Insight problem solving may have specific neural response patterns

Insight problem solving may have specific neural response patterns | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 9 January 2025 V1 Latest version Share on Insight problem solving may have specific neural response patterns Authors : Yan Chen 0009-0002-6273-1485 , Ying Li , Guanxiong Liu , Quanlei Yu , Zheng Liang , Shi Chen , and Qingbai Zhao [email protected] Authors Info & Affiliations https://doi.org/10.22541/au.173639730.03579860/v1 500 views 169 downloads Contents Abstract Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract Insight problem solving is the result of complex interactions of multiple cognitive activities and neural regulation. In addition, unconscious processing is considered important to insight problem solving, due to the difficulty in capturing, few studies have directly probed this process. The present research recruited 37 participants, recorded the EEG activities when they completed the compound remote association task. By comparing the microstates (the topographic maps formed by the clusters are used to reveal the cognitive processes that may occur at the millisecond level) of the insight, non-insight and unresolved condition in the different stages, the main results show that: (1) under the insight condition, microstate C(reflecting a part of the default mode network) shows a higher rate of occurrence and more frequent transition with microstate A(related to the cognitive process of speech information processing) and microstate D(reflecting attention and input, related to executive function). (2) microstate B (related to visual processing) occurs significantly more frequently at the beginning of both types of successful problem solving, but only in each stage of the non-insight condition. The current research results indicate that the time series of microstates corroborate the distinct neural responses observed under different problem-solving conditions at the electrophysiological level. The successful resolution of a problem depends on an adequate representation of the problem and the active participation of executive function. More importantly, when comparing the two conditions for successful problem solving, microstate C, associated with the default mode network, is only captured under the insight condition. This indicates that during insight problem solving, the brain may be engaging in unconscious processes that are not directly related to the current task. Insight problem solving may have specific neural response patterns Yan Chen a,b , Ying Li a,b , Guanxiong Liu a,b , Quanlei Yu a,b , Zheng Liang a,b,c,d,*, Shi Chen e, f,** , Qingbai Zhao a,b,* ** a Key Laboratory of Adolescent Cyberpsychology and Behavior (CCNU), Ministry of Education, Central China Normal University, Wuhan, China b Key Laboratory of Human Development and Mental Health of Hubei Province, School of Psychology, Central China Normal University, Wuhan, China c Department of Psychological and Cognitive Sciences, Tsinghua University, Beijing 100084, China d Tsinghua Laboratory of Brain and Intelligence, Tsinghua University, Beijing 100084, China e Hubei Health Industry Development Research Center, School of Medical Humanities, Hubei University of Chinese Medicine, Wuhan, China f Hubei Shizhen Laboratory, Wuhan, China All author information Qingbai Zhao (Corresponding author): [email protected] Shi Chen (Co-corresponding author): [email protected] Zheng Liang (Co-corresponding author): [email protected] Yan Chen (the first authorship): [email protected] Ying Li: [email protected] Guanxiong Liu: [email protected] Quanlei Yu: [email protected] Conflict of Interest statement The authors declare no competing financial interests. Abstract : Insight problem solving is the result of complex interactions of multiple cognitive activities and neural regulation. In addition, unconscious processing is considered important to insight problem solving, due to the difficulty in capturing, few studies have directly probed this process. The present research recruited 37 participants, recorded the EEG activities when they completed the compound remote association task. By comparing the microstates (the topographic maps formed by the clusters are used to reveal the cognitive processes that may occur at the millisecond level) of the insight, non-insight and unresolved condition in the different stages, the main results show that: (1) under the insight condition, microstate C(reflecting a part of the default mode network) shows a higher rate of occurrence and more frequent transition with microstate A(related to the cognitive process of speech information processing) and microstate D(reflecting attention and input, related to executive function). (2) microstate B (related to visual processing) occurs significantly more frequently at the beginning of both types of successful problem solving, but only in each stage of the non-insight condition. The current research results indicate that the time series of microstates corroborate the distinct neural responses observed under different problem-solving conditions at the electrophysiological level. The successful resolution of a problem depends on an adequate representation of the problem and the active participation of executive function. More importantly, when comparing the two conditions for successful problem solving, microstate C, associated with the default mode network, is only captured under the insight condition. This indicates that during insight problem solving, the brain may be engaging in unconscious processes that are not directly related to the current task. Key words : insight; problem solving; microstate; default mode network; unconscious processing 