting with immediate correct answer feedback, it is not necessarily the case that attention is directed only towards the target (Potts & Shanks, 2014); there are evidently cases where pretest-ing also increases attention to cues as well (Experiments 5–7; see also Hays et al.,…”
InfluentialThe advantages of deliberate errors in promoting college students' memory retention and transfer
Qiang, X., & Ma, X. (2026) · British Journal of Educational Psychology · doi:10.1111/bjep.70058
Puimège, E., Uchihara, T., & Oikawa, G. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.6118208
Duolingo-inspired pretesting with words and pictures improves vocabulary learning
Chua, T. J. E., & Pan, S. C. (2026) · Cognitive Research Principles and Implications · doi:10.1186/s41235-026-00708-y
“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).”
InfluentialKliegl, O. (2026) · Psychologische Rundschau · doi:10.1026/0033-3042/a000750
Generative Processing Strengthens Learning Across Errorful and Non-Errorful Contexts
Khin, T. H. K., & Pan, S. C. (2026) · The Journal of Experimental Education · doi:10.1080/00220973.2026.2636103
Dong, R., Indrajaya, A., Yin, L., & Pan, S. C. (2026) · Applied Cognitive Psychology · doi:10.1002/acp.70200
Li, K. W. Y., & Ni, Y. (2026) · TESOL Quarterly · doi:10.1002/tesq.70164
Learning from mistakes: Objective and subjective error magnitudes predict the testing effect
Shi, A., & Liu, X. L. (2026) · Psychonomic Bulletin & Review · doi:10.3758/s13423-026-03009-z · Cited by 1
“However, such evidence provides only limited support for the EDL account because these studies typically rely on a binary manipulation of error signals, contrasting conditions with salient errors against conditions with smaller or less explicit errors (Kornell et al., 2009; Potts & Shanks, 2014).”
Grüter, T., Zhu, Y. (., & Jackson, C. N. (2026) · Language Learning · doi:10.1111/lang.70061
Ritchie, M., & Wilbiks, J. M. P. (2026) · Journal of Cognition · doi:10.5334/joc.522
Metadata
- Authors
- Potts, R., & Shanks, D. R.
- Affiliations
- University College London (GB)
- Venue
- Journal of Experimental Psychology: General
- Published
- 2014
- Volume
- 143
- Issue
- 2
- Pages
- 644-667
- ISSN
- 0096-3445 (ISSN-L)
- 1939-2222
- OpenAlex
- W2032355206
- Cited by
- 166
- References
- 68
- Citations by year
- 2026: 14 · 2025: 18 · 2024: 14 · 2023: 17 · 2022: 13 · 2021: 22 · 2020: 16 · 2019: 12 · 2018: 15 · 2017: 9 · 2016: 5 · 2015: 6 · 2014: 5
- FWCI
- 2.91
- Type
- journal article
- Language
- en
- Open access
- green
- discovery.ucl.ac.uk
- Locations
- Journal of Experimental Psychology General (publishedVersion)
- PubMed (publishedVersion)
- UCL Discovery (University College London) (submittedVersion)
- Funders
- United Kingdom Economic and Social Research Council (ES/H036237/1)
- Topics
- Life Sciences › Neuroscience › Cognitive Neuroscience › Memory Processes and Influences
- Social Sciences › Psychology › Developmental and Educational Psychology › Second Language Acquisition and Learning
- Social Sciences › Psychology › Experimental and Cognitive Psychology › Visual and Cognitive Learning Processes
- Keywords
- generation effect, corrective feedback, memory encoding, vocabulary acquisition, foreign language vocabulary, metalinguistic awareness, perceived difficulty, metacognition
- MeSH terms
- Adult, Cues, Educational Measurement, Feedback, Female, Humans, Learning, Male, Memory, Mental Recall, Young Adult
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-09-28, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-09-28 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
