Researchers administered a spatial reasoning test to students aged 12 to 14. The test assessed spatial visualisation, spatial orientation and mental rotation, with each answer scored from 0 to 3. Students achieved a mean score of 7.34, with a standard deviation of 1.76. Based on the study’s classification criteria, 74.74% of students were in the medium spatial reasoning category, 9.47% were in the high category and 15.79% were in the low category.

Beliefs and metacognitive awareness were measured separately through questionnaires. The belief instrument was based on three domains: beliefs about mathematics, beliefs about oneself as a mathematics learner, and beliefs about the social context of mathematics learning. The metacognitive questionnaire addressed knowledge and regulation, including planning, information management, monitoring, debugging and evaluation. Responses were converted to scores from 1 to 4 and averaged for each student.

Two mathematics-education experts assessed the instruments’ content validity. Item-level Pearson correlations ranged from 0.785 to 0.892 for the spatial reasoning test and from 0.375 to 0.937 for the belief and metacognitive questionnaires. Cronbach’s alpha values were 0.882 for the spatial test, 0.949 for the belief questionnaire and 0.843 for the metacognition questionnaire. A pilot study involving 30 students from comparable schools preceded data collection.

Questionnaire results showed high reported confidence in mathematics learning: 52.63% of students were classified as very confident and 46.32% as confident, while 1.05% were not very confident and none were classified as not confident. For metacognition, 74.74% of students were classified as reflective, 24.21% as strategic, 1.05% as conscious and none as tacit. These self-reported categories indicate that students generally perceived themselves as confident and able to monitor their thinking, although questionnaire responses may be affected by social desirability bias.

Multiple linear regression showed that the combined model was statistically significant, with F = 21.13 and p < 0.001. Beliefs were positively associated with spatial reasoning after accounting for metacognition (β = 0.103, p = 0.023), as was metacognitive awareness (β = 0.167, p = 0.006). The larger metacognition coefficient indicates stronger statistical predictive power in this sample, but neither result demonstrates that changing beliefs or metacognition would directly cause an improvement in spatial reasoning.

The reported explained variance differs slightly across sections of the article. The abstract gives R² = 0.315, or 31.5%, while the results and conclusion give R² = 0.314, or 31.4%. The results also report R = 0.561; this is the multiple correlation for the model, not a correlation coefficient between any single pair of variables. On the results-and-conclusion value, the model accounted for 31.4% of the observed variation in spatial reasoning, leaving 68.6% associated with factors not included in the analysis.

The study was framed within Education 5.0, an approach that places human cognitive, social and emotional development at the centre while using technology to support learning. The authors discuss visual and digital resources such as Augmented Reality, three-dimensional simulations and interactive geometry software. They also describe belief-strengthening activities, metacognitive reflection, self-assessment, peer evaluation and opportunities to revise work after feedback as possible elements of learning design. These are proposed applications of the findings, not interventions tested in this study.

Important limitations include the purposive sample of 95 students from a single public school, the correlational design and the modest amount of variance accounted for by the model. Potential influences omitted from the regression include working-memory capacity, spatial experiences outside school, teacher instruction and previous geometry achievement. The direction of the relationships may also be reciprocal: stronger spatial reasoning could support more positive beliefs and metacognitive reports rather than the reverse. The authors recommend replication with larger and more geographically diverse samples, experimental or quasi-experimental intervention studies, longitudinal research and mixed-methods investigations of the processes behind the statistical associations.

Subject of Research: Correlational analysis of mathematics beliefs, metacognitive awareness and spatial reasoning among Indonesian junior high school students

Article Title: How beliefs and metacognitive awareness influence junior high students’ spatial reasoning ability for education 5.0

Article References: Hafizatunnisa, Fauzan, A., Candra, O., & Harisman, Y. (2026). How beliefs and metacognitive awareness influence junior high students’ spatial reasoning ability for education 5.0. Discover Education, 5(1), Article 926. https://doi.org/10.1007/s44217-026-02008-5

Image Credits: AI Generated

DOI: 10.1007/s44217-026-02008-5

Keywords: spatial reasoning, metacognition, mathematics beliefs, junior high students, Indonesia, multiple regression, mathematics education, Education 5.0, beliefs, metacognitive, awareness, influence

The study, led by Sanaa N. Al-Haj Ali of the Department of Orthodontics and Pediatric Dentistry and Ra’fat I. Farah of the Department of Prosthetic Dental Sciences, moves beyond the familiar accuracy benchmarks that dominate AI evaluation in medical education. Accuracy alone, the authors argue, cannot characterize how AI models engage with assessment tasks. Workload — the cost of achieving a particular level of performance, emerging from the interaction between task requirements and the operator’s skills and perceptions — offers a complementary lens. In humans, working memory is limited in capacity and duration, and when task demands exceed this capacity, performance is compromised. An operator may maintain accurate performance at an unacceptably high mental cost, which is why performance and subjective workload must be measured independently. This dissociation makes workload a valuable evaluative dimension for AI systems, where models and students may achieve similar accuracy while experiencing vastly different processing demands.

To capture that dimension, the researchers turned to the National Aeronautics and Space Administration Task Load Index (NASA-TLX), the most widely used multidimensional subjective workload scale in human factors research. The original instrument comprises six subscales — Mental Demand, Physical Demand, Temporal Demand, Effort, Performance, and Frustration — each rated on a 21-point scale. For this study, the team selected three subscales tailored to the OSPDE context: mental demand as the primary indicator of cognitive processing, temporal demand given the timed, bell-ringer station format of the examinations, and frustration to probe whether AI models show any sensitivity to task-related affective load. Physical demand was excluded as inapplicable to both the examinations, which involve no psychomotor component, and to AI; performance was excluded to avoid overlap with measured accuracy; and effort was excluded because, in a task with no physical component, it reduces largely to mental effort. Subscales were analyzed individually rather than as a composite, consistent with the diagnostic approach recommended for identifying discrete workload sources.

The experimental setup was meticulous. Data were sourced from OSPDEs administered to twenty fourth-year dental students during a preclinical orthodontics and pediatric dentistry course in 2025. The question pool comprised seventy items across four formats: ten plain multiple-choice questions without visual support, and twenty short-answer questions, twenty multiple-choice questions, and twenty True/False questions, all supported by visuals such as clinical photographs, radiographs, illustrations, and tables. By clinical discipline, the pool included thirty-three pediatric dentistry items (47 percent) and thirty-seven orthodontics items (53 percent). Correct answers were validated against three standard reference texts, and all visuals passed a three-component quality checklist scored by a single faculty member whose intra-rater reliability exceeded Cohen’s kappa of 0.8.

Ten multimodal AI models were evaluated from three providers: four from Anthropic (Claude Sonnet 3.5, Claude Sonnet 4.5, Claude Haiku 4.5, and Claude Opus 4.6), three from OpenAI (GPT-5.2 Pro, GPT-5.1, and GPT-4o), and three from Google Gemini (Gemini 3 Pro, Gemini 2.5 Pro, and Gemini 2 Flash). All models were accessed via a paid subscription to POE between January 25 and March 5, 2026. A single investigator, blinded to study objectives and student performance data, inputted each question into the models via a new chat session to prevent context contamination, transcribing questions verbatim with visuals uploaded as JPEGs at 300 dpi using a zero-shot prompting approach. The team employed a two-step elicitation protocol: after providing its answer, the complete question–response pair was resubmitted to the model with a prompt requesting ratings on the modified NASA-TLX, each subscale rated on a seven-point scale from one (very low) to seven (very high). To assess intra-model response stability, each question was tested in three independent sessions, with final responses determined by majority rule and NASA-TLX ratings by median value across trials. Students completed the same modified NASA-TLX in print immediately following the examination, with a 100 percent response rate.

The accuracy results set the stage for the workload findings. Overall accuracy differed significantly across responders (chi-square = 95.309, p < .001): students achieved the highest score at 77.9 percent, followed by GPT-5.2 Pro at 74.3 percent and Gemini 3 Pro at 72.9 percent, while Claude Haiku 4.5 scored lowest at 44.3 percent. Only GPT-5.2 Pro, Gemini 3 Pro, Gemini 2.5 Pro, and Claude Opus 4.6 did not differ significantly from students. Performance gaps were largest on visual short-answer questions (chi-square = 44.676, p < .001), where several Claude models scored at or below 25 percent, and smallest on plain multiple-choice questions, where the difference was not statistically significant. Intra-model reliability, measured by intraclass correlation coefficients, ranged from 0.677 to 0.955 across all models and dimensions (all p < .001), with mental demand showing the highest consistency and frustration the most variability — a gradient the authors attribute to frustration capturing an emotional reaction that is inherently more state-dependent than ratings referencing concrete task properties.