1 Introduction Problem solving is an important way for individuals to adapt to life, and it can test how our existing knowledge and experience are applied in changing situations. However, not all problems can be easily solved by previously established rules. Sometimes, individuals may be in a state of impasse until the problem is suddenly resolved, accompanying a positive emotional experience, this form of problem solving is known as insight problem solving (Scheerer, 1963; Kohler, 1985). Further illuminating the cognitive processes of insight problem solving and its neural mechanisms is highly necessary, because it is a special form of problem solving that involves complexity and multiple factors in the process. (Haavold & Sriraman, 2022). Like other problem solving, insight problem solving has general characteristics of problem solving, including understanding the problem, searching for information within the problem representation space according to familiar strategies, solving the problem, and evaluating the problem (Haavold & Sriraman,2022). In this process, the information most directly related to the problem in individual knowledge and experience is activated, and further activation and diffusion is selected according to the process of problem solving. Graphic problem solving is based on visual processing of physical features of the stimulus, and mainly activates visuospatial brain regions such as the parietal cortex, occipital cortex, posterior temporal cortex and cerebellum (Rominger et al., 2018; Lu & Singer, 2023; Saggar et al., 2015); Verbal problem solving mainly processed auditory or semantic information, inducing the activation of the middle temporal gyrus, superior temporal gyrus, and inferior frontal gyrus (Chen Shi et al., 2021; Aziz-Zadeh et al., 2009; Qiu et al., 2010). In addition, the executive function is responsible for the selection and integration of information in the information processing of problem solving (Lin et al., 2022; Beaty et al., 2015; Lloyd-Cox et al., 2022; Matheson et al., 2023). It can be seen that many types of problem solving share multiple cognitive processes based above the level of consciousness. However, unlike the conscious approach to a solution through logical analysis, the process of insight problem solving involves the formation and sudden breaking of a thinking impasse, which is usually spontaneous and unconscious (Ohlsson, 2011; Weisberg, 2015; Haavold & Sriraman,2022), insight often occur when people do not even realize they are thinking about it. At present, many researches use the method of setting distraction tasks or probes after thinking impasse to investigate the role of unconscious processing in insight problem solving (Sio et al., 2009; Leszczynski et al., 2017; Tan et al., 2015). Although these researches provide behavioral outcome support for unconscious processing promoting insight problem solving, they do not measure unconscious processing directly or indirectly, and insight in real-world situations is not always associated with distracted situations. EEG studies by Beeman et al. (2004) in more spontaneous (less intrusive) problem solving situations, using the compound remote association task (CRA), observed a sudden burst of high-frequency (gamma-band) neural activity 0.3 seconds before insight solution. It reflects a sudden shift in solution-related cognitive processing from unconscious to conscious states. However, before this sudden shift, how did unconscious processing change dynamically with the problem-solving process? Beeman et al. ’s research does not respond to this question. In recent years, more researches have looked at insight as a comprehensive time sequence with complex ingredients. First, insight problem solving involves multiple cognitive processes. From the perspective of content of information processing, it includes the processing of strongly related information above the level of consciousness and weakly related information below the level of consciousness (Zhao et al., 2015; Shen et al., 2018), in which executive function plays a top-down regulatory role. Secondly, from the perspective of the time course of problem solving, the process of insight problem solving is a nonmonotonic time series. When trying to solve a problem, individuals often spontaneously iterate through understanding the problem, searching the problem space, evaluating ideas, or discovering a solution (Yu et al., 2022). More importantly, the process of solving the insight problem is uncontrollable under the natural state (Luo Jin, 2004; Benedek et al., 2019), from impasse breaking to problem solving may only take a short time to complete, indicating that the process of insight problem solving is a ”slow before fast” and dynamic process. At the moment of insight, the brain may deploy cognitive processes in complex ways for a short period. Therefore, it is necessary to understand how multiple cognitive processes rapidly change and combine to support insight problem solving on a more precise temporal scale. ERP is widely used to explore the cognitive process of insight problem solving because of its high temporal resolution. Previous researches have mainly explored the possible specific components of insight problem solving. It has been found that successful insight problem solving may reflect the P2 component, which appears earlier in the stage of thought impasse and is thought to be related to metacognition, where early metacognitive alertness regulates subsequent attention allocation processes (Shen et al., 2012). In addition, insight problem solving induces a negative ERP component (N320, N350, N380, N400) in the fronto-central region of the scalp between 300-500 milliseconds, reflecting a greater