RESEARCH DOSSIER
Creativity Education, Training, and Transfer
Abstract. Creativity education seeks to develop the capacity to generate and refine original, useful ideas within authentic constraints. This critical narrative research monograph examines the history of creative-ability testing and process models, instructional approaches to problem finding, strategic search, analogy, evaluation and revision, and empirical evidence for training outcomes. Central attention is given to the distinction among task learning, near transfer, far transfer, durable retention, and authentic creative achievement. Sio and Lortie-Forgues' 2024 synthesis of 169 studies found a mean unadjusted effect of 0.53 standard deviations, reduced to approximately 0.29–0.32 after publication-bias adjustments. Meta-analyses of adult enhancement and organizational training address different intervention and outcome populations; organizational evidence reveals a pronounced distinction between learning gains and on-the-job behavioral transfer. The synthesis specifies methodological weaknesses, contrasts competing explanatory mechanisms, examines classroom, workplace, cross-cultural and human–artificial contexts, and advances a testable research agenda based on active controls, blinded products, preregistration, and delayed independent transfer assessment. The evidence supports modest teachability of certain creative performances while leaving generalized, sustained creative achievement an open research question.
Keywords: creativity education; creativity training; creative thinking; transfer of learning; near transfer; far transfer; divergent thinking; creative problem solving; analogical reasoning; instructional intervention; creative assessment; creative self-efficacy; publication bias; meta-analysis; educational psychology; Artificial Era
Introduction: Creativity as an Educational Research Problem
Education directed toward creativity attempts to alter the probability, quality, and eventual usefulness of ideas and products through structured experience. Its central outcome is creative performance: a response or artifact that is both original in a relevant comparison class and appropriate to a task, problem, or community. Instruction may change how learners discover problems, retrieve knowledge, combine representations, test alternatives, and revise solutions. The same training can improve a laboratory ideation score while leaving classroom projects or workplace decisions unchanged. This monograph therefore treats creativity education, training, and transfer as a linked empirical question about intervention, measurable change, retention, and generalization. The foundational distinction between creative potential and demonstrated creative achievement is essential to the research design. Runco & Acar (2012) Hennessey & Amabile (2010)
The inquiry proceeds through five connected questions. What psychological activities can instruction alter? Which instructional components produce effects under credible comparisons? What exactly do the common creativity assessments detect? How far do gains extend beyond practiced formats and how long do they persist? Under which individual, institutional, cultural, and technological conditions are new ideas translated into usable products? Evidence is drawn from primary experimental investigations, psychometric and transfer theory, studies of educational environments, and modern quantitative syntheses. Findings about short-term laboratory scores and findings about performance in consequential settings are treated as separate estimands. The aim is a critical narrative synthesis with explicitly identified methodological boundaries, rather than a new experiment or a protocol-based systematic review. Sio & Lortie-Forgues (2024) McKay et al. (2024)
The research object is instructional change in creative cognition and creative work. Nearby topics, including the neural basis of originality, artistic eminence, collaborative innovation, and aesthetic appraisal, enter only where they specify mechanisms or help evaluate an intervention. In the Psychology by Bogdanova architecture, the primary scientific home is Psychology of Aesthetics and Creativity, with strong connections to Educational and School Psychology, Learning and Behavioral Psychology, and Positive Psychology and Well-Being. The relevant research horizon includes learning from other people and from artificial systems, yet empirical claims about transfer require observed performance and clearly specified tasks. Educational aspiration and conceptual possibility become scientific propositions when they are made testable. OECD (2024) Barnett & Ceci (2002)
Defining Creativity, Learning, Training, and Transfer
Creativity is commonly operationalized through novelty and appropriateness, sometimes expressed as originality and effectiveness or usefulness. Novelty is relative to a specified audience, domain, and temporal point; usefulness depends upon constraints and evaluative standards. A learner can generate a statistically unusual idea that fails every engineering requirement, or a workable solution that merely reproduces familiar methods. A valid outcome assessment must decide how these dimensions contribute to the criterion and how judgment uncertainty is handled. The Four C framework distinguishes personally meaningful mini-c interpretations, everyday little-c contributions, professional Pro-c accomplishment, and historically eminent Big-C work. Such distinctions matter because a program effective for novice idea generation cannot simply be credited with producing professional innovation. Kaufman & Beghetto (2009) Amabile (1982)
Creativity education refers to extended instructional contexts in which original thinking and production are developmental objectives integrated with subject learning. Creativity training denotes more bounded activities intended to teach identifiable procedures, strategies, or habits, often evaluated before and after an intervention. Practice can be guided or unguided, individual or collaborative, domain-specific or cross-domain. Transfer means that an acquired change contributes to performance outside its trained instances. Transfer is always relative to the difference between learning and assessment situations: tasks may vary in surface content, domain, time, modality, social context, and purpose. The distinction between near and far transfer must specify these changes rather than merely call a new test 'different.' Barnett & Ceci (2002)
A useful conceptual sequence identifies instructional input, intervening psychological mechanism, performance criterion, retention interval, and transfer criterion. For example, analogical-comparison instruction might teach identification of structural relations; a same-domain analogy test would assay near transfer; a previously unpracticed design problem in another subject months later would assay a more demanding form of transfer. Independent raters could assess originality and constraint satisfaction, and the curriculum could record whether students independently retrieved the analogy procedure. Each stage allows failure even when the preceding stage succeeds. Gick & Holyoak (1980) Runco & Acar (2012)
Historical Foundations: From Creative Ability to Teachable Processes
Modern psychological study of creativity gained institutional momentum with J. P. Guilford's 1950 presidential address to the American Psychological Association, which argued for systematic investigation of creative abilities and the importance of divergent production. The subsequent psychometric tradition sought measurable indicators such as ideational fluency, flexibility, and originality. Torrance developed performance tasks intended to examine creative thinking in educational settings, extending both the practical appeal and the measurement challenges of divergent-thinking research. Neither historical innovation implies that a single score directly measures a person's future achievement across all domains. The concepts became influential partly because the tests offered replicable exercises and measurable outcomes for training studies. Guilford (1950) Runco & Acar (2012)
The second historical trajectory emphasized processes rather than fixed ability: finding and redefining a problem, gathering information, generating possibilities, selecting ideas, and elaborating solutions. Process-analytic models developed by Mumford and colleagues presented creative thinking as organized cognitive activities that could be studied separately. The creative-cognition approach subsequently examined how ordinary operations of memory, imagery, conceptual combination, and evaluation yield surprising ideas. This trajectory supports the practical hypothesis that instruction can change strategy deployment. It also predicts that gains depend on appropriate task representations, domain knowledge, and opportunities for deliberate use, rather than the repeated demand to 'be creative.' Mumford et al. (1991) Finke et al. (1992)
