Figure 1 - uploaded by Roger E Beaty
Content may be subject to copyright.
PRISMA flow diagram illustrating study identification, screening, and selection processes. Blue boxes = records interrogated for inclusion; Red boxes = excluded records; Green boxes = included in meta-analysis.
Source publication
Researchers have been studying creativity for decades, and yet controversy still surrounds the cognitive basis of creative thought. A longstanding question in the creativity literature concerns the role of memory in creative cognition. Increasing evidence suggests that specific memory systems (e.g., episodic vs. semantic) may support specific creat...
Contexts in source publication
Context 1
... we conducted a manual reference screen by extracting the references of all articles meeting inclusion criteria from the database search and applied the same three step screening procedure. The entire study identification, screening, and eligibility process is shown in Figure 1. The review and protocol for this study were unregistered and were considered exempt by the XXX Institutional Review Board. ...
Context 2
... full study identification, screening, and selection process are displayed in Figure 1. In total, we included 525 effect sizes from 79 unique empirical articles and unpublished datasets (indicated by the studies listed in effect sizes from young adults, and 12 effect sizes from older adults. ...
Citations
... Gilhooly et al.'s (2007) exploration of cognitive strategies used in alternate uses test showed that personally experienced uses are frequently retrieved during idea generation process. Growing knowledge base also impacts memory, which is in turn related to creative ideation Gerver et al., 2022;Miroshnik et al., 2023). These findings also imply the need for age-based norms when using MOTES for high-stakes decisionmaking, or using age as a covariate. ...
Creativity is highly valued in both education and the workforce, but assessing and developing creativity can be difficult without psychometrically robust and affordable tools. The open-ended nature of creativity assessments has made them difficult to score, expensive, often imprecise, and therefore impractical for school- or district-wide use. To address this challenge, we developed and validated the Measure of Original Thinking for Elementary School (MOTES) in five phases, including the development of the item pool and test instructions, expert validation, cognitive pilots, and validation of the automated scoring and latent test structure. MOTES consists of three game-like computerized activities (uses, examples, and sentences subscales), with eight items in each for a total of 24 items. Using large language modeling techniques, MOTES is scored for originality by our open-access artificial intelligence platform with a high level of agreement with independent subjective human ratings across all three subscales at the response level (rs = .79, .91, and .85 for uses, examples, and sentences, respectively). Confirmatory factor analyses showed a good fit with three factors corresponding to each game, subsumed under a higher-order originality factor. Internal consistency reliability was strong for both the subscales (H = 0.82, 0.85, and 0.88 for uses, examples, and sentences, respectively) and the higher-order originality factor (H = 0.89). MOTES scores showed moderate positive correlations with external creative performance indicators as well as academic achievement. The implications of these findings are discussed in relation to the challenges of assessing creativity in schools and research.
... Gilhooly et al.'s (2017) exploration of cognitive strategies used in Alternate Uses Test showed that personally experienced uses are frequently retrieved during idea generation process. Growing knowledge base also impacts memory, which is in turn related to creative ideation Gerver et al., 2022;Miroshnik et al., 2023). These findings also imply the need for age-based norms when using MOTES for high-stakes decision-making, or using age as a covariate. ...
Creativity is highly valued in both education and the workforce, but assessing and developing creativity can be difficult without psychometrically robust and affordable tools. The open-ended nature of creativity assessments has made them difficult to score, expensive, often imprecise, and therefore impractical for school- or district-wide use. To address this challenge, we developed and validated the Measure of Original Thinking for Elementary School (MOTES) in five phases, including the development of the item pool and test instructions, expert validation, cognitive pilots, and validation of the automated scoring and latent test structure. MOTES consists of three game-like computerized activities (Uses, Examples, and Sentences subscales), with eight items in each for a total of 24 items. Using large language modeling techniques, MOTES is scored for originality by our open-access artificial intelligence (AI) platform with a high level of agreement with independent subjective human ratings across all three subscales at the response level (rs = .79, .91, and .85 for Uses, Examples, and Sentences, respectively). Confirmatory factor analyses showed a good fit with three factors corresponding to each game, subsumed under a higher-order originality factor. Internal consistency reliability was strong for both the subscales (H = .82, .85, and .88 for Uses, Examples, and Sentences, respectively) and the higher-order originality factor (H = .89). MOTES scores showed moderate positive correlations with external creative performance indicators as well as academic achievement. The implications of these findings are discussed in relation to the challenges of assessing creativity in schools and research.