cognitive conflict between old and new thinking by locating the source in the anterior cingulate gyrus (Mai et al., 2004; Wang et al., 2009; Qiu et al., 2008). At a relatively late stage, insight problem solving was observed to induce a late positive component LPC/P600 in the right frontal and temporal regions. This component is generally thought to reflect semantic information integration (Brouwer & Hoeks, 2013). In creative verbal tasks (including insight problem solving), this component is located in the parhippocampal gyrus or distributed in the right temporal region, which is thought to reflect the formation of novel semantic associations (Qiu & Zhang, 2008; Zhao et al., 2015, Zhao et al., 2017). These results support the relevant theories of insight problem solving and provide electrophysiological evidence for the existence of key cognitive processes in insight problem solving. However, these researches are mainly concerned with the chain process of insight problem solving, such as from problem representation to the formation of thinking impasse, or from the impasse formation to breaking. The previous ERP results cannot explain the basic view that insight problem solving is an iterative and evolutionary process (Huang et al., 2019; Shen et al., 2018; Bilalić et al., 2021). For such a complex cognitive activity as insight problem solving, it is important to interpret the dynamic changes of insight problem solving from a broader perspective, rather than limiting the role of a certain cognitive process. One research used the CRA task (which has often been used in previous researches to investigate verbal insight problem solving) to fit the Hidden Markov model (HMM) to the time-frequency transformed EEG signals, and the continuously changing EEG signals were divided into seven topographic maps representing different brain states in different frequency bands. Iterations between states are also shown on a time series (Yu et al., 2022). The results show that alpha band plays an important role in insight problem solving, and this role will change dynamically with the time course of problem solving, confirming the previous conclusions about the role of alpha band in insight problem solving and even creative problem solving (Fink et al., 2009; Zhou et al., 2018) described the changes of some representative frequency bands in insight problem solving from a more overall perspective. However, this research has yet to reveal the corresponding relationship between the seven topographic maps and brain regions or brain networks related to insight problem solving, thus failing to clarify their cognitive functions. The cognitive processes involved in insight problem solving and their interactions cannot be explained more intuitively. The EEG microstate analysis can be used as a tool to investigate these problems. Besides the advantage of time, microstate analysis also provides certain spatial information. According to microstate, the spontaneous activity of the brain is not completely continuous and random, but can be divided into discrete topographic maps with certain cognitive significance in a relatively short time (Lehmann et al., 1987; Koenig et al., 2002; Lehmann et al., 2009). Microstate analysis is generally divided into task state microstate analysis and resting state microstate analysis. Task state microstate analysis is usually combined with event-related potentials. After identifying ERP components, topographic maps of different components can be analyzed to investigate the stability and other features of ERP components in a certain cognitive process (Schiller et al., 2016). Microstate analysis in the resting state classifies all time points into several categories by way of clustering (the most common number of clusters is 4-6 categories) and is expressed in the time series. In general, the topographic maps formed by these clusters are highly similar across studies and can explain about 80% of the variation in EEG activity. Many studies have proved that there is a certain correspondence between microstate topology and large-scale resting state brain networks obtained from fMRI (Britz, 2010; Musso, 2010; Betzel, 2012). For example, microstate A is associated with BOLD activation in the bilateral superior temporal gyrus and middle temporal gyrus, which are critical in speech processing and reflect resting auditory networks; microstate B is involved in the processing of representational thinking and reflects the resting visual network. Microstate C is generally associated with activation of the angular gyrus and internal parietal sulcus, and is thought to reflect part of the default mode network, while microstate D is associated with BOLD activation in the right frontal and parietal cortex dorsal and ventral regions, possibly reflecting cognitive processes related to attention and control functions. Microstate E is associated with the salience network, plays a role in certain cognitive functions, and is involved in the processing of interoceptive and emotional information… the known topographic maps with more studies include but are not limited to the above categories, which need to be further analyzed and judged according to the data clustering. At present, there are also researches applies resting-state microstate analysis to task-initiated spontaneous thinking, comparing the microstates induced by autobiographical memory and computational tasks (Bréchet et al., 2019).This kind of research has extended the application of resting state microstates properly, facilitating the study of the dynamic organization of brain networks on time series in millisecond, especially for the tasks that require spontaneous thinking under certain guidance for a long time, such as the