The monitoring of one's own learning progress is a key process in models of self-regulated learning and a key predictor of self-regulated learning and academic success. Judgments of learning (JOLs) are an established measure for assessing people's monitoring of learning and have been found to predict learners' subsequent performance as well as effort regulation. However, most studies have been conducted in laboratory settings, involving relatively artificial learning materials and low-stakes tests. We evaluate the predictive validity of JOLs for learning performance and effort regulation in an ecologically valid learning environment by requesting aggregate JOLs in an intelligent tutoring system. 90 German university students used an intelligent tutoring system that provided practice exercises for self-regulated preparation for a statistics exam over the course of a semester. Aggregate JOLs for each chapter of the statistics course were assessed once per week (279 assessments in total). Dependent variables were learning performance as well as absolute and relative learning effort for each chapter, derived from the intelligent tutoring system's log files. JOLs significantly predicted learning performance ( β = 0.20, p < .001) and effort regulation ( β absolute = −0.12, p < .001, β relative = −0.07, p = .002). The present research demonstrates that JOLs have predictive power in real-world learning. It thus bridges the gap between experimental cognitive research and applied educational research on metamemory and self-regulation. • Judgments of learning (JOLs) are predictions of one's own future performance. • Little is known about JOL accuracy in ecologically valid learning environments. • We examined JOL accuracy in an intelligent tutoring system used for exam preparation. • JOLs predicted effort regulation and performance during exam preparation.

Neural Correlates of Metamemory Judgments in Worry: An ERP Study.

Judgments of Learning (JOLs) are prospective metamemory judgments critical for intentional learning, and have been theorized to be, in part, supported by implicit memory processes. This study tested the boundaries of subliminal implicit memory’s influence on JOLs by presenting masked primes, antiprimes, and neutral stimuli either before encoding (Pre-Study) or before JOL elicitation (Pre-JOL) in 96 neurotypical undergraduates across the two presentation conditions (n = 48 each). We assessed JOL calibration, resolution, recall accuracy, and reaction times. Subliminal stimuli showed no effects (equivalence testing, p < .01, with an equivalence bound of ηp2 = .03), regardless of delay condition, though delayed JOLs showed higher resolution and recall (the delayed JOL effect). These null findings suggest JOLs are not significantly influenced by subliminal stimuli under the tested conditions (50 ms exposure, younger adults), highlighting potential limits to implicit memory’s role in metamemory.

Metacognitive judgments and decisionsinvolve uncertainty and rely on probabilistic cues. Prior research shows that people integrate multiple cues when making judgments of learning (JOLs). The present study examined whether metacognitive control decisions are influenced by multiple cues as well. In each of two experiments, participants studied 60 words varying on two cues (Experiment1: concreteness, emotionality; Experiment2: font format, word frequency). In Experiment1, all participants made restudy choices to maximize later recall, whereas in Experiment2, half made restudy choices, and the other half provided JOLs. Participants who made restudy choices restudied their selected items, and all participants completed a recall test at the end of the experiment. At the group level, both cues influenced restudy choices in Experiment1, but only one cue did so in Experiment2. Individual-level analyses of Experiment2 revealed that most participants used both cues, yet the direction of cue use differed across participants: Some participants more often selected items with cue values associated with lower JOLs, whereas others more often selected items with cue values associated with higher JOLs. Overall, effect sizes for cue effects on restudy choices were smaller than those for JOLs. These findings suggest that multiple cues guided metacognitive control decisions, but that cue integration and cue use were weaker and varied more across individuals than in metacognitive judgments. This pattern indicates that the alignment between monitoring and control is reduced by other factors influencing restudy choices.