... Factor analytic studies of human cognitive abilities have operationalized goal-directed retrieval as a facet of intelligence, termed broad retrieval ability [32,33]. Broad-retrieval ability (i.e., verbal fluency) robustly predicts performance on creative thinking tasks, according to two recent meta-analyses [34,35], and requires executive control to extract information from long-term memory, avoiding repetitions and inappropriate responses along the way. ...
Creativity has long been thought to involve associative processes in memory: connecting concepts to form ideas, inventions, and artworks. However, associative thinking has been difficult to study due to limitations in modeling memory structure and retrieval processes. Recent advances in computational models of semantic memory allow researchers to examine how people navigate a semantic space of concepts when forming associations, revealing key search strategies associated with creativity. Here, we synthesize cognitive, computational, and neuroscience research on creativity and associative thinking. This Review highlights distinctions between free- and goal-directed association, illustrates the role of associative thinking in the arts, and links associative thinking to brain systems supporting both semantic and episodic memory – offering a new perspective on a longstanding creativity theory.
... A and Sect. B. Since memory is one of the prerequisites of creativity [30,31], we employ the ESN algorithm described in Sect. C to demonstrate in Sect. ...
... Consequently, this creative task is especially challenging for AI system because it requires a machine to have some of the key features of the human intelligence such as the ability to associate ideas, perceive, think, search for answers and criticise results of own work [45]. Yet, creativity relies on cognitive memory [30,31] and is closely linked to cultural context and personality, also being influenced by motivation and emotions of the artist [45]. Interestingly enough, the ability to appreciate, arrange and compose heavy metal music has also been associated with high intellectual abilities [46], which means that the production of heavy metal style music should be a particularly challenging task for AI. ...
Producing original and arranging existing musical outcomes is an art that takes years of learning and practice to master. Yet, despite the constant advances in the field of AI-powered musical creativity, production of quality musical outcomes remains a prerogative of the humans. Here we demonstrate that a single bubble in water can be used to produce creative musical outcomes, when it nonlinearly oscillates under an acoustic pressure signal that encodes a piece of classical music. The audio signal of the response of the bubble resembles an electric guitar version of the original composition. We suggest, and provide plausible theoretical supporting arguments, that this property of the bubble can be used to create physics-inspired AI systems capable of simulating human creativity in arrangement and composition of music.
... Studies have unveiled the constructive, dynamic nature of memory and its important role in imagination and creative thinking 3,21 . In parallel, mounting evidence has linked creative performance to memory processing 20,22 , and to differences in memory structure 23,24 and underlying brain networks 25,26 . These findings have converged to offer a unique understanding of the role of memory in creativity. ...
... Furthermore, the MemiC framework serves to contextualize well-known cognitive effects related to creative performance such as cognitive fixation and incubation 19,173 that can especially impact the search stage. Notably, the ideation stages differ in how memory is applied, but basic memory functions such as memory search and retrieval might still influence the memory usage at each ideation stage 20 . ...
Creativity reflects the remarkable human capacity to produce novel and effective ideas. Empirical work suggests that creative ideas do not just emerge out of nowhere but typically result from goal-directed memory processes. Specifically, creative ideation is supported by controlled retrieval, involves semantic and episodic memory, builds on processes used in memory construction and differentially recruits memory at different stages in the creative process. In this Perspective, we propose a memory in creative ideation (MemiC) framework that describes how creative ideas arise across four distinguishable stages of memory search, candidate idea construction, novelty evaluation and effectiveness evaluation. We discuss evidence supporting the contribution of semantic and episodic memory to each stage of creative ideation. The MemiC framework overcomes the shortcomings of previous creativity theories by accounting for the controlled, dynamic involvement of different memory systems across separable ideation stages and offers a clear agenda for future creativity research.