The workload comparisons produced the study’s central contribution. Seven of ten models reported significantly higher mental demand than students, six reported significantly higher temporal demand, and nine reported significantly higher frustration. The greatest human–AI divergence occurred on visual short-answer questions, where nine of ten models reported significantly higher mental demand and all ten reported significantly higher temporal demand. Yet frustration told a different story on this format: only three models (Gemini 2.5 Pro, Gemini 3 Pro, and GPT-5.1) reported significantly higher frustration than students, compared with seven to nine models on other formats, despite visual short-answer questions producing the lowest AI accuracy. On plain multiple-choice questions, models performed comparably to students on both accuracy and mental demand, indicating that human–AI convergence is greatest when visual processing is absent and recognition-based options constrain response demands.

This dissociation between demand ratings and frustration on the hardest format is not paradoxical, the authors argue, once the cue sources underlying each dimension are distinguished. Mental and temporal demand are stimulus-facing judgments: their cues derive from observable task properties such as inferential steps, visual integration requirements, and time constraints. Frustration, by contrast, is a self-evaluative judgment that requires comparing output quality against some performance standard. In multiple-choice and True/False formats, fixed alternatives provide an implicit benchmark against which the model can gauge correctness; short-answer questions strip away this external scaffold entirely, leaving the model with impoverished evaluative cues. Koriat’s cue-utilization framework supports this interpretation: metacognitive judgments lose diagnostic sensitivity when available cues are restricted. Additionally, the higher processing demands on visual short-answer items may consume resources that would otherwise support self-evaluative reflection, analogous to Sweller’s finding that heavy cognitive load diminishes capacity for secondary processes. Hence, the models had ample cues to detect that the task was difficult, but insufficient cues to evaluate how poorly they performed.

The authors also caution against anthropomorphic readings of the frustration data. The NASA-TLX frustration item asks how insecure, discouraged, irritated, or stressed the respondent felt — wording that presupposes an affective state current models do not possess. A model’s response is therefore generated text about frustration rather than a report of an experienced emotion, and the researchers treat the affective subscale strictly as a behavioral output rather than as evidence of internal experience. A parallel caveat applies to temporal demand: language models generate output without an internal clock or representation of elapsed time, so the instruction that two minutes were available was processed as semantic content rather than experienced as time pressure. Elevated temporal demand ratings are therefore more plausibly an inference from task characteristics — item length, information density, reasoning steps, and visual parsing load — than a report of temporal stress. This mirrors the broader pattern: models tracked content-derived indices of difficulty while showing attenuated responses on the affective dimension, which carries no comparable task signal.

Correlational analyses between workload and accuracy revealed a predominantly inverse relationship among both students and AI models, though among students significant correlations were weak and confined to frustration (rs = −0.08, p < .01). Among AI models, the pattern was architecture-dependent: Claude Sonnet 4.5 and Claude Opus 4.6 showed significant negative correlations across all three dimensions overall, while GPT-4o, Claude Sonnet 3.5, and the Gemini models showed none. Claude Haiku 4.5 showed significance for temporal demand (rs = −0.35) and frustration (rs = −0.40), and GPT-5.2 Pro for temporal demand (rs = −0.30) and frustration (rs = −0.33). At the format level, significant negative correlations were most concentrated on visual multiple-choice and visual short-answer items and least on visual True/False items and plain multiple-choice questions, suggesting that the diagnostic sensitivity of AI workload ratings scales with format complexity. This variability has practical significance: models with significant negative load–accuracy associations and acceptable accuracy — notably Claude Opus 4.6 and GPT-5.2 Pro — could serve as automated item difficulty estimators, flagging disproportionately demanding questions before student administration during examination development.

The findings carry implications for educators designing assessments intended to differentiate human from AI performance. Because human–AI convergence was greatest on plain multiple-choice questions and diverged most on visually integrated open-ended formats, educators seeking to distinguish human from AI capability should favor visually integrated open-ended question types. The study also establishes workload profiling as a complementary dimension of AI evaluation that extends beyond accuracy, offering insight not only into whether AI can answer correctly but into how it represents the difficulty of doing so — and where that representation breaks down.

Several limitations warrant acknowledgment. AI answers were determined by majority rule across three sessions whereas students had a single attempt, an asymmetry that, for accuracy, can only maintain or improve an AI’s score relative to a single attempt — meaning the finding that students outperformed most models is conservative. The student sample comprised twenty participants from a single institution, representing the entire eligible cohort but limiting generalizability, and the seventy-item question pool was smaller than licensing examinations. Potential data contamination cannot be fully excluded, as the training corpora of commercial models are proprietary, though the authors note that the study’s central finding — a systematic cross-item pattern in which workload tracked question format and visual complexity with a consistent demand–affect dissociation — would not be expected from memorization of individual images. Whether AI workload ratings reflect genuine metacognitive self-monitoring or statistically patterned output remains open; the reliability and correlational data favor a structured interpretation but cannot establish introspective capacity. Future work, the authors suggest, should compare both parties under identical single-attempt conditions and explore whether format-specific time allocations reduce the human–AI gap on open-ended items.

Subject of Research: Science News

Article Title: NASA TLX workload profiles of multimodal artificial intelligence models and dental students during objective structured practical dental examinations

Article References: Al-Haj Ali, S. N., & Farah, R. I. (2026). NASA TLX workload profiles of multimodal artificial intelligence models and dental students during objective structured practical dental examinations. Discover Education, 5(1), Article 835. https://doi.org/10.1007/s44217-026-02116-2

Image Credits: AI Generated

DOI: 10.1007/s44217-026-02116-2

Keywords: artificial intelligence, NASA-TLX, workload, dental education, multimodal AI, large language models, metacognition, cognitive load, objective structured practical dental examinations, AI evaluation, NASA, profiles

The study, published in the journal Discover Artificial Intelligence, was conducted by Ece Leventoğlu and Burak Ayçiçek of Tokat Gaziosmanpaşa University. It focused on a construct called metacreativity, defined as the awareness and regulation of one’s own creative process. Rather than treating creativity as a spontaneous spark, the researchers framed it as something learners can deliberately monitor and steer, much like the planning, monitoring, and evaluation that define metacognition more broadly. In practice, metacreativity shows up when students understand problems more deeply, reorganize relationships among ideas, generate alternative strategies, and judge the quality of their emerging answers.

To test whether AI could scaffold that kind of thinking, the team ran a convergent parallel mixed-methods study in an undergraduate Philosophy of Education course. Fifty students participated in the quantitative phase: 22 in an experimental group from the Special Education Teaching program and 28 in a control group from the Psychological Counseling and Guidance program. Both groups studied the same philosophical topics with the same instructor over the same five weeks, but only the experimental group worked with generative AI tools. The researchers measured attitudes toward AI, perceived metacreativity, and academic achievement before and after the intervention, and they supplemented the numbers with weekly reflective journals and an open-ended questionnaire.

The AI toolkit was broad. Students worked with ChatGPT, Gamma, Claude, NotebookLM, Suno, and Perplexity, using them to generate visual metaphors, creative texts, songs, and alternative representations of philosophical concepts. Crucially, the pedagogical design positioned AI as a scaffold rather than a substitute for thinking. Students were explicitly instructed not to accept generated content automatically but to examine its relevance, accuracy, and originality before revising or incorporating it into their final products. The instructor introduced weekly objectives and facilitated discussion but never supplied ready-made answers, leaving students responsible for questioning, selecting, revising, and evaluating what the machines produced.

The quantitative results revealed a distinctive pattern. In the experimental group, positive attitudes toward AI rose significantly from an average of 3.18 to 3.69 on a five-point scale, while negative attitudes fell from 3.20 to 3.00, both small-to-medium effects. Four of the six metacreativity subdimensions improved significantly: Comprehending the Main Problem and Sub-Problems, Reconstructing Connections, Finding Solutions, and the Affective Dimension, with effect sizes ranging from medium to medium-to-large. Notably, two subdimensions, Developing Additional Strategies and Evaluation, showed no significant within-group change at all. The control group, meanwhile, remained statistically flat across every attitude and metacreativity measure.