The act of making judgments of learning (JOLs), which involves predicting one's likelihood of remembering information, can modify memory performance, a phenomenon known as JOL reactivity. This study investigates the conditions under which JOL reactivity occurs, focusing on the interplay between encoding processes and retrieval contexts. Across three experiments, participants generated target words either from their phonological or semantic cues or read them in a within-subjects design. For testing JOL reactivity, one group provided JOLs, whereas the other group made random ratings. Experiment 1 used a recognition test, Experiment 2 employed a cued-recall test, and Experiment 3 introduced a process-matched cued-recall test to examine how JOL reactivity may differ across test types. Across experiments, participants produced higher memory performance for the generate than read condition and for semantic processing than phonological processing. Moreover, participants who provided JOLs outperformed participants in the random-rating group, showing the overall JOL reactivity effect. However, the magnitude and scope of this effect depended on the retrieval type. The recognition test boosted memory performance across both phonological and semantic processing conditions. Cued recall selectively favored semantic processing over phonological processing. Finally, process-matched cued recall showed JOL reactivity emerged primarily for semantic retrieval cues. These findings suggest that while JOL reactivity generally enhances the quality of encoding, the type of retrieval test critically determines whether and to what extent these improvements in memory performance are observable. The results are discussed in light of the cue-strengthening hypothesis, enhanced relational processing, and elaborative processing. (PsycInfo Database Record (c) 2026 APA, all rights reserved).

This study investigates the hypothesis that the theoretical framework underlying the Self-Consistency Model of subjective confidence can be extended to feeling of knowing (FOK) judgments and to judgments of learning (JOLs). According to this model, when individuals make a binary choice, their confidence in the selected option derives from the internal consistency among cues that are sampled randomly from a shared pool - termed collective wisdomware. Consequently, confidence should increase with the popularity of the chosen option (the "consensuality effect") and should predict the likelihood that others will select the same option (the "replicability effect"). Applying this framework to FOK and JOL revealed that both judgments exhibit robust consensuality and replicability effects. Results for JOLs further indicated that items differ reliably in the extent to which they elicit replicable cues across different encounters, and that signs of shared collective wisdomware exist even among first graders. The implications of these findings for the characterization and development of collective wisdomware are discussed.

Design effects constitute an important boundary condition for numerous memory effects: The structure of the study lists moderates the influence of these phenomena on free-recall tests. Previous research has shown that judgment of learning (JOL) reactivity-the effects of eliciting JOLs on memory performance-is also moderated by list composition. More specifically, Kaya and Mulligan (2025) showed that JOL reactivity in free recall is more positive in mixed than in pure lists. In this study, we examined why JOL reactivity shows design effects by testing the predictions from two prominent accounts of design effects: item-specific/relational processing and rehearsal borrowing. In Experiments 1 and 2, we demonstrate that, when memory is assessed via cued recall (Experiment 1) and recognition (Experiment 2), JOL reactivity produces the same positive effect for both mixed and pure lists, and no design effect occurs. These findings are consistent with the item-specific/relational and inconsistent with the rehearsal-borrowing account. Experiments 4 and 5 tested the rehearsal-borrowing account more directly by examining the self-paced study times for the JOL and no-JOL trials in mixed versus pure lists. The results showed that study times were not affected by the requirement to make a JOL or the list type, a finding inconsistent with the rehearsal-borrowing account. Taken together, the present study provides evidence for the item-specific/relational processing and against the rehearsal borrowing as the mechanism underlying the design effects in JOL reactivity. (PsycInfo Database Record (c) 2026 APA, all rights reserved).

ABSTRACT Providing Judgments of Learning (JOLs) for cue-target pairs at study often improves recall for related pairs versus a silent reading control task, a pattern termed JOL reactivity. While reactivity patterns have been found consistently using a variety of stimuli and test types, they appear to only occur when participants are explicitly instructed to provide JOLs. Whether JOL reactivity can carry over to cue-target pairs when JOLs are not directly applied is unclear. We evaluated this possibility by comparing cued-recall between two study-test blocks. In two experiments, participants either provided JOLs on both blocks or only in the first block with instructions to provide covert JOLs on the second block. Although JOL reactivity was found with explicit JOLs in the first block, it did not carry over to covert JOLs in the second block (Experiment 1), nor did reactivity carry over when participants were frequently reminded to make covert JOLs (Experiment 2). Reactivity therefore appears to be a byproduct of overtly provided judgments.