... Moreover, cooperation between divergent and convergent thinking (Cropley, 2006;Wigert, Murugavel, & Reiter-Palmon, 2022), as well as associational processes and cognitive control (Beaty et al., 2018;Beaty, Benedek, Kaufman, & Silvia, 2015;Frith, Elbich, et al., 2021;Lloyd-Cox, Chen, & Beaty, 2022), is needed to account for both generative and evaluative phases of the creative process (Ivancovsky, Shamay-Tsoory, Lee, Morio, & Kurman, 2019;Runco & Chand, 1995). The recent attempts to synthesize empirical evidence elucidated how fluid and crystallized intelligence contribute to cognitive creative potential (Gerwig et al., 2021), as well as how various memory components, including semantic memory, episodic memory, working memory, and short-term memory, support idea generation (Gerver, Griffin, Dennis, & Beaty, 2022). Yet, to the best of our knowledge, so far no study endeavored to meta-analyze data on the individual contributions of broad retrieval ability (Gr) and processing speed (Gs) in DT performance. ...
... Second, the presence of meta-analyses on the relationship between DT and major cognitive abilities from the CHC framework (Gerver, Griffin, Dennis, & Beaty, 2022;Gerwig et al., 2021), and reliance on DT tasks as measures of creative potential Long, 2014;Long, Kerr, Emler, & Birdnow, 2022), might lead researchers to focus on more ecologically valid and domain-specific measures of creative potential and production. Currently, a limited number of studies explored the role of CHC abilities in specific creative domains, such as drawing and writing (Avitia & Kaufman, 2014;Smith et al., 2022;Taylor & Barbot, 2021), humor production (Kellner & Benedek, 2017), or jazz improvisation (Beaty, Smeekens, Silvia, Hodges, & Kane, 2013). ...
The present study aimed to integrate evidence on the relationship among broad retrieval ability (Gr), processing speed (Gs), and divergent thinking (DT) with a three-level meta-analytic approach. The analysis was conducted on 560 effect sizes obtained from 47 studies with an overall sample of 10,391 participants. Results indicated moderate mean correlations for both the Gr–DT (r = 0.47, 95% CI: [0.38, 0.54]) and the Gs–DT relationship (r = 0.31, 95% CI: [0.20, 0.41]). Notably, the correlation between DT and Gr was significantly higher than between DT and Gs, and the former remained significant even after controlling for the Gr–Gs correlation (r = 0.35, 95% CI: [0.26, 0.44]). Moderation analyses revealed that the Gr–DT link was moderated by the modality of DT tests and type of DT indicator, whereas the Gs–DT link was moderated by the modality of DT tests and type of DT instruction. Overall, these findings support the claim on the essential role of broad retrieval ability and processing speed in creative idea production.
... Analysis is a stepwise and logical problem solving strategy that stands in opposition to the suddenness of insight (Metcalfe & Wiebe, 1987;Newell & Simon, 1972), allowing us to examine whether insight problem solving is specifically linked to semantic memory structure. Altogether, our goal was to extend research on the role of memory in creative problem solving (Gerver et al., 2022) by exploring how semantic memory structure relates to CT ability and insight problem solving. ...
... The compound remote associates (CRA) test is a modern variation of the RAT, designed to illicit both insightful and analytical problem solving strategies (Bowden & Jung-Beeman, 2003a;Bowden et al., 2005;Jung-Beeman et al., 2004). Notably, people who more often use insight to solve CRA problems generally solve more problems than people who tend to use analysis (Ellis et al., 2021;Salvi et al., 2016). ...
... Individual differences research has revealed several cognitive factors that may support CT performance (Sawyer, 2011). For example, CRA accuracy is strongly predicted by crystallized intelligence-the ability to employ general knowledge (e.g., vocabulary) to solve problems (Cattell, 1963)-over and above other cognitive abilities (Ellis et al., 2021). More broadly, CT has been associated with working memory capacity and executive control (Chein & Weisberg, 2014;Ellis & Brewer, 2018), raising questions about the extent to which CT tasks (e.g., the RAT) measure intelligence or creativity (Lee & Therriault, 2013; see also Beaty, Nusbaum et al., 2014). ...