A third trajectory investigated social context. Amabile's research connected creative products to evaluative circumstances, expertise, task motivation, and social incentives, and her consensual assessment technique established expert judgments as a prominent outcome method. Educational approaches increasingly explored the interaction between autonomy, disciplinary knowledge, feedback, peer relationships, and constraints. These strands produce a broader definition of instructional success: learners must recognize useful opportunities for originality, possess applicable knowledge, and inhabit environments where ideas can be tested and revised. The history of these theories establishes a set of plausible mechanisms; modern controlled studies determine how reliably the proposed components improve performance. Amabile (1982) Hennessey & Amabile (2010)
Creative Potential, Products, and Achievement
Training studies often substitute a divergent-thinking task for the complex behavior they ultimately hope to change. Participants may list unusual uses for a common object, complete a visual figure, invent captions, or propose solutions to an open problem. Scoring can quantify the number of responses, diversity of categories, statistical rarity, elaboration, or subjective quality. These are related but nonidentical attributes. Greater fluency creates more opportunities to produce rare responses; novelty ratings can be influenced by the size of a person's response pool and by rater familiarity with the population. Runco and Acar explicitly frame divergent thinking as an indicator of creative potential rather than a full representation of creative accomplishment. Runco & Acar (2012)
Products supply another assessment level. A design prototype, scientific explanation, short story, algorithm, visual composition, or organizational proposal can be judged against domain-relevant originality and appropriateness criteria. Consensual assessment asks suitable experts to judge creative quality, with agreement among judges contributing evidence for reliability. A valid study documents the judges' expertise, whether they were blind to condition, and how discrepancies were resolved. Social recognition and long-term accomplishments add still more contextual conditions, including access to resources, institutional sponsorship, and historical timing. A training program can alter the quality of a student artifact without producing measurable effects on prizes, occupations, or lifetime achievement. Amabile (1982) Kaufman & Beghetto (2009)
Construct-valid evaluation therefore uses a family of outcome measures. The proximal measure tests the trained strategy; a near-transfer measure uses different examples and formats; a far-transfer measure changes a meaningful domain or function; longer follow-up evaluates retention; and authentic products test usefulness beyond a scoring exercise. These levels should be declared in advance. When a study reports many creativity subscores and selects favorable findings afterward, the apparent effect can partly reflect multiplicity rather than general improvement. The appraisal of any intervention begins with what the outcome can legitimately show. Barnett & Ceci (2002) Sio & Lortie-Forgues (2024)
Divergent Production and Strategic Search
Divergent production involves generating multiple possibilities under an open criterion. Its commonly trained elements include deliberate category switching, breaking a habitual association, combining features across examples, and suppressing the first familiar response long enough to explore additional alternatives. These are operational processes rather than an indivisible creative faculty. A lesson might ask students to identify hidden assumptions, search several semantic categories, or transform a problem before proposing solutions. The predicted result is improvement in strategic access to candidate ideas, potentially producing higher quality responses when judgment and domain knowledge are available. Studies of executive switching support the proposition that cognitive control contributes to divergent thinking, though that evidence cannot establish that every trained strategy produces durable transfer. Nusbaum & Silvia (2011)
Nusbaum and Silvia investigated relations among fluid intelligence, category switching, and divergent-thinking performance, and experimentally provided a strategy during an unusual-uses task. Their results suggested that strategy use interacts with executive resources. The inference for instruction is specific: explicit strategies can change performance in a defined ideation paradigm, and learners differ in their capacity to apply those strategies under interference. This finding provides a plausible mechanism for some short-term training effects. It does not by itself warrant claims about increased creativity in science, the arts, or employment, where the relevant work also requires experience, extended evaluation, and external feedback. Nusbaum & Silvia (2011)
A rigorous training contrast should distinguish instruction from mere exposure to the task. Both treatment and control groups can spend equal time generating ideas while only one learns explicit category switching. The outcome can then test whether the taught procedure is used on a new problem, whether judged originality increases after controlling for response count, and whether the advantage survives a delayed test. Failure to control practice, time, and expectations leaves open the possibility that improved scores merely reflect familiarity with the instrument. This distinction is prominent in the modern reassessment of creativity training research. Sio & Lortie-Forgues (2024) Runco & Acar (2012)
Conceptual Combination, Analogy, and Problem Representation
Many creative solutions arise when known elements are reorganized under a newly recognized relation. Conceptual combination can connect distant categories; analogy maps a structural pattern from one problem to another; and problem reframing changes which constraints appear essential. Instruction can make these operations explicit through comparative examples, prompts to identify deeper structure, and exercises requiring alternatives to an initial formulation. The creative-cognition tradition treats such outcomes as products of identifiable ordinary processes. The explanatory question is what features enable a learner to retrieve and adapt a useful representation outside the practice situation. Finke et al. (1992) Mumford et al. (1991)
Gick and Holyoak's original experimental work on analogical problem solving showed that participants could apply the structure of a military problem story to a medical problem, especially when prompted to recognize the analogy. Across their experiments, the success of transfer depended on retrieval, structural correspondence, and cueing. This evidence offers a particularly important boundary condition for creativity instruction: a learner can possess a relevant solution representation and still fail to notice its relevance at the moment of use. Training should therefore be evaluated not only for recall of strategies but for spontaneous recognition of opportunities to apply them. Gick & Holyoak (1980)
Instructional comparisons should distinguish the contribution of structural alignment from the effect of simple reminders. A more demanding test presents an unannounced new problem and observes whether learners independently retrieve the analogy, select relevant constraints, and adapt the solution. Even such a positive result must be interpreted in light of shared problem structure: what appears distant in surface features may be near in causal relations. Transfer classifications require detailed task analysis, and the more distant the context, the more important it becomes to document how the underlying principle remained useful. Barnett & Ceci (2002) Gick & Holyoak (1980)
Knowledge, Expertise, and the Conditions of Useful Originality
Original ideas are produced in relation to accumulated knowledge. Expertise creates a repertoire of techniques, concepts, standards, materials, and failure modes that novices have yet to acquire. It can also promote routine solutions when familiar procedures dominate attention. Education for creativity therefore requires a productive relation between knowledge building and exploratory revision. Domain specificity is especially relevant: creative accomplishment in mathematics, dance, narrative writing, or clinical service design depends on partly different expertise and community criteria. Baer's domain-specificity argument cautions against assuming that a general creative-thinking lesson yields interchangeable skill across all fields. Baer (1998) Hennessey & Amabile (2010)