ABSTRACT Metacognitive monitoring of cognitive processes is not always accurate. Metacognitive illusions occur when metacognitive judgments rely on invalid information or fail to rely on valid information. This study tested the effectiveness of different forms of feedback in mending metacognitive illusions in judgments of learning (JOLs). Across four experiments, participants completed three study-test cycles with JOLs in which they studied different word lists. Participants received feedback or no feedback after each cycle. In Experiments 1 and 2, cognitive feedback about recall performance and JOL for each item was provided. In Experiments 3 and 4, additional metacognitive feedback about metacognitive illusions during the task was provided. Results showed that cognitive feedback was not effective for mending the font size illusion (Experiments 1 and 2), the stability bias (Experiment 1), or the font format illusion (Experiment 2). Additional metacognitive feedback partially remedied the stability bias in Experiment 3, but this effect did not replicate in Experiment 4. Regardless of whether participants received feedback and what type it was, the font size illusion decreased across cycles when manipulated orthogonally to a valid cue (Experiments 1, 3, and 4). In conclusion, this study shows that neither cognitive nor metacognitive feedback remedy metacognitive illusions.

ABSTRACT Judgments of learning (JOLs) can improve memory for cue-target word pairs (i.e., JOL reactivity). Prior research suggests that this effect does not extend to text passage learning, especially when memory is assessed using short-answer tests (e.g., Ariel, R., Karpicke, J. D., Witherby, A. E., & Tauber, S. K. (2021). Do judgments of learning directly enhance learning of educational materials? Educational Psychology Review, 33(2), 693–712. https://doi.org/10.1007/s10648-020-09556-8). However, JOL reactivity is often moderated by test format and JOL phrasing, and recent findings suggest that reactivity effects are greater when recognition tests are used. We tested whether global and term-specific JOLs would improve text passage learning when memory was assessed using short-answer (Experiment 1) and multiple-choice tests (Experiments 2 and 3). Across test formats, global JOLs were non-reactive. However, term-specific JOLs produced negative reactivity on short-answer tests (Experiment 1) but positive reactivity on multiple-choice tests (Experiments 2 and 3). This positive reactivity effect was greater in Experiment 3 when term-specific JOLs used a target-present phrasing. Importantly, this effect was also observed relative to a restudy group, suggesting that positive JOL reactivity was not solely driven by increased exposure to the target information. Taken together, JOLs can improve text passage learning, but their effectiveness is linked to test format and JOL phrasing.

Judgements of learning (JOLs) are metacognitive evaluations of future memory for newly learned information (Fiacconi et al., 2020; Koriat, 1997). The cue utilization view of JOLs states that individuals use a variety of cues when predicting future memory performance (Koriat, 1997). Critically, however, the majority of research aimed at understanding how different types of cues influence individuals' JOLs has focused on immediate memory assessments based on individuals' in-the-moment experiences or has utilized very brief retention intervals and relied on the representation of previously studied material (Rhodes & Tauber, 2011). Importantly, individuals' assessments of new learning may also be coloured by information learned further in the past when it is similar to the current information. Using a letter set training procedure (Fiacconi et al., 2020), we manipulated the fluency of to-be-learned material to examine whether previous learning would influence JOLs for new material over a 24-hr time period. As hypothesized, our results showed that previous learning did impact individuals' metamemory predictions, as JOLs for distinct but similar items were indeed higher than those for novel dissimilar items both immediately following training and 24 hr later. (PsycInfo Database Record (c) 2026 APA, all rights reserved).