Between-group comparisons sharpened the picture further. After controlling for pretest scores, the experimental group scored significantly higher than the control group on Reconstructing Connections, Evaluation, Finding Solutions, and the Affective Dimension, with partial eta-squared values as large as 0.226 for Reconstructing Connections, a large effect by conventional standards. Interestingly, the Evaluation subdimension showed a significant adjusted between-group difference even though the experimental group had not improved significantly within itself, a result the authors attribute partly to the groups’ different baseline levels. Academic achievement rose significantly in both groups, suggesting ordinary course learning contributed to the gains, but the adjusted posttest scores still favored the experimental group, with a medium effect size of Cohen’s d equal to 0.66.

The qualitative data traced a clear developmental arc across the five weeks. First-week journals were dominated by novelty and excitement, laced with hesitation; one student admitted lingering prejudice because AI presents everything in an overly easy and ready-made way. By the second week, confidence grew, and by the third, students described practical applications, using an AI tool to build a presentation and directing its visuals themselves. By the fourth week, reflections turned pedagogical, with students imagining how song-generating AI might serve musically inclined learners and mind-mapping tools might help visual thinkers in their own future classrooms. One student captured a conceptual shift neatly: I used to think AI harmed creativity, but I later realized that creativity and creation are not the same.

Yet the students did not become uncritical enthusiasts. In the open-ended questionnaire, they flagged reduced mental effort, dependence on AI, misinformation risks, privacy and security concerns, and possible creativity suppression as genuine dangers. This balanced stance matters for interpretation. The authors argue that the educational consequences of AI are not inherently positive or negative; they depend on how AI is integrated, what tasks students perform, and whether the design explicitly encourages reflection, monitoring, and critical evaluation of machine-generated output. When AI replaced rather than supported self-regulated thinking, prior research has warned of weakened metacognitive engagement, but this study suggests such risks are contingent on pedagogy, not baked into the technology.

The selective pattern of gains is perhaps the study’s most instructive finding. The dimensions that improved, problem framing, connection-making, solution generation, and emotional engagement, align closely with what iterative AI use naturally affords: repeated cycles of idea generation, comparison, revision, and reflection. The dimensions that lagged, strategic diversification and evaluative judgment, represent higher-order regulatory processes that likely demand more practice, explicit evaluation criteria, and sustained scaffolding than a five-week intervention can provide. The authors recommend that future AI-supported activities push preservice teachers to compare alternatives, revise AI-generated content, justify their decisions, and evaluate their final products, and they suggest prompt templates, reflective journals, and structured reflection questions as practical ways to make students’ reasoning visible.

The researchers also stress the limits of what can be concluded. The sample was small and drawn from a single public university in Türkiye using convenience sampling, the two groups came from different academic programs, and metacreativity was measured through self-report rather than direct behavioral evidence. There was no delayed posttest to check whether the gains persisted, and because the intervention bundled AI use with creative production, discussion, and active learning, the effects reflect the overall instructional design rather than AI in isolation. Still, within those boundaries, the study offers a concrete template for teacher education: AI works best not as an answer machine but as a thinking partner, embedded in reflective, metacognitively oriented designs that keep students firmly in charge of their own creative judgment.

Subject of Research: AI-supported instructional design and preservice teachers' metacreativity, attitudes, and achievement in teacher education

Article Title: AI-supported instructional design and preservice teachers’ perceived metacreativity: a mixed-methods study

Article References: Leventoğlu, E., & Ayçiçek, B. (2026). AI-supported instructional design and preservice teachers’ perceived metacreativity: a mixed-methods study. Discover Artificial Intelligence, 6(1), Article 1239. https://doi.org/10.1007/s44163-026-02319-4

Image Credits: AI Generated

DOI: 10.1007/s44163-026-02319-4

Keywords: artificial intelligence, metacreativity, teacher education, preservice teachers, instructional design, generative AI, metacognition, academic achievement, mixed methods, higher education, AI attitudes, creativity

The study set out to compare three tiers of factors that might shape how accurately students judge their own math performance. At the broadest level sat the Big Five personality traits—openness, conscientiousness, extraversion, agreeableness, and neuroticism—the stable dispositions that psychologists have long studied as predictors of academic outcomes. One tier down sat subject-specific person attributes, represented by individual math anxiety, the domain-bound dread that many students feel when confronted with numbers and equations. At the most contextual level sat task difficulty itself, a property of the learning materials rather than of the learner. The central question was which of these tiers would most reliably predict self-estimated performance, actual performance, and—crucially—the accuracy of the match between the two, a property researchers call calibration.

Calibration is one of the most consequential concepts in educational psychology, and its importance is hard to overstate. A student with poor calibration faces a double bind: those who are overconfident stop studying too soon, believing they have mastered material they have not, while those who are underconfident waste precious study time rehearsing what they already know and may experience corrosive anxiety about their competence. Decades of research, stretching back to classic work on the illusion of knowing and on confidence calibration in judgment under uncertainty, have shown that monitoring accuracy feeds directly into self-regulated learning. If your internal gauge of your own knowledge is broken, every downstream decision—what to study, when to stop, whether to attempt a harder course—rests on faulty data.

The experimental design was elegantly straightforward. College students were assigned to solve one of two sets of math problems: an easy set deliberately pitched significantly below college level, or a relatively difficult set that better matched the expectations of college coursework. Before attempting the problems, participants estimated how they expected to perform; afterward, they estimated how they believed they had performed. They then completed the Big Five personality inventory developed by John and Srivastava, along with a standardized math anxiety scale. This structure allowed the researchers to examine metacognitive judgments both prospectively and retrospectively, and to ask whether the same factors predicted the raw judgments, the actual scores, and the alignment between the two.

The headline finding was that specific factors were more consistent predictors of students’ calibration accuracy than general traits. In other words, knowing a student’s overall personality profile told you less about whether their confidence would track their actual performance than knowing how anxious they were about math in particular, or how hard the problems they faced actually were. This pattern echoes a growing recognition in the literature that domain-specific measures often outperform broad dispositional ones when the outcome of interest lives inside a particular academic subject. General traits may set a backdrop, but the action happens in the foreground of the specific domain.

That said, personality was not irrelevant. Neuroticism—the trait characterized by emotional volatility, worry, and susceptibility to negative affect—emerged as a significant predictor of metacognitive judgments and their accuracy. This makes theoretical sense, because neuroticism and math anxiety share overlapping features related to general anxiety. A student high in neuroticism tends to experience worry across many situations, and that pervasive apprehension appears to color how they estimate their mathematical performance. Yet the study found that subject-specific math anxiety was the more consistent predictor of the two. When the researchers pitted the broad disposition against its domain-specific cousin, the cousin won. The implication is that anxiety does its damage to metacognitive accuracy most powerfully when it is anchored to the material at hand, not merely when it floats freely as a general temperament.

The third tier of factors, task difficulty, also proved to be a stable predictor of ability judgments and calibration accuracy. This finding aligns with a long line of research showing that the properties of the materials themselves shape metacognitive monitoring. Students confronting genuinely difficult problems face a harder inferential task when judging their own performance: the cues they rely on—fluency, familiarity, ease of processing—are less trustworthy when the material is demanding. Conversely, easy problems can breed their own distortions, since smooth processing may generate confidence that outstrips what the performance actually warrants. The fact that difficulty predicted both the judgments students made and how well those judgments matched their scores underscores that metacognition is never a purely internal affair; it is a dialogue between the mind and the task.

The study’s framing draws on influential theoretical accounts of self-regulated learning, including models that place metacognitive monitoring at the intersection of cognition, motivation, and affect. Within such frameworks, the current results suggest a layered architecture: general personality traits contribute a unique but less reliable signal, domain-specific anxiety contributes a stronger and steadier one, and contextual task factors anchor the whole system to the immediate situation. For educators, this hierarchy carries practical weight. Interventions aimed at improving students’ self-assessment in mathematics may achieve more by addressing math anxiety directly and by calibrating students’ expectations to the actual difficulty of their materials than by attempting to reshape broad personality dispositions, which are famously resistant to change.

The research also speaks to a persistent concern in mathematics education. Math anxiety has been documented for decades as a force that consumes working memory resources, degrades performance under pressure, and steers capable students away from quantitative careers. The new findings extend this picture into the metacognitive realm: anxiety does not only impair the doing of math, it impairs the knowing of how well one has done. A student whose fear of mathematics systematically distorts their self-assessments is doubly disadvantaged, because poor calibration undermines the very self-regulation strategies—strategic study, judicious practice, informed course selection—that could otherwise help them improve.