The educational challenge is to teach conventional elements well enough for learners to transform them. Students cannot critically redesign an experiment without understanding measurement, cannot write an effective original argument without language resources, and cannot build a workable artifact without material constraints. Conversely, instruction that rewards only exact reproduction restricts opportunities to generate and examine alternatives. Teaching should alternate between obtaining domain knowledge, exploring a problem space, constructing candidate solutions, and receiving technically informed feedback. The resulting creative product is assessed simultaneously for originality and successful satisfaction of constraints. Finke et al. (1992) Amabile (1982)
The transfer implication is that instruction must make explicit which elements are generalizable and which are specific to a discipline. A heuristic for broadening search may travel among settings, but the evaluation of a chemical hypothesis or a historical interpretation requires domain-specific criteria. The research program must separate general strategy learning from situated competence and assess their interaction. When interventions mix subject instruction with creativity prompts, stronger performance can result from either improved knowledge or improved exploratory strategy; factorial designs can help identify the distinct contributions. Baer (1998) Barnett & Ceci (2002)
Instructional Architecture and Active Ingredients
Creativity instruction is often packaged as an undifferentiated program even though its constituent activities target different psychological mechanisms. A defensible instructional architecture specifies the opportunity to identify a problem, the technique for expanding possible responses, the criteria used to test alternatives, the practice schedule, and the feedback regime. It also identifies which outcomes are expected to change. In their quantitative review, Scott, Leritz, and Mumford argued that creativity training could be effective and emphasized cognitive process instruction coupled with appropriate exercises. Their conclusions helped establish the value of comparing actual training components rather than aggregating programs solely by brand name. Scott et al. (2004) Mumford et al. (1991)
Training activities can be grouped by the cognitive transition they seek: exposure to new materials widens available representations; strategies such as analogical mapping alter retrieval and search; iterative critique improves evaluation; and social practices establish whether a proposed idea receives serious examination. Repeated practice alone may improve speed on the familiar task; transfer requires that learners recognize relevant principles beyond the practiced setting. The same intervention can contain several activities, making causal attribution difficult. A dismantling design removes one component at a time, while a factorial design crosses instructional elements and estimates their interactions. Multisite replication then assesses whether the combination depends on particular facilitators or learner populations. Sio & Lortie-Forgues (2024) Barnett & Ceci (2002)
Claims that a named method is universally superior require randomized contrasts with active alternatives, aligned total instructional time, sufficiently precise outcome measurement, and a predeclared target population. Published comparisons often vary simultaneously in instructor attention, task novelty, task content, encouragement, and scoring instruments. In such cases the estimated effect belongs to a bundle rather than an identified mechanism. Evidence synthesis should document what was actually delivered, whether fidelity was checked, and how instructors responded when students became stuck. The same standards apply to digital systems that present creative prompts or provide feedback. McKay et al. (2024) Sio & Lortie-Forgues (2024)
Theoretical Models of Creative Learning and Motivation
Componential accounts of creativity distinguish expertise, creativity-relevant processes, motivation, and the social environment. In instructional practice these elements interact. Greater knowledge expands what can be recombined; a generative strategy helps examine unconventional possibilities; and task engagement sustains effort through revision. Motivation affects the probability of attempting difficult alternatives and accepting feedback, while evaluative pressures can change the kinds of risks students regard as acceptable. The evidence supports studying these factors as interacting variables instead of presuming that motivation is a sufficient cause of original work. Hennessey & Amabile (2010) Amabile (1982)
The Four C framework is educationally useful because it places the learner's novel interpretation within a developmental trajectory. Mini-c understanding can be a legitimate outcome of learning even when the resulting insight is unfamiliar only to the learner. More public forms of creativity require progressively broader evaluative standards, increasingly sophisticated knowledge, and opportunities for production. A training study should report its intended level of creativity rather than treating all novelty as equally significant. Increases in creative confidence, classroom participation, and expert-reviewed artifact quality are distinguishable constructs requiring different instruments and interpretations. Kaufman & Beghetto (2009) Tierney & Farmer (2011)
Self-efficacy research adds a temporal question. Belief in one's ability to produce creative work can influence engagement and persistence, yet self-rated confidence may rise independently of the quality of actual products. Tierney and Farmer examined creative self-efficacy development and creative performance over time in organizational settings, demonstrating the relevance of longitudinal analysis rather than a single post-intervention perception measure. Educational evaluations should therefore pair subjective judgments with blinded product ratings, observe repeated work, and assess whether increased self-efficacy mediates the effects of training under plausible temporal ordering. Tierney & Farmer (2011) Amabile (1982)
Training for Problem Finding and Question Formulation
Many programs begin after an instructor has supplied a problem. Creative work frequently begins earlier, when a person identifies a neglected difficulty, constructs a meaningful question, or challenges the premises of an inherited task. Problem finding can involve observation, comparison of conflicting evidence, stakeholder interviews, and examination of constraints. Teaching students to produce alternative problem formulations changes the search space before solution generation begins. Process models of creative thinking recognize problem construction as a consequential activity rather than a decorative preliminary step. Mumford et al. (1991) Finke et al. (1992)
An intervention can operationalize this skill through paired tasks. Before training, students may generate interpretations of an ambiguous situation. During instruction they learn to distinguish symptoms from causes, identify affected groups, list missing information, and reformulate the objective. A subsequent assessment presents a new domain and evaluates whether the reformulated problems are coherent, consequential, and open to novel yet viable solutions. Raters must distinguish a genuinely improved problem representation from a verbally elaborate one. The scoring rubric should be established before viewing treatment assignments and should incorporate independent judgment. Amabile (1982) Barnett & Ceci (2002)
Problem finding is particularly sensitive to institutional affordances. A lesson may encourage reframing while a standardized assessment rewards only one predetermined answer. Students respond to the operative reward system, making experimental context and real classroom conditions essential to interpretation. A transferable problem-finding skill also requires the ability to recognize where reformulation is warranted and where careful execution of an existing procedure is appropriate. Research designs that measure both creative reframing and conventional task accuracy can identify whether instructional gains reflect expanded competence or indiscriminate departure from constraints. Hennessey & Amabile (2010) Vincent-Lancrin et al. (2019)
Guided Practice, Feedback, and Revision
Learning a creativity strategy requires opportunities to apply it, observe consequences, and improve an initial product. Feedback may address the number of alternatives, distinctiveness of conceptual categories, feasibility of proposals, clarity of assumptions, or quality of final expression. Its value depends on its relation to the target: a suggestion that increases novelty while undermining necessary safety constraints may reduce the quality of the work. Instructional feedback should specify which dimension is being discussed and should preserve a reasoned role for learner judgment. The central empirical question is whether feedback changes the learner's subsequent independent decisions rather than only the instructor-assisted artifact. Amabile (1982) Scott et al. (2004)