compound remote association test (CRA) in verbal insight problem solving. In this task, participants are given three words to think of a word that can be combined with all three words to form a phrase or compound (Bowden and Jung-Beeman, 2003). In view of the correlation between topographic map in microstate analysis and cognitive processes, such as information processing and executive control in insight problem solving, this technique can be used to explore the dynamic changes of various cognitive processes and their interactions in insight problem solving. In particular, microstate C reflects part of the default mode network, which is thought to represent the unconscious processing of insight problem solving that departs from the attention of the current task (Darsaud et al., 2011; Baird et al., 2012; Ritter & Dijksterhuis, 2014; Beaty et al., 2015), therefore, the topographic map (microstate C) may provide a window for us to explore the role of unconscious processing in insight problem solving and its dynamic changes. Based on this, the current study uses CRA testing as a task to explore the interaction patterns of unconscious processing and multiple cognitive activities in spontaneous insight problem solving. The EEG activity of the participants while attempting to solve CRA was recorded, and the differences in topographic maps of different answer types (insight, non-insight, and unsolved) at three time periods (early, middle, and late) were analyzed using microstate. We expect that there are various cognitive processes involved in the three answer types, but the participation weights and conversion modes of various cognitive processes are different among conditions. Insight problem solving may exhibit specific processing patterns in time series, especially the continuous participation of microstate C related to the default mode network. 2 Methods 2.1 participant The sample size required for the experiment was calculated according to G-power, assuming that the effect size was 0.25, the α level was 0.05, the statistical power was 0.8, and the total sample size of repeated measurement ANOVA was 28. Considering that previous studies using the CRA task have shown that this task is difficult(Bowden & Beeman, 2003; Beeman et al., 2004; Yu et al., 2022), our research recruited as many participants as possible, 37 right-handed volunteers (average age 21.2 years old, 17 females) who spoke Chinese as their mother tongue and English as their second language, reached certain standards(IELTS≥7 or TOEFL≥95 or the major of study at the university is English, and TEM4≥80) to participate in this experiment, and signed informed consent before the experiment and receive a cash reward upon completion. 2.2 Materials We used the compound distance association (CRA) test as the experimental material (Bowden & Beeman, 2003), which requires participants to come up with a word (e.g. cheese) that can be combined with three words on a screen (e.g. cottage, swiss, cake) to form a phrase or compound (e.g. swiss cheese, cake-cheese, cottage cheese). This material has been widely used to explore verbal insight problem solving (Beeman et al., 2004; Kounios et al., 2006; Yu et al., 2022). 2.3 Experiment procedure Fig.1 is a schematic diagram of a trial as an example. First, the screen will display ”Ready?” An empty screen with a ”+” will be presented for 1s to focus the participants’ attention. Then the participants will see three problem words presented parallelly on the screen, and are asked to generate a solution word, which can be combined with the three problem words to form a familiar compound word or phrase. Each problem can take up to 15s to solve, and the time limit is determined by previous studies (Erickson et al., 2018; Yu et al., 2022). Participants were asked to press the button immediately after coming up with an answer and report the method by which the solution to the problem was accomplished (insight/non-insight). In the subsequent analysis, we categorized all answers into insight and non-insight solutions based on the participants’ subjective reports. If the participant does not response within 15s, it is considered unsolved, and the question will disappear and be replaced by an empty screen with a ”+” and proceed to the next question after 1s. The experiment consisted of 10 practice trials and 134 formal trials. Fig.1. Example of a CRA trial 2.4 Data acquisition and preprocessing The EEG was recorded with the BrainAmp MR64 system (Brain Products, Münich, Germany) with 64 electrodes embedded in an elastic cap using the extended International 10–20 system. The positions of grounding electrode GND and online reference electrode REF are AFz and FCz respectively. The data were sampled at 500 Hz. Using the eeglab toolbox 2024.0 version to run on Matlab 2023b, the preliminary channel location of data and the removal of useless electrodes are carried out. Subsequently, the data were preprocessed using Happe (Lopez et al., 2022), a preprocessing Pipeline developed by Harvard University (GitHub-PINE-Lab /HAPPE: EEG Pre-Processing Pipeline). Specifically, we downsample the data to 250Hz, filter it with a band-pass filter with a high-pass filter of 40Hz and a low-pass filter of 1Hz, and filter the noise at 50Hz with a notch filter (the Cleanline method). Bad tracks and artifacts are detected, the rejected channels are interpolated and the whole brain average is performed off-line re-reference, and the data segments determined to be artifacts are deleted. The continuous data is segmented to extract the time window of interest: (1)T1: the problem present from 0 to 2s; (2)T2: 2~5s after the problem presents; (3) T3:1s before the reaction to 0.5 after the reaction. The response time is less than 6s of the test is excluded, to avoid data overlap sampling between the stages. The amplitude of ±100μV was used as the standard for wavelet threshold denoising, and independent component analysis (ICA) was used to further remove the artifacts on the segmented data. Finally, the data is visually examined, and segments with residual artifacts are flagged, and excluded from further analysis. 