The authors, writing in Current Psychology, note that factors with domain and contextual specificity had more reliable predictive effects on metacognitive assessment accuracy in math, although general personality traits also play a unique role. The work was supported by the Eagle Scholars Program for Undergraduate Research through the Drapeau Center for Undergraduate Research at the University of Southern Mississippi, and all procedures were approved by the university’s Institutional Review Board. As classrooms increasingly emphasize self-directed learning and students are asked to monitor their own progress, studies like this one offer a sharper map of where the errors in self-knowledge come from. The gauge on the dashboard matters—but so do the driver’s history with that particular road, and the road itself.

Subject of Research: How personality traits, math anxiety, and task difficulty affect the accuracy of college students' metacognitive self-estimates in mathematics

Article Title: The role of personality, anxiety, and task difficulty in math metacognition

Article References: Thigpen, T., Agler, L.-M. L., & Noguchi, K. (2026). The role of personality, anxiety, and task difficulty in math metacognition. Current Psychology, 45(17), Article 1438. https://doi.org/10.1007/s12144-026-10002-1

Image Credits: AI Generated

DOI: 10.1007/s12144-026-10002-1

Keywords: metacognition, calibration, math anxiety, Big Five personality, neuroticism, task difficulty, self-regulated learning, college students, mathematics education, self-estimates, Current Psychology, role

The story began when the teenager developed a progressively enlarging neck mass causing discomfort and airway compression. Magnetic resonance imaging showed a left lower neck multi-cystic lesion extending into the supraclavicular region and down into the mediastinum, findings read as consistent with a macrocystic lymphatic malformation. Local otolaryngologists attempted aspiration but obtained insufficient fluid to relieve the compression, prompting transfer out of state for specialized care at a vascular anomaly center. His past medical, family, and psychosocial history offered no red flags, with no known genetic syndromes and no prior malignancy. On the surface, everything pointed toward a benign vascular anomaly.

At the specialty center, evaluation confirmed left cervical cystic masses, the largest measuring seven by seven centimeters. The lesion was non-tender but firm, a detail that raised concern for intralesional hemorrhage. Then came the first cracks in the diagnostic narrative. Additional history revealed a forty-pound weight loss, vomiting, and early satiety over the preceding months, symptoms that are distinctly atypical for an isolated lymphatic malformation. On examination, the boy’s abdomen was markedly distended with a palpable mass spanning the left upper and lower quadrants. Full-body MRI and magnetic resonance angiography revealed a large multiloculated cystic abdominal mass, initially interpreted as additional sites of lymphatic malformation. The differential diagnosis included complex lymphatic anomaly, venous-lymphatic malformation, and germ cell tumor, with evolving findings progressively arguing against a vascular cause.

Interventional radiology was consulted for aspiration and possible sclerotherapy, the standard treatment for macrocystic lymphatic malformations. What the procedures yielded should have been a turning point. The abdominal fluid was brown-tinged and the cervical fluid was clear, both appearances atypical for lymph or chyle, the milky lymphatic fluid expected from a true lymphatic lesion. Sclerotherapy was attempted on the cervical lesion, but fluid analysis told a different story: a triglyceride level of just 13 milligrams per deciliter and a lymphocyte fraction of only 2 percent, values wholly inconsistent with lymphatic fluid. Further sclerotherapy attempts were discontinued. The treatment was failing not because the technique was flawed, but because the underlying diagnosis was wrong.

Multidisciplinary consensus led to surgical resection of the abdominal mass, a twelve-hour operation that removed 95 percent of a retroperitoneal tumor measuring 25 by 25 by 20 centimeters. The mass had both solid and cystic components and encased critical structures, including the aorta, inferior vena cava, kidneys, ureters, and iliac vessels. Pathology revealed a mixed germ cell tumor, approximately 98 percent immature teratoma, with additional mature teratomatous elements including squamous, neural, and gastrointestinal-type epithelium. Foci of embryonal carcinoma and areas consistent with yolk sac tumor were also identified. Tumor markers confirmed the diagnosis: alpha-fetoprotein was 1,240 nanograms per milliliter against a reference range of zero to eight, and human chorionic gonadotropin was 268 milli-international units per milliliter against a reference of zero to five. Following pathologic confirmation, the patient transitioned to oncology care and began standard bleomycin, etoposide, and cisplatin chemotherapy without dose modification or delay. Follow-up imaging showed regression of residual disease, tumor markers declined appropriately, and treatment was well tolerated without significant toxicity or interruptions.

The authors of the report argue that this case is a vivid illustration of diagnostic anchoring, the cognitive tendency to rely too heavily on initial information and interpret everything that follows in relation to that first reference point. Once the lymphatic malformation label was attached, subsequent findings were filtered through that lens. Confirmation bias compounded the problem: clinicians sought and favored information that supported the existing hypothesis while downplaying evidence that contradicted it. The firm texture of the mass, the systemic symptoms, the multifocal disease across non-contiguous anatomic compartments, the atypical aspirate characteristics, and the failure to respond to sclerotherapy were all signals that something did not fit. Each was initially absorbed into the vascular anomaly framework rather than triggering a fundamental reconsideration.

The report lays out explicit diagnostic checkpoints that should prompt clinicians to question a lymphatic malformation diagnosis. These include systemic symptoms such as weight loss or early satiety, which are atypical for isolated lymphatic disease; multifocal involvement across separate anatomic compartments; aspirate characteristics inconsistent with chyle or lymph; and lack of biochemical or clinical response to sclerotherapy. When an initial diagnosis fails to account for the full clinical picture, the authors contend, structured re-evaluation is essential to avoid diagnostic delay. In this case, the presence of gastrointestinal symptoms and dramatic weight loss deviated from the typical lymphatic malformation course, yet an atypical complex vascular anomaly with possible intestinal lymphangiectasia was initially considered before the team pivoted toward malignancy.

The stakes of such cognitive errors are far from abstract. Diagnostic errors account for approximately 75 percent of malpractice lawsuits and 38 percent of total payouts, and pediatricians report diagnostic errors affecting patient outcomes at least annually. Misdiagnosis remains a leading cause of avoidable harm and cost in medicine. Lymphatic malformations are typically diagnosed through history, examination, and imaging, but their radiologic features can be nonspecific and can be confused with teratomas, particularly when cystic morphology is interpreted without integrating systemic symptoms or treatment response. A limited number of prior reports describe teratomas mimicking vascular anomalies, but few address the diagnostic delay caused specifically by cognitive bias. Unlike cases diagnosed early through surgery, this case shows how diagnostic mimicry and specialty silos can delay recognition when no one actively re-evaluates the working diagnosis.

The authors emphasize that overcoming bias requires deliberate tools rather than good intentions alone. Metacognition, the practice of actively reflecting on one’s own clinical reasoning, allows clinicians to recognize when their thought processes may be drifting toward premature closure. Cognitive forcing strategies offer structured ways to break out of entrenched thinking: case-based discussions with peers, red team reviews inviting external expert critique, prospective hindsight that imagines future outcomes to anticipate pitfalls, the five whys technique for root-cause analysis, decision trees and flowcharts, deliberate consideration of contrary evidence, and simulation-based reviews. Clinicians should also avoid the pseudo-diagnostic effect, in which irrelevant data are used to shore up a diagnosis. The report suggests two questions worth asking whenever a diagnosis fails to explain the clinical course: What else could this be? And what doesn’t fit?

Ultimately, the case is a story of both failure and rescue. The initial reliance on typical referral pathways without questioning atypical features delayed definitive treatment, a limitation the authors acknowledge candidly. But clinical intuition, combined with a structured multidisciplinary approach drawing on radiology, interventional radiology, pediatric surgery, pathology, and oncology, broke through the anchoring. The authors argue that balancing intuitive and analytical thinking, while remaining vigilant for bias, is central to diagnostic excellence, and that medical education should emphasize debiasing strategies, guided reflection, and structured re-evaluation. Time pressures, sleep deprivation, and cognitive overload all work against consistent analytical reasoning, which is why they also point to technology such as clinical decision support systems and artificial intelligence tools as potential safeguards. Improving the diagnostic process, they conclude, is both a clinical and a moral imperative, and recognizing diagnostic mimicry while applying critical thinking throughout the clinical course can improve care in complex pediatric conditions such as mixed germ cell tumors and lymphatic malformations.