Process feedback can be designed as a sequence of elicitation and evidence. The teacher first asks students to identify the problem as understood, then to generate alternative approaches, then to provide reasons for evaluating those alternatives. Students compare the limitations of candidate solutions, revise their work, and explain the decisions that made the revision meaningful. Such routines offer multiple points of measurement: quantity of alternatives, diversity, accuracy of evaluation, and final product quality. If only the final version is scored, the process by which an improvement occurred remains uncertain. Mumford et al. (1991) Finke et al. (1992)
Transfer tests should remove the exact prompts and feedback used during instruction. The same rubric can be used to score a new assignment, but its application must respect the destination domain. A useful delayed measure asks whether a learner independently revisits an idea, seeks external evidence, and corrects defects without a specific reminder. For school and workplace settings, this often requires longer observation periods and systematic collection of authentic products. Short laboratory posttests remain valuable as proximal evidence, but their evidential scope is narrower. Barnett & Ceci (2002) McKay et al. (2024)
Constraints, Evaluation, and the Relation Between Novelty and Usefulness
Creative instruction is sometimes interpreted as the removal of constraints, yet all consequential products must function within a domain of material, ethical, technical, and social requirements. A research essay needs evidence, a design must meet its users' needs, and an organizational proposal must respect resources and duties. The ability to identify legitimate constraints can focus search and make previously invisible alternatives discoverable. Teaching creativity therefore includes the evaluation of a proposal against criteria and the revision of ideas that fail those criteria. Unrestricted idea quantity is only one component of an effective process. Hennessey & Amabile (2010) Finke et al. (1992)
A study comparing unconstrained brainstorming with constraint-guided work should measure both originality and task effectiveness, then interpret their relationship. A constraint may initially reduce the number of candidate responses while increasing the proportion that are usable. Alternatively, a poorly designed constraint can suppress exploratory search. These outcomes depend on domain structure and participants' knowledge. The scoring procedure must not reward compliance so heavily that originality becomes immaterial, or privilege strangeness to such a degree that task performance disappears from the criterion. Amabile (1982) Runco & Acar (2012)
Good evaluation instruction teaches learners to distinguish between judgments that can be tested now and claims that require new evidence. Students can make predictions, develop prototypes, obtain feedback from relevant users, or revise a hypothesis when observations contradict it. These practices resemble disciplined inquiry and design iteration. Their creative contribution is measured by the quality and novelty of the resulting changes under actual domain criteria. To determine whether this benefit transfers, researchers should assess later instances in which participants encounter a new constraint without scripted instructions on how to respond. Barnett & Ceci (2002) Vincent-Lancrin et al. (2019)
Attention, Incubation, and Temporal Organization of Practice
Creative work unfolds through episodes of concentrated effort, interruptions, reconsideration, and revision. The incubation hypothesis proposes that an interval away from an unresolved task can facilitate later solutions under some circumstances. Proposed mechanisms include reduction of fixation, continued associative processing, and altered retrieval opportunities. Reviews of incubation research emphasize differences among task types, break duration, and activities during the interval. These patterns do not establish that passive delay will improve any problem; their instructional significance is that time structure may alter the efficacy of strategies already learned. Ritter & Dijksterhuis (2014) Finke et al. (1992)
Training schedules can compare uninterrupted work, a neutral break, a related switching task, and a delayed return. To interpret a benefit, investigators must control for total time, fatigue, additional exposure, and the possibility that participants simply learn to reinterpret the instructions on the second attempt. A delayed return may be helpful for one class of problems and counterproductive for tasks requiring retention of complex details. For longitudinal creativity education, the more important practical question may be whether students routinely revisit tentative ideas after acquiring new information rather than whether a fixed number of minutes yields the best incubation effect. Ritter & Dijksterhuis (2014) Barnett & Ceci (2002)
Temporal structure is also central to retention. Immediate posttests estimate accessibility of a recently practiced technique; later assessment tests whether it remains available when reminders have faded and competing demands have intervened. Evidence from organizational training indicates that effects measured after a delay can be smaller than effects measured immediately. That observation warrants deliberate follow-up measurement but does not identify a universal decay rate because studies differ in outcomes, populations, and timing. McKay et al. (2024)
Learning Environments, Autonomy, and Psychological Safety
Instruction occurs inside systems of evaluation, authority, access, and peer response. Classroom norms influence whether students can share incomplete ideas, question interpretations, and disclose unsuccessful attempts. An environment that invites exploration also needs standards for criticism and revision so that experimentation leads to learning. Learners benefit from clarity about which risks are intellectual, which constraints remain binding, and how contributions will be judged. The social psychology of creativity links motivational and evaluative conditions to creative expression; however, a favorable climate cannot be equated directly with demonstrated improvements in transferable skill. Hennessey & Amabile (2010) Amabile (1982)
Teacher behavior provides multiple possible causal channels. An instructor may model how to reconsider a problem, give students a choice of methods, select unfamiliar materials, or offer feedback on the structure rather than the superficial novelty of an answer. If all these elements occur together, the intervention's effective ingredient is unknown. Classroom studies should document instructional fidelity, baseline student ability, classroom composition, educator expertise, and the extent to which treatment and comparison groups receive comparable attention. Cluster randomization and analysis respecting classroom dependence are often necessary because students share teachers, peers, and evaluation practices. Sio & Lortie-Forgues (2024) Vincent-Lancrin et al. (2019)
Equity matters at the level of measurement and opportunity. A student unfamiliar with the cultural references of a prompt may appear less original because the task demands specialized experiences rather than a general creative capacity. Other students may possess relevant ideas but have fewer resources to prototype, communicate, or exhibit them. A responsible training evaluation uses accessible materials, specifies language and cultural conditions, and interprets output in context. The educational ideal is widened participation in disciplined original work; the empirical requirement is evidence that the intervention improves meaningful performance for the learners actually studied. OECD (2024) Amabile (1982)
Evidence Across Five Decades of Creativity Training
The empirical history of creativity training contains classroom curricula, short laboratory interventions, professional workshops, cognitive exercises, and programs combining multiple forms of instruction. These interventions differ in treatment duration, participant age, task content, instructors, control conditions, and outcome measures. Earlier quantitative syntheses often concluded that training improves creative performance and examined program content as a moderator. Scott, Leritz, and Mumford's review remains historically influential because it connected effectiveness to the cognitive processes taught and the type of practice employed. The resulting research tradition made intervention effects plausible, but its collective evidence must be assessed using methods that account for selective publication, correlated outcomes, and uneven study quality. Scott et al. (2004) Sio & Lortie-Forgues (2024)