The Cross-Race Effect (CRE) refers to the repeated finding that individuals are better at recognizing faces belonging to members of their own racial or ethnic ingroup. While this effect has been widely studied on memory, fewer studies have explored whether judgments of learning (JOLs) are also sensitive to this pattern. In two experiments, Black and White participants made JOLs while studying high typicality same and other-race faces (Experiment 1) or a mix of high and low typicality faces (Experiment 2). We replicated the CRE on memory, as participants were better at recognizing and showed greater discriminability for same-race faces. Importantly, the CRE pattern extended to JOLs. However, this pattern was linked to participant ethnicity, as only Black participants' JOLs were sensitive to the CRE. For White participants, mean JOLs did not differ between same and other-race targets, regardless of typicality. Responses on racial attitude measures showed that White participants displayed stronger internal motivations to appear non-prejudiced; however, correlational and mediation analyses revealed that this JOL pattern was not driven by differences in racial attitudes or motivations. Thus, while JOLs can be sensitive to the CRE, this effect varies by participant ethnicity, underscoring the influence of social experience on metacognitive judgments.

Students worldwide struggle to learn in STEM subjects, highlighting the need for effective interventions to enhance learning in these areas. Evidence on whether judgments of learning (JOLs) combined with retrieval practice can support learning from text is mixed. Notably, previous studies have been carried out with adult learners. Thus, it remains unclear to what extent younger students in intermediate school tracks, who typically perform at low to medium levels, benefit from making JOLs in addition to engaging in retrieval practice. Moreover, a potential positive influence of JOLs and retrieval practice on the effectiveness of self-regulation has not yet been examined. Therefore, we investigated the impact of JOLs and retrieval practice on text comprehension, judgment accuracy, and improvements in text comprehension after restudying (as an indicator of self-regulation). The study involved N = 315 9th school students ( M age = 15.10) who made can-judgments as a special form of JOLs and had a 2 (JOLs vs. no JOLs) × 2 (retrieval vs. no retrieval practice) factorial design. The results showed no significant effects of retrieval practice, can-judgments, or their combination on text comprehension, JOL accuracy, and improvements in text comprehension. This means that can-judgements and retrieval practice, without feedback or discussion afterwards, do not support self-regulated learning from text.

Semi-open-ended questions (SOE) are commonly used in online foreign language assignments; however, the effects of online feedback on these question types have received limited research attention. To address this gap, the present study employed an instructor survey and an in-class experiment to investigate how different forms of immediate feedback to SOE questions influence learning outcomes and judgment of learning (JoL). Survey findings revealed that although most instructors had access to platforms enabling immediate feedback, few utilized this feature for open-ended questions. Moreover, a majority of instructors believed that the most effective approach was to provide both correct responses (CR) and elaborated feedback (EF). Contrary to these beliefs, our experimental findings demonstrated that CR alone—feedback that simply provides the correct answer—was more effective than more complex feedback conditions, including a combination of CR and EF. Students who received CR required the least time, achieved the highest learning gains and reported stronger JoL, while CR and EF together did not yield additional benefits despite requiring more time to process. These findings challenge the common assumption that “more feedback is always better” and highlight the efficiency and pedagogical value of CR feedback for SOE questions in online foreign language learning. The study offers practical implications for instructors designing feedback strategies in technology-mediated environments.

Social comparison information has an impact on self-evaluation, and previous research has explored the effect of this type of information on relative prediction. However, studies on the absolute prediction of one’s own learning state, which is necessary for effective learning, are lacking. This study examined how social comparison information affects judgments of learning (JOLs). The study involved two experiments, which in turn comprised two sessions. During the solo session, the participants completed a memory task in a noncompetitive context. In the competition session, the participants were asked to compete with an opponent who signed in just before them, and they were provided with either upward or downward comparison information. The results of Experiments 1 and 2 showed that participants who received downward comparison information predicted that they would win against their opponent, whereas participants who received upward comparison information predicted that they would lose to their opponent. By contrast, participants in the upward comparison condition made higher JOLs than those in the downward comparison condition. We discuss these findings in terms of the change in confidence and distortion in the conversion process between confidence and JOLs, and present implications for effective learning.