Subject of Research: Cognitive bias and diagnostic error in the diagnosis of a pediatric mixed germ cell tumor mimicking a lymphatic malformation

Article Title: Case Report: When bias clouds judgment in diagnosing a pediatric mixed germ cell tumor

Article References: Elsabbagh, E. M., Maloney, L., Kulungowski, A., Annam, A., Siegele, B., & Nakano, T. A. (2026). Case Report: When bias clouds judgment in diagnosing a pediatric mixed germ cell tumor. Frontiers in Cancer Control and Society, 4, Article 1638457. https://doi.org/10.3389/fcacs.2026.1638457

Image Credits: AI Generated

DOI: 10.3389/fcacs.2026.1638457

Keywords: cognitive bias, diagnostic error, anchoring bias, confirmation bias, pediatric germ cell tumor, immature teratoma, lymphatic malformation, diagnostic mimicry, interventional radiology, metacognition, pediatric oncology, multidisciplinary diagnosis

The framework builds on a biopsychosocial perspective but pushes it further than the familiar three-pillar model. Rather than treating biology, mind, and society as separate influences whose contributions can simply be added together, Neroni proposes that creativity arises from dynamic, interdependent relationships that shift across individuals, tasks, developmental stages, and historical moments. The unit of explanation, in this view, is not the isolated determinant but the evolving configuration of determinants. This distinction matters because the same ingredient can play radically different roles depending on who is using it, what stage of creative work they are in, and what technological or social environment surrounds them. The perspective draws on earlier systemic models, such as the influential account locating creativity in the relationship among person, domain, and field, and on sociocultural approaches that dissolved the boundary between mind and material world. What is new is the extension of this relational logic across all levels of analysis, from neural networks to algorithms.

The neuroscience section of the argument illustrates the problem vividly. Over the past two decades, creativity research has moved away from hunting for a discrete creative brain area toward network-based models. Attention has focused on interactions between the default mode network, which supports internally directed thought, episodic memory, and spontaneous associative processing, and the executive control network, which contributes to working memory, attentional regulation, and the evaluation of emerging ideas. A third system, the salience network, appears to coordinate shifts between internally and externally oriented processing. Studies of divergent thinking, in which people generate multiple unusual uses for everyday objects, have repeatedly found functional coupling between these networks, supporting the idea that creative thought depends on cooperation between idea generation and cognitive control.

Yet recent critiques caution against replacing one reductionism with another. As the framework notes, studies differ substantially in the tasks they use, how they score creativity, and the neural patterns they report, and individual differences in how people approach creative tasks are often overlooked. Evidence that the default mode and executive networks cooperate during divergent thinking does not establish a general neural system for creativity. A broader concern holds that creativity should not be treated as a single cognitive faculty with a dedicated neural substrate, because creative activities recruit very different combinations of cognitive, emotional, perceptual, motor, and domain-specific processes. Findings from lesion studies reinforce this picture: damage to anatomically distinct brain regions can alter fluency, flexibility, or originality, yet distinct sites may connect to common functional circuits, suggesting distributed systems rather than a creative command center. Genetic research tells a similar story. Associations between creative performance and dopaminergic gene variants exist, but creativity is a complex behavioral phenotype that cannot map directly onto individual genes; genetic influences operate indirectly through temperament, reward sensitivity, and neural plasticity, and their expression depends on developmental and environmental conditions.

Development itself dismantles the notion of a biologically fixed creative capacity. Lifespan research distinguishes creative potential, everyday creative behavior, domain-specific performance, and socially recognized achievement, each following different trajectories. Childhood offers neural plasticity and exploratory behavior that favor experimentation, but children lack accumulated domain knowledge, and whether their explorations yield outcomes judged creative depends on task, educational environment, materials, and evaluative criteria. Later adulthood may bring declines in processing speed alongside continued growth in expertise and semantic knowledge, changing the resource mix rather than simply raising or lowering creativity. Research on sex and gender drives the point home: recent meta-analyses find only minimal average gender differences in creative ability, and differences in variability cannot explain the much larger disparities in eminent creative achievement, which instead reflect educational opportunities, institutional barriers, professional networks, and systems of cultural recognition. The significance of any biological characteristic, the framework argues, depends on the developmental, social, and cultural conditions in which it operates.

Psychological research reveals the same conditionality. Executive functions such as working memory, inhibition, and cognitive flexibility can help maintain goals, suppress dominant responses, and shift perspectives, yet stronger executive control is not always beneficial. Control helps when a task demands evaluation, refinement, or adherence to constraints, while less constrained processing may better facilitate access to unusual associations during idea generation. Cognitive flexibility can restructure an ineffective representation, but constant shifting becomes counterproductive when creative work requires sustained elaboration. Metacognition offers a way to understand this coordination: creative activity requires not only generating ideas but assessing progress, selecting strategies, and deciding whether to continue exploring or change direction. Emotion follows the same pattern. Positive affect has been linked to broad attention and associative flexibility, while some negative states support persistence and analytical processing, but recent experimental work shows that cognitive reappraisal facilitated creativity particularly among individuals lower in openness to experience, demonstrating how an emotion-regulation strategy interacts with personality rather than acting universally.

Social and cultural environments do not merely facilitate or inhibit individual creativity; they help shape the possibilities people encounter, the forms of novelty they recognize, and the outcomes that acquire value. Ideas about originality, usefulness, authorship, and innovation vary across societies and historical periods, and recent reviews caution against treating Western and Eastern cultures as homogeneous categories. Multicultural experience has been associated with greater cognitive flexibility, but exposure alone is insufficient; its effects depend on openness, language competence, domain knowledge, and the quality of intercultural engagement. In education, the long-standing debate over whether schooling stifles creativity through standardization or provides the knowledge needed for creative development remains unresolved, and the framework argues the opposition between structure and freedom is too simple. Structure can limit exploration when rigid, but it also supplies the knowledge, strategies, and feedback needed to develop an idea. Autonomy encourages exploration but may be less useful when learners lack the metacognitive skills to evaluate and refine what they generate.

Organizational research adds further layers. Psychological safety, autonomy, and diversity can support creativity, but none guarantees it; actor-context relationships can be synergistic, compensatory, inhibitory, or characterized by diminishing returns. Conditions that increase individual idea generation do not necessarily improve collective innovation, because teams must coordinate expertise, manage disagreement, and select among alternatives. Digital environments complicate the picture further. Recommendation systems and platform architectures shape what information people encounter and which ideas become visible, sometimes expanding access to diverse material while reinforcing particular patterns of attention and popularity. Generative artificial intelligence makes technology’s role even more direct, participating in ideation, recombination, evaluation, and production rather than merely providing access to information.

The evidence on human-AI collaboration illustrates the framework’s central claim with striking clarity. One influential study found that access to generative AI improved evaluations of individual creative stories, particularly among less creative writers, but simultaneously increased similarity across outputs and reduced collective diversity. Another line of research compared large language models used as ghostwriters, taking the primary role in content generation, versus sounding boards providing feedback on human-generated material. Sounding boards improved outcomes for non-experts, while ghostwriting provided no comparable advantage and was detrimental to expert users, possibly by anchoring subsequent human contributions. A separate experiment found that participants produced less creative poetry when given an AI-generated poem to edit than when writing independently, but this disadvantage disappeared when the system supported iterative co-creation with human and AI alternately contributing, and the co-creative configuration preserved higher creative self-efficacy. A meta-analysis of human-AI collaboration found substantial variation in outcomes depending on the relative strengths of humans and machines and how their contributions were combined. The same technology, in other words, can benefit individual performance while constraining diversity at the collective level.

From this configurational account, the framework derives four mechanisms—constraint, affordance, regulation, and feedback or selection—that describe how relationships across levels develop over time. Constraints can limit exploration but also reduce an unmanageable search space and focus attention. Affordances are relational possibilities for action that emerge between a person and their environment, such as an AI system enabling rapid generation of alternatives whose usefulness depends on the user’s expertise and evaluative skills. Regulation encompasses metacognitive monitoring, cognitive control, motivation, and emotion at the individual level, and feedback, norms, and collaboration at the institutional level. Feedback and selection make creative activity recursive, as products are accepted or rejected and systems amplify some outputs while overlooking others. The framework yields testable propositions: the effect of any creativity-related factor should vary with conditions, factors supporting one phase of creative activity should not necessarily support another, the same intervention can have opposite effects at different levels, and the impact of opportunities or constraints should change across development. Rather than asking which factors universally increase creativity, researchers should now ask for whom, under which conditions, at which stage and level, and over what timescale particular factors support or constrain creative activity—a question that demands ecological, longitudinal, multilevel, and interdisciplinary research designs capable of following creative processes as they unfold in the real world.