Sio and Lortie-Forgues' 2024 meta-analysis integrated 169 studies across five decades and 844 effect sizes, including 48 unpublished studies that supplied 262 effect sizes. The reported unadjusted average training effect was 0.53 standard deviations. Multiple publication-bias adjustments reduced the estimate to a range of 0.29–0.32 standard deviations. The authors also reported prevalent methodological shortcomings and limited evidence of improvement in study quality over time. These statistics establish that a positive mean in the assembled literature is compatible with materially smaller expected benefits under more conservative assumptions. The adjusted values are statistical model-based estimates, not proof that every training program improves performance by a constant amount. Sio & Lortie-Forgues (2024)
These findings should change interpretation and design. An intervention that produces a significant advantage on a locally created immediate test contributes one kind of evidence; a preregistered multisite comparison with active controls, blinded raters, and delayed out-of-domain products contributes a stronger and different kind. The meta-analysis identifies reasons to discount simple effect-size averages, yet it does not demonstrate that creativity education is ineffective. It supports a more specific proposition: a modest average improvement on the outcomes commonly measured is plausible, and the degree to which those improvements persist and generalize remains insufficiently established. Sio & Lortie-Forgues (2024) Barnett & Ceci (2002)
Alternative Meta-Analytic Frameworks and Their Apparent Disagreements
Haase, Hanel, and Gronau examined creativity enhancement methods for adults and synthesized 332 effect sizes, reporting an overall Hedges' g of 0.53 with a 95% confidence interval from 0.44 to 0.61. Their analysis grouped interventions into multiple categories, including training courses and other procedures, and found substantial variation across effects. Their reported publication-bias diagnostics did not produce evidence of bias. This result sits alongside Sio and Lortie-Forgues' evidence of substantial bias in a different training corpus. The discrepancy is scientifically informative because the reviews differ in eligibility, intervention definition, population, analytic choices, and the studies available for each outcome. Haase et al. (2025) Sio & Lortie-Forgues (2024)
Two meta-analyses cannot be directly compared by looking only at the size of their grand means. A synthesis combining meditation, cultural exposure, and training courses answers a question about short-term creativity enhancement broadly defined. A training-specific meta-analysis of five decades addresses a narrower intervention class with its own methodological problems. Differences in dependence correction, study-selection models, small-study-effect methods, and the definition of unpublished work further affect results. A credible comparative synthesis starts with the estimands and source populations, then asks why conclusions differ and what information each supplies about real-world instruction. Haase et al. (2025) Sio & Lortie-Forgues (2024)
The appropriate conclusion is conditional. Different forms of creative activity may be enhanced in controlled circumstances, but the magnitude and reliability of benefit depend on what is taught, who receives it, what is measured, and how publication and measurement practices are handled. Independent replication of specific training mechanisms and genuine out-of-sample prediction would be more decisive than treating any single pooled effect as a universal constant. Review conclusions also need attention to whether individual studies themselves were randomized and whether outcome assessors were blinded. Haase et al. (2025) McKay et al. (2024)
Organizational Training: Learning Gains Versus Behavioral Transfer
Organizational settings provide a crucial test because creative learning is expected to influence actual work rather than only training-room responses. McKay, Reiter-Palmon, Coombes, and Coombs assembled 40 creativity-training studies involving 3,522 participants. Their quantitative synthesis reported an overall Hedges' g of 0.68 and a learning-outcome effect of 0.73. The estimated effect on on-the-job behavior and transfer was g = 0.34 and was not statistically significant. These outcome categories differ in the causal distance from the instruction: recalling strategies or completing learning tests is easier to observe than independent use in complex organizational routines. McKay et al. (2024)
The same synthesis found smaller effects for delayed evaluation than for immediate evaluation: g = 0.40 and g = 0.86, respectively, in the authors' moderator comparison. This contrast does not necessarily estimate individual decay, since different studies, timing, and outcome measures may contribute to the subgroups. It nevertheless shows why immediate improvement should be kept analytically distinct from retention and actual behavior. If training changes what employees know but organizational incentives, time pressure, leadership, or resources inhibit implementation, the causal chain from learned strategy to innovation remains incomplete. McKay et al. (2024) Tierney & Farmer (2011)
Organizational transfer studies require clear performance criteria, comparable opportunities to produce novel work, and measures that are not defined exclusively by participants' satisfaction with a workshop. Blind appraisal of artifacts, observed changes to operating procedures, and recorded decisions on proposed improvements can complement self-reports. Such endpoints may depend on supervisors and colleagues as well as trainees, motivating cluster or multilevel analyses. The available evidence justifies continued experimental study with delayed outcomes; it provides weaker support for broad assurances that a workshop will reliably transform workplace creative achievement. McKay et al. (2024) Barnett & Ceci (2002)
Near Transfer, Far Transfer, and the Geometry of Difference
Transfer is not a single binary variable. Barnett and Ceci's taxonomy distinguishes multiple dimensions along which source and target conditions differ. A creativity strategy might be applied to a new object in the same classroom the next day, to a new subject during the next semester, or to an occupational challenge several years later without prompting. The comparison changes in knowledge domain, physical and social setting, temporal distance, function, and modality. A scientifically informative study should state these distances in advance so that readers know what kind of generalization has been tested. Barnett & Ceci (2002)
Near-transfer tasks retain important underlying structure while varying surface materials. They are necessary for distinguishing skill acquisition from memorization of a single answer. Far-transfer tasks alter meaningful features of the target problem, perhaps demanding different disciplinary standards, forms of representation, or social coordination. Whether a task is far depends on the mechanism being claimed. Two differently worded unusual-uses tests may be perceptually distinct while relying on nearly identical strategies, whereas production of a research protocol involves knowledge and evaluation requirements absent from both. Runco & Acar (2012) Baer (1998)
Scientific reports should present a transfer map showing which features of the trained task were shared with each assessment and which were changed. Spontaneous transfer ought to be distinguished from cued transfer: a reminder to use a technique changes the instructional environment and partly supplies the retrieval decision that independent transfer requires. The classic analogy experiments found that hints could dramatically affect application of a previously seen solution. That result illustrates why a program that works when the task names the learned strategy may have an uncertain effect when the learner must recognize the occasion alone. Gick & Holyoak (1980) Barnett & Ceci (2002)
Retention, Delayed Measurement, and Sustained Creative Practice
Retention describes the persistence of an acquired strategy or change in performance; transfer describes its application in a different situation. A learner might remember the mechanics of a creativity exercise after several months without choosing to use it in independent work. Conversely, successful new behavior could arise from ordinary practice or changing opportunity rather than persistence of the original training mechanism. Strong studies therefore measure both the retained ability and its spontaneous use outside the training context. This distinction is especially consequential when policy or institutional budgets are justified by claims of enduring skills. McKay et al. (2024) Barnett & Ceci (2002)
Delayed assessments require explicit timelines: time from instruction to posttest, to follow-up, and to genuine performance opportunity. Attrition must be reported and compared across conditions, because missing participants may differ in motivation or skill. Additional instruction, ordinary maturation, repeated assessments, and exposure to creative work during the interval may contribute to observed results. Researchers should record such exposure and avoid attributing all long-term change to a brief initial course. For school research, longitudinal administrative outcomes can complement repeated classroom tasks, but they introduce their own confounding and measurement challenges. Sio & Lortie-Forgues (2024) OECD (2024)