The associative theory of creativity has long held that creative thinking involves connecting remote concepts in semantic memory. Network science tools have recently been applied to map the organization of concepts in semantic memory, and to study the link between semantic memory and creativity. Yet such work has largely overlooked the domain of convergent thinking, despite the theoretical importance of semantic memory networks for facilitating associative processes relevant for convergent problem solving (e.g., spreading activation). Convergent thinking problems, such as the Compound Remote Associates (CRA) test, can be solved with insight (the sudden “aha” experience) or analysis (deliberately and incrementally working towards the solution). In a sample of 477 participants, we adopted network science methods to compare semantic memory structure across two grouping variables: 1) convergent thinking ability (i.e., CRA accuracy), and 2) the self-reported tendency to solve problems with insight or analysis. Semantic memory networks were constructed from a semantic fluency task, and problem solving style (insight or analysis) was determined from judgments provided during solving of CRAs. We found that, compared to the low-convergent thinking group, the high-convergent thinking group exhibited a more flexible and interconnected semantic network—with short paths and many connections between concepts. Moreover, participants who primarily solved problems with insight (compared to analysis) showed shorter average path distances between concepts, even after controlling for accuracy. Our results extend the literature on semantic memory and creativity, and suggest that the organization of semantic memory plays a key role in convergent thinking, including insight problem solving.
... Creativity has previously been associated with several cognitive processes (Benedek & Fink, 2019;Runco & Chand, 1995). A recent meta-analysis examining several decades of creativity research found a small but significant relationship between creativity and memory, including both working memory and long-term memory (Gerver et al., 2022). The authors concluded that the ability to efficiently retrieve information from long-term memory primarily influenced the relationship between creativity and memory. ...
... Although creativity has been conceptualized in many different ways, one prominent perspective considers creativity as associative thinking, or linking related concepts into original and useful ideas (Gerver et al., 2022;Mednick, 1962). Further refinement of the idea of creativity led to the delineation of two types of creativity: convergent thinking and divergent thinking (Guilford, 1950). ...
... Further refinement of the idea of creativity led to the delineation of two types of creativity: convergent thinking and divergent thinking (Guilford, 1950). Convergent thinking in particular has been linked to working memory (Gerver et al., 2022). In a typical convergent thinking task, the Remote Associates Test (Mednick, 1962;Mednick & Mednick, 1971), participants are asked to identify an item that is associated with other items. ...
While holding items in working memory has been shown to improve delayed long-term recall, the mechanisms underlying this relationship remain unclear. One potential mechanism is working memory consolidation, which may facilitate the formation of novel associations between items during learning and lead to improved memory search at delayed retrieval. Forming novel associations via consolidation may share mechanisms with creative ability. The present research aims to explore how an individual’s creativity relates to the relationship between working memory consolidation and long-term memory. In Experiment 1, participants completed a stimulus identification task that manipulated the need for consolidation followed by a surprise delayed recognition task and measures of objective and self-reported creativity. While creativity scores were correlated with general performance on memory tasks, this effect was not related to working memory consolidation. In Experiment 2, participants were induced into either a creative or a non-creative state prior to completing the stimulus identification and delayed recognition tasks. Performance on these tasks was not significantly different between the groups and was again unrelated to working memory consolidation. The results of these two experiments suggest that creativity is not related to the mechanism underlying the effect of working memory consolidation on delayed recognition.
Generating creative ideas takes time: the first idea to come to mind is usually obvious, and people need time to shift strategies, enact executive processes, and evaluate and revise an idea. The present research explored the role of time in creative humor production tasks, which give people a prompt and ask them to create a funny response. A sample of 152 young adults completed four joke stems prompts. Their response times were recorded, and the responses were judged for humor quality (funniness) by six independent judges and by the participants themselves. Mixed-effect models found that, at the within-person level, response time’s link to humor quality diverged for judges and participants. The judges’ ratings of funniness predicted longer response times (relatively funnier responses took longer to create), but participants’ self-ratings of their own responses predicted shorter response times (relatively funnier responses were created faster). Controlling for elaboration (quantified via word count of the response) diminished the effect of judge-rated humor but not participant-rated humor. Taken together, the results suggest that the role of time in humor generation is complex: judges may be weighting elaboration more heavily when judging funniness, whereas participants may be weighting metacognitive cues like ease-of-generation when judging their own ideas.