Large language models (LLMs) increasingly mimic human cognition in various language-based tasks. However, their capacity for metacognition—particularly in predicting memory performance—remains unexplored. Here, we introduce a cross-agent prediction model to assess whether ChatGPT-based LLMs align with human judgments of learning (JOL), a metacognitive measure where individuals predict their own future memory performance. We tested humans and LLMs on pairs of sentences, one of which was a garden-path sentence—a sentence that initially misleads the reader toward an incorrect interpretation before requiring reanalysis. By manipulating contextual fit (fitting vs. unfitting sentences), we probed how intrinsic cues (i.e., relatedness) affect both LLM and human JOL. Our results revealed that while human JOL reliably predicted actual memory performance, none of the tested LLMs (GPT-3.5-turbo, GPT-4-turbo, and GPT-4o) demonstrated comparable predictive accuracy. This discrepancy emerged regardless of whether sentences appeared in fitting or unfitting contexts. These findings indicate that, despite LLMs’ demonstrated capacity to model human cognition at the object-level, they struggle at the meta-level, failing to capture the variability in individual memory predictions. By identifying this shortcoming, our study underscores the need for further refinements in LLMs’ self-monitoring abilities, which could enhance their utility in educational settings, personalized learning, and human–AI interactions. Strengthening LLMs’ metacognitive performance may reduce the reliance on human oversight, paving the way for more autonomous and seamless integration of AI into tasks requiring deeper cognitive awareness.

Metacognitive monitoring plays a crucial role in self-regulated learning, as accurate monitoring enables effective control, which in turn impacts learning outcomes. Most studies on metacognitive monitoring have focused on learners' monitoring abilities when they are explicitly prompted to monitor. However, in real-world educational settings, learners are more often prompted to control their learning, such as deciding whether to allocate additional time to a learning target. The primary goal of this study was to investigate whether retrieval is engaged when learners are explicitly prompted to control their learning processes by making study decisions. To address this, three experiments were conducted. In Experiment 1, participants (N = 39) studied 70 Swahili-English word pairs in a learning task. Each trial displayed a word pair for 8 s, followed by a distractor task (a two-digit mental addition) and a study decision intervention (choose "Study Again" or "Next"). After learning, participants provided a global judgment of learning (JOL), estimating their overall recall accuracy. Finally, they completed a cued recall test (Swahili cue). Responses were scored for accuracy and analyzed alongside study decisions, study decision reaction time (RT), and metacognitive judgments. Reaction times (RTs) for study decisions correlated positively with test accuracy, global judgments of learning (JOLs), and judgments of confidence (JOCs), suggesting retrieval likely underlies these decisions. Experiment 2 (N = 74, between-subjects) compared memory performance and intervention response time between single-study, restudy, retrieval (explicit recall prompt), and study decision (study decision prompt) groups to have better control over study time and cognitive processes. Although no significant group differences in test accuracy emerged, the retrieval group took longer to respond than the study decision group. Within-subject analyses revealed similar recall accuracy patterns: participants recalled successfully retrieved or "no restudy" items better than failed-retrieval or "restudy" items, implying shared cognitive processes underlying retrieval and study decision interventions. Experiment 3 (N = 74, within-subject, three learning conditions: single-study, retrieval, and study decision) replicated these findings, with no condition effects on test accuracy but longer RT for retrieval than study decisions. The similar recall accuracy patterns between retrieval and study decision interventions further supported shared cognitive processes underlying both tasks. Self-reports across experiments confirmed retrieval engagement in both retrieval and study decision interventions. Collectively, the results suggest that retrieval likely supports study decisions but may occur less frequently or less deeply than under explicit monitoring prompts. Additionally, this study explored objective, online measures to detect retrieval-based metacognitive monitoring.