Durable creative practice is likely to depend partly on routines that prompt periodic problem reconsideration, access to substantive feedback, and opportunities to enact ideas. This is a reasoned instructional hypothesis drawn from the distinction between knowledge and behavior, not a demonstrated universal mechanism. Research can test it by assigning maintenance prompts or continued practice independently of initial training and measuring whether novel, useful products persist at subsequent intervals. Such designs identify whether a one-time intervention is sufficient or whether benefits require an organized ecology of continued use. McKay et al. (2024) Scott et al. (2004)
Causal Identification and Research Design
The strongest test of a training effect compares plausible potential outcomes for similar learners under different instructional conditions. Random assignment is valuable, but implementation decisions affect what it identifies. An inactive control may differ from an active training condition in expectations, attention, time, novelty, and social encouragement. An active comparator can instead provide similar opportunities and effort while excluding the specific strategy believed to produce change. A design that separates strategy teaching from additional practice yields a sharper causal inference than a simple treated-versus-untreated contrast. Sio & Lortie-Forgues (2024) Scott et al. (2004)
Allocation concealment, baseline measurement, preregistration, sample-size planning, and blinded outcome judgment are important when creative scores permit flexibility. Researchers should predefine whether they analyze individual ideas, average product ratings, creative process behavior, or a composite. If instructors deliver the intervention to intact classrooms, the assignment and analysis should respect classroom clustering; treating correlated student observations as independent can exaggerate precision. Multiple creativity measures also require clear hierarchy or multiplicity control. Intervention adherence and exposure should be reported alongside outcome estimates, making it possible to distinguish a weak method from unsuccessful delivery. Sio & Lortie-Forgues (2024) Amabile (1982)
Follow-up analyses ought to preserve the randomized comparison when possible, document missing data, and avoid interpreting posttreatment mediators as experimentally manipulated causes. Claims of subgroup effectiveness require formal interaction tests rather than selective reporting of significance within one group. Effect estimates should include uncertainty intervals and baseline context, because a modest standardized mean difference can imply very different practical consequences depending on measurement, population, and cost. The field's recurring methodological problems make openly specified estimands and shareable instruments central to cumulative science. Sio & Lortie-Forgues (2024) McKay et al. (2024)
Publication Bias, Researcher Degrees of Freedom, and Replication
Publication bias arises when the probability that a study appears in the accessible literature depends partly on its findings. Creativity training is particularly vulnerable because interventions can be evaluated with numerous overlapping fluency, originality, flexibility, elaboration, and self-rating outcomes. An investigator may also try different scoring rules or compare multiple subgroups. If successful contrasts are more likely to be written up or accepted, the published mean can overstate the effect typical of all conducted studies. The 2024 meta-analysis using unpublished material and multiple adjustment methods establishes that this is a material concern in the historical training record. Sio & Lortie-Forgues (2024)
A rigorous synthesis should distinguish selection for publication from small-study effects, variation in intervention fidelity, and genuine moderation. Different statistical correction methods impose different assumptions, so convergent adjusted estimates increase confidence without eliminating model uncertainty. The presence of a nonsignificant bias diagnostic in another review does not prove that no selective reporting occurred, especially when the review uses a different evidence base or has limited diagnostic power. Both negative and positive studies must be represented to estimate average effects responsibly. Sio & Lortie-Forgues (2024) Haase et al. (2025)
Preregistered studies with adequate sample sizes, independent replication, registered-report publication, transparent datasets, and access to full measurement protocols provide practical routes toward stronger evidence. Replications should preserve essential intervention components while varying instructor, site, age, and disciplinary context. This allows researchers to identify both reproducibility under similar conditions and generalizability under changed conditions. Future meta-analyses would benefit from hierarchically modeling dependent outcomes, coding control-group quality, and separating proximal from delayed transfer outcomes, rather than attempting to average unlike criteria into an unqualified creativity effect. Sio & Lortie-Forgues (2024) Barnett & Ceci (2002)
Classroom Implementation Across Developmental Stages
Educational settings require creativity instruction to be organized around learners' existing knowledge and developmental capabilities. Young students may benefit from concrete comparisons, imaginative reformulation, and explicit explanation of how a novel answer can still meet a task's purpose. Adolescents can work with increasingly complex disciplinary constraints, group critique, and the documentation of design choices. University students may need strategies for integrating evidence across fields and revising research questions. These are design considerations based on differences in educational demands; they do not imply that a single training method has demonstrated the same causal effect at every age. Kaufman & Beghetto (2009) Vincent-Lancrin et al. (2019)
Within a subject, a useful sequence begins with accurate foundational knowledge, presents a genuinely open question, models a search or reframing procedure, and moves students through production, evidence-based evaluation, and revision. Teachers can distinguish occasions for original inquiry from occasions for mastering established procedures. Creative performance is then represented by artifacts and decisions rather than by confidence or enjoyment alone. The OECD's work on fostering student creativity and critical thinking offers a framework for connecting such instructional goals to school practice, while emphasizing the need to define how the abilities are demonstrated and assessed. Vincent-Lancrin et al. (2019) OECD (2024)
Implementation research should observe how much of the intended curriculum teachers deliver, whether instructions are understandable to learners, and how ordinary assessment pressures affect participation. It should compare similar classrooms receiving credible alternative instruction, collect both student and teacher characteristics, and report distributions of effects rather than only an average. A method that appears beneficial for students already confident in a subject may be less accessible to novices or multilingual learners. Identifying these interactions allows curricula to be adapted to observed needs without assuming that differences represent immutable creative ability. Sio & Lortie-Forgues (2024) OECD (2024)
Educational Assessment, International Comparison, and PISA
Large-scale educational assessments offer descriptive evidence about the distribution and correlates of creative-thinking performance. The OECD's PISA 2022 creative-thinking assessment, reported in 2024, compared student responses to tasks involving the generation, evaluation, and improvement of ideas across specified contexts. Such research can reveal differences among educational systems and associations with background conditions, but the data are predominantly observational and do not identify the causal effect of a particular creativity-training program. A high-performing system may differ from another in socioeconomic resources, language, culture, selection, curricular organization, and numerous unmeasured factors. OECD (2024)
Comparative scores also depend on test design. Cultural familiarity with prompts, accepted forms of argument, scoring conventions, translation, and the opportunity to practice comparable tasks can influence observed performance. A construct that is intended to be broadly comparable requires careful evidence about measurement invariance and the meaning of performance in different contexts. Training experiments can complement system-level comparison by manipulating instructional content and using both common measures and locally meaningful products. The two methods answer different questions and become most useful when their differences are preserved. OECD (2024) Runco & Acar (2012)