Subject of Research: A biopsychosocial and sociotechnical framework explaining creativity as a configurational phenomenon across biological, psychological, cultural, and technological levels

Article Title: Creativity beyond the isolated mind: toward a biopsychosocial and sociotechnical perspective

Article References: Neroni, M. A. (2026). Creativity beyond the isolated mind: toward a biopsychosocial and sociotechnical perspective. Frontiers in Cognition, 5, Article 1893765. https://doi.org/10.3389/fcogn.2026.1893765

Image Credits: AI Generated

DOI: 10.3389/fcogn.2026.1893765

Keywords: creativity, neuroscience, biopsychosocial, generative AI, cognitive flexibility, default mode network, executive control, sociocultural, developmental psychology, organizational creativity, human-AI collaboration, metacognition

The project was led by Li-Chun Chang of Chang Gung University of Science and Technology, together with Hsiao-Fang Hsiung, Li-Ling Hung, and Li-Ling Liao, and it was explicitly designed as a feasibility study rather than a definitive trial of effectiveness. That distinction matters. The researchers wanted to know whether such a module could actually be delivered inside a real, crowded nursing curriculum: would students sign up, would they stay, and would they complete the work? The answer on all three counts was yes. Of 49 eligible first-year students in a two-year baccalaureate nursing program, 35 enrolled, a recruitment rate of 71.4 percent. Every single one of those 35 students completed the study, a 100 percent retention rate, and adherence to unit-level submissions reached 95.2 percent, with 100 of 105 expected submissions received. For anyone who has tried to run an educational intervention in a professional degree program, those numbers are remarkable.

The intellectual architecture of GenAI-GUIDE rests on three pillars: AI literacy, metacognition, and self-directed learning. Metacognition, the capacity to monitor and regulate one’s own thinking, is the crucial ingredient here. Rather than treating generative AI as a forbidden fruit or an unquestioned shortcut, the module asked students to plan how they would use it, monitor what it was actually contributing, and evaluate whether the output deserved a place in their evidence-based work. Self-directed learning theory contributed the idea that adult learners thrive when they set goals and assess their own progress, and AI literacy supplied the technical vocabulary: understanding what large language models can and cannot do, where their failure modes lie, and how the phrasing of a prompt shapes the quality of the response.

Structurally, the one-semester module alternated between AI-free and AI-supported stages, a deliberate rhythm that let students first grapple with evidence-based nursing tasks on their own and then compare that experience with AI-assisted attempts. The module imposed explicit boundaries on when AI could be used, required students to refine and iterate their prompts, built in risk awareness about the tools’ well-documented tendency to fabricate plausible-sounding but false content, and, perhaps most distinctively, required contribution disclosure. Students had to state, in percentage terms, how much of each completed task they believed came from themselves versus the machine. This forced accounting transformed a vague sense of ‘the AI helped’ into a concrete, debatable judgment that could be examined in class.

The evidence-based nursing context gave the module its clinical teeth. Students worked through the canonical steps of EBN, including constructing PICO questions, which frame a clinical query in terms of population, intervention, comparison, and outcome; searching the literature; and synthesizing the evidence they found. Before and after the semester, they completed validated measures of their beliefs about evidence-based nursing and of their self-directed learning capacity. They also reported their perceived human-versus-AI contribution percentages across the three EBN task types, and they submitted written reflections that the team analyzed inductively for recurring themes.

The quantitative results form a nuanced picture. Self-directed learning scores rose significantly from pretest to posttest, suggesting that a semester of structured, reflective AI engagement may have strengthened students’ sense of ownership over their own learning. Total scores for beliefs about evidence-based nursing, by contrast, did not change significantly, a reminder that attitudes toward a professional practice framework are slower to shift than study skills. On the contribution question, students’ perceived AI contribution differed across the three tasks, with the pattern suggesting that students saw the machine playing different roles in question building, searching, and synthesis, although no pairwise comparison survived the stringent Bonferroni correction for multiple testing. The authors are careful to note that these are within-group pre-post patterns in a single-group design, not evidence of intervention effects, and that the study did not directly measure EBN task performance or the actual quality of AI use.

The qualitative analysis is where the study becomes genuinely vivid. Six themes emerged from the student reflections, and together they read like a field guide to the psychology of human-AI collaboration. The first was teacher guidance: students repeatedly credited the instructor’s framing for keeping their AI use honest and purposeful. The second was contribution disclosure itself, which several students described as an uncomfortable but clarifying exercise, because putting a number on the AI’s share forced them to confront how much thinking they had actually outsourced. The third was prompt structure, with students discovering that vague questions yield vague answers and that iterative refinement of prompts was itself a learnable skill.

The fourth theme, reappraisal of AI-generated content, captures a subtle cognitive shift: students reported going back over chatbot output with new skepticism, checking claims against course materials and questioning answers they would once have accepted at face value. The fifth theme concerned tool-specific risk and verification needs, as students learned that different AI tools carry different failure profiles and that every output required independent verification against primary sources, a habit that maps directly onto the evidence-appraisal mindset evidence-based nursing tries to instill. The sixth theme, collective monitoring through group discussion, showed that metacognition need not be a solitary affair; when students compared AI outputs and disclosed contributions in groups, the class effectively became a distributed error-detection system, catching fabrications and weak reasoning that individual students had missed.

Why does this matter beyond one Taiwanese nursing program? Because the dilemma GenAI-GUIDE addresses is universal. Health professional education is built on the premise that learners must earn their competence through effortful practice, yet generative AI makes it trivially easy to bypass that effort. Outright bans are proving unenforceable, and unstructured permission risks producing graduates who can prompt fluently but reason poorly. The Taiwanese study offers a third path: embed AI inside the curriculum, but wrap it in metacognitive routines, explicit boundaries, and honest accounting, so that using the tool becomes itself an exercise in critical thinking. The significant gains in self-directed learning hint that this approach may do more than teach AI etiquette; it may cultivate the self-regulated habits that lifelong clinical learning demands.

The authors are appropriately measured about the limits. The study was small, single-group, and unregistered, conducted at one institution with students who all had prior nursing education experience, so the findings describe feasibility and preliminary learning responses rather than proven effectiveness. They call for future iterations that strengthen critical appraisal and clinical application, the two EBN domains the module touched least, and for controlled studies that can isolate the module’s effects with comparison groups and objective performance measures. Still, the implementation indicators all cleared their provisional thresholds, and the retention rate alone will make curriculum designers sit up. As generative AI continues its relentless advance into every professional field, this study suggests that the most important technology to teach may not be the model itself, but the metacognitive machinery of the student standing in front of it. The nurses of the near future will work alongside algorithms; the question this research begins to answer is whether their education can teach them to remain the ones in charge of the thinking.

Subject of Research: A mixed-methods feasibility study of a metacognitively guided generative AI teaching module in evidence-based nursing education

Article Title: GenAI-GUIDE: a mixed-methods feasibility study of metacognitively guided generative AI in evidence-based nursing education

Article References: Chang, L.-C., Hsiung, H.-F., Hung, L.-L., & Liao, L.-L. (2026). GenAI-GUIDE: a mixed-methods feasibility study of metacognitively guided generative AI in evidence-based nursing education. BMC Nursing. https://doi.org/10.1186/s12912-026-05383-4

Image Credits: AI Generated

DOI: 10.1186/s12912-026-05383-4

Keywords: generative artificial intelligence, nursing education, evidence-based nursing, metacognition, self-directed learning, AI literacy, feasibility study, mixed methods, prompt engineering, health professional education, Taiwan, BMC Nursing

The motivation for the work lies in a genuine measurement gap. Classic instruments of self-regulated learning, such as the Motivated Strategies for Learning Questionnaire and the Learning and Study Strategies Inventory, were built for classrooms where the learner, the teacher, and the textbook were the only relevant actors. They were never designed to probe what happens when a student must decide how much to trust a chatbot, when to verify its output, or how to integrate machine-generated material with their own reasoning. Recent reviews of ChatGPT’s impact on education have documented both promising gains, including improved creative problem-solving performance in experimental settings, and worrying risks such as cognitive offloading, in which students outsource thinking they should be doing themselves. Without a validated measure of AI-integrated learning strategies, researchers and educators could not distinguish adaptive, sophisticated AI use from passive dependency.