Recent studies established that engaging metacognitive monitoring via making judgments of learning (JOLs) can directly enhance young adults' recognition memory, a phenomenon termed the reactivity effect of JOLs. The present study explored the reactive influence of making JOLs on older adults' recognition memory and probed the potential age-related differences in this effect. In three experiments, participants were instructed to study four lists of words, with two lists studied with concurrent JOLs and the other two without, followed by a recognition test. The results provided strong evidence that making JOLs improves older adults' recognition performance (Experiments 1-3) through enhancing both recollection- and familiarity-based recognition (Experiment 3). But the positive reactivity effect on recognition memory for older adults was weaker than that for young adults (Experiments 2 and 3). To elucidate potential mechanisms underlying age-related differences in the reactivity effect, the present study also measured participants' learning engagement and cognitive abilities. The model results substantiated the mediating role of learning engagement, supporting the enhanced learning engagement theory, rather than the dual-task hypothesis, as an account for the reactivity effect on recognition memory. (PsycInfo Database Record (c) 2025 APA, all rights reserved).

The present study examined how making judgments of learning (JOLs) vs. not making judgments of learning (no-JOLs) influences item and sequential memory in collaborative contexts. According to the item-order hypothesis, making JOLs improves memory for specific items (i.e., item memory) but disrupts sequential memory where memory for temporal relationships between items is required. If JOLs do enhance item memory performance, the study predicts they may effectively eliminate collaborative inhibition through a compensatory enhancement mechanism. Specifically, the magnitude of JOL-induced memory improvement appears to be greater in collaborative groups than in nominal groups. This differential enhancement likely offsets the typical memory impairment caused by collaborative retrieval interference, resulting in statistically equivalent final performance between groups. Consequently, the collaborative inhibition effect may disappear under JOL conditions. This study employed a 2 (group: collaborative vs. nominal; between-subjects) × 2 (metamemory monitoring: with vs. without judgments of learning; within-subjects) × 2 (test type: recognition vs. sequential reconstruction; within-subjects) mixed factorial design. The findings indicated that making judgments of learning significantly enhanced item memory performance while having no noticeable effect on sequential memory. It suggests that the reactivity effect is only present in item memory. Additionally, it was found that both item recognition and sequential memory performance were lower in the collaborative group compared with the nominal group, highlighting the presence of collaborative inhibition. These results suggest that the reactivity effect and collaborative inhibition are two distinct memory phenomena that do not affect each other.

To uncover how people monitor their learning and memory, researchers often elicit learners' predictions of their future memory performance immediately after studying, so-called immediate judgments of learning (JOLs). Previous research on immediate JOL reactivity shows that soliciting immediate JOLs can change memory. It is not well-understood, however, under which conditions making JOLs improves, impairs, or does not affect memory and which processes contribute to immediate JOL reactivity. The present study provides a comprehensive meta-analysis of immediate JOL reactivity based on 344 effect sizes from 175 independent experiments reported in 49 records with a total sample size of 15,079 adults. Overall, making JOLs improved memory performance (g = 0.22, 95% confidence interval [0.17, 0.27]). Whereas a meta-analysis of a subset of effect sizes with word pairs yielded virtually identical results (g = 0.23), we also found that JOL reactivity differed considerably depending on word-pair relatedness: Related pairs revealed small-to-moderate positive reactivity (g = 0.44), whereas unrelated pairs showed small negative reactivity (g = -0.09). Moderator analyses indicated that the size and direction of JOL reactivity varied depending on experimental characteristics (e.g., study materials, memory tests, or manipulating making JOLs within or between participants). Moderator effects were only partially consistent with existing theoretical accounts. Results revealed indications of publication bias favoring positive reactivity over no reactivity or negative reactivity. This study attests that making JOLs can change memory and highlight the need for refined theories to explain when and why these effects occur. (PsycInfo Database Record (c) 2025 APA, all rights reserved).