School-level policy should therefore resist translating descriptive ranking differences directly into prescriptions. Comparative surveys can identify environments worthy of closer study and generate hypotheses about resources or pedagogy. Experimental and longitudinal work can then determine whether a proposed practice contributes to improvement, which learners benefit, and whether the effects persist. Bringing these approaches together is a research agenda for cumulative educational science, not a completed demonstration that any one national curriculum causes creativity. OECD (2024) Vincent-Lancrin et al. (2019)
Individual Variation, Emotion, and Motivation
Learners approach open-ended tasks with different knowledge, cognitive strategies, tolerance of ambiguity, interests, and histories of evaluation. These differences may moderate training uptake and subsequent performance. A strategy requiring active switching between semantic categories, for example, imposes demands on executive control and working knowledge. An approach relying on intrinsic engagement may be more effective when a task is meaningful and the learner has opportunities to exercise choice. Such propositions need direct moderation tests within sufficiently powered interventions. A correlation between individual characteristics and creative scores does not automatically reveal how training should be tailored. Nusbaum & Silvia (2011) Hennessey & Amabile (2010)
Emotion interacts with creative cognition in more complex ways than the popular claim that a positive mood invariably stimulates originality. A meta-analysis by Baas, De Dreu, and Nijstad examined mood and creativity with attention to activation and regulatory processes. Its contribution to instructional theory is that affective states need to be described in terms of underlying motivational and attentional features rather than as broad labels. A productive research design records task demands and emotional context and investigates whether different activities—generation, evaluation, or revision—show distinct relationships with affect. Baas et al. (2008)
Self-efficacy is another candidate mediator. A learner who anticipates successful idea production may attempt more problems and persist after an unsuccessful trial; confidence may also become inflated without corresponding gains in evaluated products. Longitudinal evidence relating creative self-efficacy and performance illustrates the need to study both constructs and their temporal relationship. In training research, subjective ratings are useful as process variables and outcomes in their own right, while independent creative achievement needs its own assessment. Tierney & Farmer (2011) Amabile (1982)
Equity, Cultural Context, and Accessibility
Creativity education operates in multilingual, culturally varied, and materially unequal environments. Standards of originality and appropriateness partly reflect cultural practices and disciplinary communities. A task requiring familiarity with particular consumer products or artistic conventions can advantage learners who have encountered those materials. The resulting difference may reflect differential opportunity rather than differences in a domain-general ability to reorganize information. An inclusive scientific approach specifies the cultural and linguistic content of stimuli, the expertise of raters, and the communities for whom outcomes are meaningful. OECD (2024) Amabile (1982)
Access to tools, time, space, mentors, and publicly visible platforms also matters. Two learners who acquire similar ideation strategies can differ greatly in their ability to test prototypes or receive specialized feedback. Evaluations should therefore report implementation resources and investigate whether the intervention changes opportunity as well as cognition. When a creativity curriculum requires proprietary software, extracurricular time, or expensive materials, its observed effect may be entangled with access. Researchers can test equitable implementation by designing parallel low-resource conditions and by predefining outcomes that do not simply reward resource-intensive presentation. Vincent-Lancrin et al. (2019) Hennessey & Amabile (2010)
Assessment should preserve disciplinary standards while making the evidence used for judgments transparent. Multiple qualified raters, appropriate instructions, and sample products can help establish reliability, although agreement alone does not prove cultural neutrality. When originality is evaluated by one dominant community, the scope of the criterion should be described accordingly. Participatory development of tasks and rubrics, combined with cross-group validation, provides a basis for evaluating whether a training program broadens meaningful participation in creative work. Amabile (1982) Runco & Acar (2012)
Digital Learning and Human–Artificial Creative Collaboration
Digital systems can provide varied prompts, rapid examples, iterative feedback, and access to unfamiliar representations. Artificial generative systems add the possibility of suggestions and partial artifacts that students may inspect, combine, reject, or revise. In principle these tools can alter the cost of producing alternatives, the burden of evaluation, and the distribution of activity between learner and system. The appropriate psychological question concerns what the human learner subsequently knows, can do, and chooses to do. A successful co-produced artifact does not alone demonstrate independent learning or a durable transfer effect. Hennessey & Amabile (2010) Barnett & Ceci (2002)
A useful experiment can separate unaided performance, assisted performance, and later unaided transfer. One group might receive algorithmically generated candidate ideas; another might receive prompts to explain its own alternatives; a comparison group might use ordinary reference materials. Outcome measures would assess originality, correctness, provenance, learner explanation, and retention after the system is removed. Researchers could also observe whether assistance increases reliance on stereotyped suggestions or promotes independent strategy discovery. These are proposed research designs for the Artificial Era, not conclusions about a universal effect of current artificial systems. Vincent-Lancrin et al. (2019) Runco & Acar (2012)
The conceptual horizon From Homo to Artificial introduces further questions about distributed creative organization and attribution of contributions. Scientific investigation should specify operational units—prompts, generated options, evaluation decisions, revisions, and resulting products—and establish which interventions change their sequence. Psychological transfer belongs to the learner or system whose behavior is actually measured; claims about artificial cognition require dedicated evidence and cannot be inferred merely from fluent outputs. This research dossier therefore treats human–artificial co-production as an emerging test environment for creativity education and keeps its specific joint-production science for a separate research object. Barnett & Ceci (2002) McKay et al. (2024)
A Research Agenda for Cumulative Creativity Education Science
The most consequential unresolved question is how to connect short-term improvement on creativity tasks with durable, useful action in a different setting. A cumulative program should establish a limited number of carefully defined intervention mechanisms, publish independent protocols, and align outcome categories before collecting data. Ideally, participants would be randomly assigned to active, expectancy-matched conditions; instructors would be trained and fidelity recorded; outcomes would include blinded proximal tasks, near-transfer tasks, and delayed authentic products. Such a design would make progress beyond a literature dominated by short, heterogeneous programs and flexible assessment choices. Sio & Lortie-Forgues (2024) McKay et al. (2024)
Research should explicitly test potential pathways: instruction may improve strategy retrieval, problem representation, category switching, knowledge integration, idea evaluation, persistence, or contextual opportunity. Mechanism claims require measures temporally and conceptually distinct from the outcomes they purport to explain. A mediation analysis performed after training can support a causal model only under additional assumptions; factorial manipulation of components can provide stronger identification. Individual differences and school or workplace contexts should be analyzed as predicted moderators, not retrospective explanations for isolated positive results. Nusbaum & Silvia (2011) Mumford et al. (1991)