Aldalham chose gifted upper-secondary students as the target population for both theoretical and practical reasons. Gifted education has increasingly shifted from a static conception of fixed ability toward talent development models that emphasize evolving complexity and creative productivity, frameworks associated with scholars such as Dai, Renzulli, and Subotnik and colleagues. Within these models, high-ability learners are expected to self-regulate intensely, pursue independent inquiry, and push beyond standard curricula, making them an ideal group in which to observe the most advanced forms of AI-integrated strategy. The study also aligns with Saudi Arabia’s National Strategy for Data and AI, which positions artificial intelligence capability as a national priority, giving the research a clear policy context alongside its psychological one.

Methodologically, the study followed an exploratory mixed-methods scale-development design, a sequence widely recommended in contemporary guidelines for constructing psychological measures. Item generation began with a literature-based definition of the construct, refined through expert input and then subjected to quantitative content evaluation using content validity index procedures rooted in the classical work of Lawshe and later refinements by Polit, Beck, and Owen. Only after the item pool survived this content scrutiny did the study move to psychometric testing, which was conducted with a sample of 504 gifted students. Crucially, the sample was randomly split into two halves, one for exploratory analyses and one for confirmatory analyses, a practice that guards against the common pitfall of discovering a factor structure and then validating it in the very same data.

The exploratory phase combined traditional exploratory factor analysis with exploratory graph analysis, a newer psychometric network technique developed by Golino and Epskamp that estimates the number of underlying dimensions by examining the connectivity structure among items. Both approaches converged on the same conclusion: AI-integrated adaptive learning is not one thing but four related things. The first dimension, Autonomous AI-Driven Learning Agency, reflects the degree to which students take initiative in directing their own learning with AI, setting goals and pursuing inquiry rather than waiting for instruction. The second, Human-AI Metacognitive Calibration, captures the delicate skill of monitoring and adjusting trust in AI outputs, knowing when to accept a machine’s answer and when to challenge it, a theme that echoes the adaptive trust calibration literature of Okamura and Yamada.

The third and fourth dimensions extend the construct into higher cognitive territory. Creative-Adaptive Knowledge Innovation measures the extent to which students use AI as a springboard for original thinking, transforming and recombining machine-generated content into something genuinely new rather than merely reproducing it. AI-Augmented Cognitive Orchestration, a term borrowed conceptually from Dillenbourg’s work on classroom orchestration, describes the strategic coordination of cognitive resources, in which the learner allocates tasks between their own mind and the AI system to manage cognitive load and maximize understanding. Together, these four dimensions sketch a portrait of the AI-era learner as an active conductor of a hybrid human-machine cognitive system, not a passive consumer of generated answers.

The confirmatory phase provided strong statistical support for this architecture. Confirmatory factor analyses endorsed both a correlated first-order model, in which the four dimensions relate laterally, and a higher-order model, in which they load onto a single general factor of AI-integrated adaptive learning. Reliability estimates, including modern internal-consistency coefficients such as omega alongside traditional measures, were satisfactory for all four subscales, and a two-week test-retest assessment demonstrated that scores are stable over time, a key requirement for any instrument intended to track students across semesters or interventions. Convergent and discriminant validity indices, evaluated with procedures such as the Fornell-Larcker criterion, fell within acceptable ranges, indicating that the scale measures what it claims and does not merely duplicate existing constructs.

Two additional findings strengthen the scale’s practical credentials. First, tests of measurement invariance showed configural, metric, and scalar invariance across gender, meaning the instrument functions comparably for boys and girls and that score differences between groups can be interpreted substantively rather than as measurement artifacts, following the sensitivity framework articulated by Chen. Second, AI-ALSS scores correlated positively with self-reported grade point average, offering preliminary concurrent criterion-related evidence that students who deploy more sophisticated AI-integrated strategies also tend to achieve more academically. While a self-reported correlation cannot establish causation, it suggests the scale taps something educationally meaningful rather than a stylistic tendency to answer questionnaires favorably.

The implications reach well beyond gifted education in Saudi Arabia. For researchers, the AI-ALSS provides a continuous measure that can finally connect the self-regulated learning tradition, from Zimmerman through Pintrich to contemporary work by Järvelä and Hadwin, to the emerging science of human-AI collaboration championed by Shneiderman’s human-centered AI framework. For teachers and school leaders, the four dimensions offer a diagnostic vocabulary: a student may show strong creative innovation with AI but weak metacognitive calibration, blindly trusting outputs, while another may orchestrate tools expertly yet rarely initiate independent inquiry. Such profiles could guide targeted coaching instead of blanket policies that either ban or uncritically embrace AI. For policymakers navigating the tension between AI literacy goals and fears of cognitive offloading, the scale offers a way to monitor whether integration strategies are actually adaptive.

The author is careful to frame the results as initial rather than definitive. The validation was conducted in a single cultural context with a specific population, and the study itself calls for future work on predictive validity, sensitivity to intervention-related change, and cross-cultural replication. Whether the four-dimension structure holds for mainstream students, for younger learners, or in classrooms with different levels of AI access remains an open empirical question. Yet the study arrives at a moment when such tools are urgently needed, as generative AI reshapes homework, assessment, and the very definition of studying. By transforming a vague intuition, that students learn differently with AI, into a measurable, multidimensional construct with documented reliability and validity, the research gives the field something it has lacked: a ruler for the age of artificial intelligence. How educators, psychologists, and students wield that ruler may shape how the next generation learns to think alongside machines.

Subject of Research: Development and validation of a scale measuring AI-integrated adaptive learning strategies in gifted students

Article Title: Adaptive learning strategies in the age of artificial intelligence: development and validation of a scale for gifted students

Article References: Aldalham, M. S. (2026). Adaptive learning strategies in the age of artificial intelligence: development and validation of a scale for gifted students. Current Psychology, 45(20), Article 1606. https://doi.org/10.1007/s12144-026-10137-1

Image Credits: AI Generated

DOI: 10.1007/s12144-026-10137-1

Keywords: artificial intelligence, adaptive learning, gifted students, scale development, self-regulated learning, metacognition, generative AI, psychometrics, education, Saudi Arabia, cognitive offloading, measurement invariance

The study, conducted by Meysam Muhammadpour of the University of Guilan and published in BMC Psychology, set out to answer a question that has long divided researchers in second language acquisition: is listening comprehension limited primarily by cognitive capacity, the sheer amount of information a learner can hold and manipulate in mind, or by metacognitive regulation, the learner’s ability to plan, monitor, and evaluate their own comprehension? Rather than treating these as rival explanations, the research adopted an integrative framework that views them as complementary, one governing resource availability and the other governing resource regulation.

To test this idea, the researcher recruited 103 male learners aged 15 to 18 from a private language institute in Babol, Iran, all classified as intermediate English as a Foreign Language students. After eligibility screening, participants were randomly assigned to one of four conditions: a working memory training group, a metacognitive intervention group, a combined group receiving working memory training embedded in metacognitive instruction, and a control group that continued with regular instruction. Following attrition-related exclusions, 83 learners completed every phase of the experiment, leaving final group sizes of 19, 21, 20, and 23 respectively.

The design was a randomized pretest-posttest controlled experiment, a gold standard structure for isolating causal effects in educational research. Before the interventions began, all participants completed measures of listening comprehension and working memory capacity, including a Digit Span Test, a classic assessment in which participants must recall increasingly long sequences of digits, either forward or backward. The same battery was administered after the training period, allowing the researcher to compare gains across the four groups while statistically adjusting for initial differences.

The working memory training condition targeted the cognitive machinery underlying listening. Working memory capacity is conventionally understood as a limited pool of attentional resources that determines how much linguistic material a learner can activate, store, and integrate in real time. For a language learner, this capacity constrains everything from holding a clause in mind while parsing its grammar to retaining a speaker’s earlier point long enough to connect it with what follows. Training tasks of the kind used in this study are designed to stretch that capacity through repeated, adaptive practice on demanding memory and attention exercises.

The metacognitive intervention, by contrast, did not aim to enlarge any mental muscle. Instead, it taught learners to regulate the resources they already possess. Metacognition in listening research typically encompasses planning before listening, such as activating background knowledge and setting a purpose; monitoring during listening, such as noticing when comprehension breaks down and deploying repair strategies; and evaluating afterward, such as judging how well the message was understood and what might be done differently next time. Learners in this condition were guided through explicit instruction and practice in these self-regulatory processes, transforming listening from a passive act of reception into an active, strategic performance.

The combined condition integrated the two approaches, embedding working memory training within the metacognitive intervention. The theoretical logic is elegant: a larger cognitive reservoir should make strategic regulation more effective, while better regulation should allow learners to deploy their available capacity more efficiently. If the resource availability and resource regulation frameworks truly capture distinct contributions to listening, then addressing both simultaneously should yield the strongest outcomes of all.