Replication priorities include longer assessment intervals, transparent measurement scoring, publication of null findings, multilingual and cross-cultural validation, and realistic economic evaluation. Publicly available de-identified data and analysis code can allow researchers to examine correlated outcomes and selective reporting. Future quantitative reviews should distinguish program effects on ideation, evaluated products, self-efficacy, and real implementation; only then can comparisons among instruction types become causally and practically informative. The immediate scientific task is to improve the precision of questions, not simply to accumulate more programs bearing the label creativity training. Sio & Lortie-Forgues (2024) Haase et al. (2025)
Critical Scientific Discussion
Three propositions emerge from the evidence. First, certain forms of creative performance are teachable at least to a modest degree under the conditions captured by existing studies. Second, the size of the average effect depends materially on inclusion criteria, outcome definition, study quality, and adjustments for publication bias. Third, near-term learning outcomes support weaker inferences about distant and sustained creative practice than public discussion frequently assumes. This synthesis is anchored in the 2024 training meta-analysis, the adult enhancement review, and the organizational training analysis, which address related but different questions and therefore cannot be collapsed into one global estimate. Sio & Lortie-Forgues (2024) Haase et al. (2025) McKay et al. (2024)
The strongest theoretical accounts connect teachable processes to their contexts of use. Explicit strategies may influence divergent search, analogical mapping, or problem definition; disciplinary expertise constrains whether an idea is useful; motivation and social conditions affect the probability that an idea will be attempted and evaluated; and subsequent opportunity determines whether learning is enacted. Each link is independently testable. Difficulties in transfer may result from failure to recognize relevance, insufficient task knowledge, weak retention, or environments that provide no realistic opportunity to apply a skill. These possibilities imply different educational interventions and experimental contrasts. Gick & Holyoak (1980) Baer (1998) Tierney & Farmer (2011)
Several recurring limitations prevent a stronger conclusion. The field contains heterogeneous training bundles, often small studies, incomplete active comparisons, varied scoring systems, short follow-up intervals, and selective outcome reporting. Published aggregate effects cannot settle which mechanism works best in a particular classroom or institution. Some contemporary reviews reach different judgments about publication bias because their samples and methods differ. The responsible response is to make research designs more explicit, preserve uncertainty around each estimand, and evaluate whether observed performance changes meet meaningful educational objectives. Sio & Lortie-Forgues (2024) Haase et al. (2025) Barnett & Ceci (2002)
Conclusion
Creativity education, training, and transfer form a single research domain centered on the instructional organization of original, useful thought and production. The evidence supports the possibility of improving certain creative performances through structured instruction and practice. It also shows that many conventional outcome measures capture potential or immediate test behavior, leaving the transfer of learning to new tasks and sustained work incompletely demonstrated. Once publication bias and study quality are considered, the expected average benefits of generic creativity training are more modest than older unadjusted summaries implied. Sio & Lortie-Forgues (2024) Runco & Acar (2012)
The most defensible educational architecture develops subject knowledge, explicitly teaches search and reframing processes, provides guided opportunity for evaluation and revision, and measures independent performance under changed conditions. Its scientific evaluation separates acquisition, near transfer, far transfer, retention, and consequential achievement, with transparent scoring, appropriate comparison groups, and meaningful follow-up. Different creative domains impose different standards and opportunities, so their commonalities should be established empirically rather than stipulated. Scott et al. (2004) Barnett & Ceci (2002) Baer (1998)
For Psychology by Bogdanova, the enduring conclusion is methodological and substantive: creative capacity becomes a scientific object of education when instructional operations, psychological mechanisms, products, and conditions of generalization are independently specified and connected by evidence. Artificial systems may expand the environments in which human creative learning takes place, but their educational value depends on measured learning, attribution of contributions, and transferable competence. Future research can advance by establishing what changes, in whom, under which instruction, for how long, and with what independently verifiable consequence. McKay et al. (2024) OECD (2024)
References
- Amabile, T. M. (1982). Social psychology of creativity: A consensual assessment technique. Journal of Personality and Social Psychology, 43(5), 997–1013. https://doi.org/10.1037/0022-3514.43.5.997
- Baas, M., De Dreu, C. K. W., & Nijstad, B. A. (2008). A meta-analysis of 25 years of mood–creativity research: Hedonic tone, activation, or regulatory focus? Psychological Bulletin, 134(6), 779–806. https://doi.org/10.1037/a0012815
- Baer, J. (1998). The case for domain specificity of creativity. Creativity Research Journal, 11(2), 173–177. https://doi.org/10.1207/s15326934crj1102_7
- Barnett, S. M., & Ceci, S. J. (2002). When and where do we apply what we learn? A taxonomy for far transfer. Psychological Bulletin, 128(4), 612–637. https://doi.org/10.1037/0033-2909.128.4.612
- Finke, R. A., Smith, S. M., & Ward, T. B. (1992). Creative cognition: Theory, research, and applications. MIT Press. https://mitpress.mit.edu/9780262061506/creative-cognition/
- Gick, M. L., & Holyoak, K. J. (1980). Analogical problem solving. Cognitive Psychology, 12(3), 306–355. https://doi.org/10.1016/0010-0285(80)90013-4
- Guilford, J. P. (1950). Creativity. American Psychologist, 5(9), 444–454. https://doi.org/10.1037/h0063487
- Haase, J., Hanel, P. H. P., & Gronau, N. (2025). Creativity enhancement methods for adults: A meta-analysis. Psychology of Aesthetics, Creativity, and the Arts, 19(4), 708–736. https://doi.org/10.1037/aca0000557
- Hennessey, B. A., & Amabile, T. M. (2010). Creativity. Annual Review of Psychology, 61, 569–598. https://doi.org/10.1146/annurev.psych.093008.100416
- Kaufman, J. C., & Beghetto, R. A. (2009). Beyond big and little: The Four C model of creativity. Review of General Psychology, 13(1), 1–12. https://doi.org/10.1037/a0013688
- McKay, A. S., Reiter-Palmon, R., Coombes, S. M. T., & Coombs, J. E. (2024). A meta-analysis of creativity training in organizational settings. Creativity and Innovation Management, 33(4), 587–602. https://doi.org/10.1111/caim.12605
- Mumford, M. D., Mobley, M. I., Reiter-Palmon, R., Uhlman, C. E., & Doares, L. M. (1991). Process analytic models of creative capacities. Creativity Research Journal, 4(2), 91–122. https://doi.org/10.1080/10400419109534380
- Nusbaum, E. C., & Silvia, P. J. (2011). Are intelligence and creativity really so different? Fluid intelligence, executive processes, and strategy use in divergent thinking. Intelligence, 39(1), 36–45. https://doi.org/10.1016/j.intell.2010.11.002
- OECD. (2024). PISA 2022 results (Volume III): Creative minds, creative schools. OECD Publishing. https://doi.org/10.1787/765ee8c2-en
- Ritter, S. M., & Dijksterhuis, A. (2014). Creativity—the unconscious foundations of the incubation period. Frontiers in Human Neuroscience, 8, Article 215. https://doi.org/10.3389/fnhum.2014.00215
- Runco, M. A., & Acar, S. (2012). Divergent thinking as an indicator of creative potential. Creativity Research Journal, 24(1), 66–75. https://doi.org/10.1080/10400419.2012.652929
- Scott, G., Leritz, L. E., & Mumford, M. D. (2004). The effectiveness of creativity training: A quantitative review. Creativity Research Journal, 16(4), 361–388. https://doi.org/10.1080/10400410409534549
- Sio, U. N., & Lortie-Forgues, H. (2024). The impact of creativity training on creative performance: A meta-analytic review and critical evaluation of 5 decades of creativity training studies. Psychological Bulletin, 150(5), 554–585. https://doi.org/10.1037/bul0000432
- Tierney, P., & Farmer, S. M. (2011). Creative self-efficacy development and creative performance over time. Journal of Applied Psychology, 96(2), 277–293. https://doi.org/10.1037/a0020952
- Vincent-Lancrin, S., González-Sancho, C., Bouckaert, M., de Luca, F., Fernández-Barrerra, M., Jacotin, G., Urgel, J., & Vidal, Q. (2019). Fostering students' creativity and critical thinking: What it means in school. OECD Publishing. https://doi.org/10.1787/62212c37-en