The results were revealing, and in some respects more nuanced than the integrative hypothesis predicted. All three active treatments produced improvements, confirming that both cognitive and metacognitive training are worthwhile. For listening comprehension specifically, the combined condition and the metacognitive intervention alone were the standout performers, both significantly enhancing learners’ ability to understand spoken English. For working memory capacity, the combined condition and working memory training alone were comparably effective, with no statistically significant difference between them. Intriguingly, in both outcome measures the combined group posted a descriptively higher adjusted mean than either single treatment, but that descriptive advantage fell short of statistical significance when compared with metacognitive intervention for listening or working memory training for memory capacity.

These findings carry a double message. On one hand, they validate the integrative framework: learners who received both forms of training never did worse than those who received either alone, and they tended to score highest on both outcomes. On the other hand, they temper expectations about additive benefits. The metacognitive intervention proved nearly as powerful as the full package for listening, and working memory training alone matched the package for building memory capacity. The practical implication is that educators seeking efficient interventions may achieve most of the benefit by choosing the training matched to their goal, while the combined approach remains a theoretically attractive option whose extra value awaits stronger empirical confirmation.

For the millions of English learners worldwide who struggle with listening, the study offers a hopeful reframing. The skills that determine whether a fast-moving conversation in a second language becomes comprehensible are not fixed traits but trainable systems. Teaching learners to plan, monitor, and evaluate their listening appears to be among the most reliable ways to improve comprehension, while targeted cognitive exercise can measurably expand the memory capacity that listening depends on. The next generation of research, the author notes, must continue to probe whether the combined approach delivers benefits beyond its components, but the evidence already suggests that listening instruction should look beyond the ear and toward the mind that interprets what the ear receives.

Subject of Research: The combined effects of working memory training and metacognitive intervention on English listening comprehension and working memory capacity in adolescent EFL learners

Article Title: Beyond hearing: investigating the combined impact of working memory and metacognitive training on EFL listening comprehension and working memory capacity

Article References: Muhammadpour, M. (2026). Beyond hearing: investigating the combined impact of working memory and metacognitive training on EFL listening comprehension and working memory capacity. BMC Psychology. https://doi.org/10.1186/s40359-026-05746-1

Image Credits: AI Generated

DOI: 10.1186/s40359-026-05746-1

Keywords: working memory, metacognition, listening comprehension, EFL, second language acquisition, cognitive training, metacognitive intervention, working memory capacity, randomized controlled trial, language learning, psycholinguistics, BMC Psychology

The research, led by Muhammad Noor Kholid of Universitas Muhammadiyah Surakarta together with colleagues from Indonesia, Vietnam, and Singapore, focused on reflective thinking, a concept with a long intellectual pedigree stretching back to John Dewey. Reflective thinking, in the framework used by the team, is a mental activity that begins with a state of perplexity, whether doubt, difficulty, or error, and is resolved by drawing on prior knowledge and experience through repeated evaluation until the confusion is overcome. In mathematics, this means students do not simply apply procedures; they monitor their reasoning, adjust strategies mid-course, and integrate conceptual understanding as they work toward a solution.

To capture this process as it unfolded, the researchers adopted a descriptive qualitative design involving 35 high school students. Participants were selected precisely because they were reflective thinkers: students who experienced doubt or difficulty when solving mathematical problems but persisted until they reached a correct solution. Each student tackled non-routine problems on arithmetic series, a topic chosen because it sits at a critical juncture in secondary mathematics, underpinning later work on linear equations, statistics, and mathematical modeling. Crucially, the problems could not be solved by plugging numbers into a formula; they demanded interpretation of the problem’s structure and the invention of an appropriate strategy.

The methodological centerpiece was the think-aloud protocol. Students verbalized their reasoning while solving the problems, and the sessions were recorded using audiovisual methods. Answer sheets, observation sheets, and recordings were then analyzed through the constant comparison method, an inductive procedure in which data incidents are compared with one another and with emerging categories until stable classifications emerge. The instruments themselves were rigorously vetted: three experts in mathematics education, assessment, and evaluation rated the tasks, yielding an Aiken’s V index of 1, above the minimum criterion of 0.95, and 100 percent inter-rater agreement on reliability.

The analysis, anchored in the reflective thinking framework of Zehavi and Mann, which identifies four components, techniques, monitoring, insight, and conceptualisation, revealed three clearly distinguishable problem-solving patterns. The first, labeled Employ Variable, describes students who translate the problem’s key quantities into algebraic symbols. The second, Visual Representation, captures students who sketch the situation to build understanding. The third, Direct, characterizes students who plunge straight into manual calculation without constructing any symbolic or visual model of the problem.

In the Employ Variable pattern, exemplified by a student coded PS1, the reflective journey began with a systematic inventory of what was known and what was asked. The student assigned variables, n for the number of rows, b for the difference in seats between rows, and a for the seats in the first row, precisely because the initial reading of the problem produced confusion. In an interview, the student explained that equating critical information with variables made the problem easier to understand and easier to solve. This symbolic scaffolding then allowed the student to apply the arithmetic series sum formula, re-examine each step, and, when confusion resurfaced over how many additional rows of seats were needed, reread the problem repeatedly until the correct first step re-emerged. Every component of reflective thinking was activated: techniques in selecting the strategy, monitoring in re-checking the work, insight in breaking through the impasse, and conceptualisation in linking the variables to the underlying structure.

The Visual Representation pattern told a different but equally complete story. A student coded PS2 responded to the same problem by sketching the seating arrangement of a performance hall, row by row, writing in the seat counts from 10 in the first row to 22 in the last. The sketch was not decoration; it was a core cognitive tool. As the student put it during the interview, drawing the rows allowed him to imagine what the situation actually looked like. From the sketch he computed total seats, multiplied by the ticket price, discovered the revenue fell short of the target, and then extended his tally row by row until the target was met, concluding that six additional rows were required. Here, too, all four reflective components were engaged, with visualization serving to stabilize understanding, reduce confusion, and guide monitoring of each calculation.

The Direct pattern was strikingly different. A student coded PS3 simply wrote down the seat counts and added them up manually, with no formula, no variables, and no sketch. Perplexity did appear, but only later, at the stage of interpreting intermediate results, when the student paused over whether the first revenue figure represented the full answer. Reflection in this pattern was contextual and partial: the student re-checked calculations, adjusted steps when results mismatched the target, and verified the final answer, but never restructured the problem symbolically or visually. The analysis found that Direct-pattern students missed several key reflective indicators, including articulating what was being asked, restructuring information, revisiting the problem when confused, and integrating conceptual understanding. Notably, the researchers caution that this does not signal low ability; rather, it reflects a situational, minimal level of reflection confined largely to monitoring.

The broader significance of the findings lies in what they mean for classrooms. Prior research cited by the team shows that teachers often struggle to recognize the specific strategies students use, tending to identify only general difficulties with comprehension or procedure. By classifying problem-solving patterns through the lens of reflective thinking, the study offers teachers a diagnostic vocabulary: a student stuck in the Direct pattern may need experiences that encourage representation and conceptual reconstruction, while students already employing variables or sketches may benefit from challenges that deepen their monitoring and evaluation. The authors acknowledge limitations, including the modest sample size and the single topic of arithmetic series, and call for future work with open-ended and ill-structured problems, as well as instructional models designed to infuse reflective learning. For now, the study delivers a vivid message: the path a student chooses through a hard problem, whether symbolic, visual, or direct, shapes not just the answer they reach, but the quality of the thinking they do along the way.

Subject of Research: Reflective thinking patterns in students' problem-solving of non-routine arithmetic series tasks

Article Title: Students’ problem-solving patterns in arithmetic series tasks viewed through reflective thinking

Article References: Kholid, M. N., Febriyanti, R., Santosa, Y. T., Faiziyah, N., Hidayati, Y. M., Budi, A. B. S., Sukmawati, T., Suharna, H., Tong, D. H., & Lam, T. T. (2026). Students’ problem-solving patterns in arithmetic series tasks viewed through reflective thinking. Discover Education, 5(1), Article 991. https://doi.org/10.1007/s44217-026-02176-4

Image Credits: AI Generated

DOI: 10.1007/s44217-026-02176-4

Keywords: reflective thinking, arithmetic series, problem-solving, mathematics education, non-routine problems, think-aloud protocol, qualitative research, metacognition, visualization, algebraic variables, high school students, pedagogy