2.5 Microstate analysis For the purpose of the experiment, we did not pay attention to the ERP components that could be isolated in a short period. Therefore, this study referred to the idea of Bréchet et al. (2019) to analyze task-based spontaneous thinking by using the EEG microstate resting state method. To observe the temporal dynamics of states represented by specific mental activities in verbal insight problem solving (CRA test). In this study, the eeglab microstate analysis toolbox based on the Matlab programming platform is used (Poulsen et al., 2018) (https://github.com/ atpoulsen/ microstate-eeglab-Toolbox) to perform Microstate analysis of the pre-processed data. The toolkit is fully transparent to all steps of the analysis and allows the integration of any clustering algorithm. The toolkit consists of a set of features that can be used and modified independently or as an interactive plug-in to the widely used open-source EEG analysis software EEGLAB (Delorme & Makeig, 2004). First, we use k-means to cluster the data of different time Windows under various conditions. K-means methods typically require the number of clusters (or microstate categories) to be set in advance, and then the EEG data is divided into a fixed number of clusters. The EEG data is then iteratively repositioned into these clusters and the Global explained variance of the clustering results is calculated until the optimal cluster assignment is achieved (Rokach & Maimon, 2010). In this study, 4-6 initial topographic map classes are set, and the topographic map at the time point with the highest SNR is regarded as the only topographic map in a given time by GFP peak clustering method. The optimal number of clusters was selected by comparing the GEV (global explained variance) of 4-6 topographic maps. After microstate analysis of each participant’s data for each condition, the data were combined to complete group-level analysis. There are individual-level clustering topographic map templates and group-level maximum mean clustering topographic map templates to choose from. The maximum mean template measures all data content in the same pattern, but this pattern may not be fully applicable to the problem of exploring the characteristics of individual data. Individual templates can better restore the detailed characteristics of individual data, but there may be large differences between participants in micro-state templates, which may cause the problem of difficult comparison across participants. The purpose of this study is to explore the cognitive process of problem-solving. Individual differences are not the focus of our research, so we choose the maximum mean template, use the four types of topographic maps formed by group-level clustering as the template, and reverse-fit the data of all participants. To explore the cognitive processes that these microstate categories may reflect, as well as the corresponding physiological basis (Koenig et al., 2002; Tarailis et al., 2024; Michel et al., 2024; Michel & Koenig, 2018; Khanna et al., 2015), SPSS 27.0 was used for statistical analysis of the mean duration, occurrence rate per second, and transition probabilities in the time parameters of microstate. 3 Results 3.1 Behavioral results The data of 12 participants were excluded (the data of 2 participants had more artifacts, and the answers of 10 participants were insufficient in the number of available trials under each condition, ≤10), the data of 25 participants were included in the subsequent analysis. An average of 109.36 attempts were included in the analysis ( sd =15.92), of which 40.88 failed attempts ( sd = 25.40), 31.68 enlightened attempts ( sd = 17.44), and 36.80 non-enlightened attempts (sd= 14.37) were included. The average correct rate of all subjects was 51% ( sd =13%), 83% ( sd =14%) for insight solution, and 70% ( sd =14%) for non-insight solution. The average accuracy of insight solution was significantly higher than that of non-insight solution condition t (24)=4.925, p <.001, Cohen’s d =0.97. The mean response time of insight solution was 8136.97ms ( sd =1664.90ms) and that of non-insight solution was 8992.42ms ( sd =1134.21ms). The mean reaction time of non-insight solution was significantly longer than that of insight solution t (24)=3.223, p =.004, Cohen’s d =0.65. 3.2 Microstate analysis results According to GEV (global explained variance), our research identified four types of topographic microstates with the best data fitting effect (mean interpretation rate 71.47%). We carried out repeated measurement ANOVA on the results of the mean duration, occurrence rate, and transition probabilities of the microstates of microstate clustered under the three answer types. The results of each parameter are as follows. 3.2.1 Duration The results of repeated measurement ANOVA showed that the main effect of answer type(insight/non-insight/miss) was not significant F (2,48)=3.001, p =.059. The main effect of time window(T1/T2/T3) was significant, F (2,48)=7.613, p =.001, partial η 2 =0.24. The main effect of microstate (microstate A/ microstate B/ microstate C/ microstate D) was significant, F (3,72)=3.025, p =0.035, η 2 =0.112(Table.1). Table 1. Interaction of different conditions on the mean duration AT*TW 4 1.132 .346 .045 AT*microstate 6 5.915 <.001 .198 TW*microstate 6 15.476 <.001 .392 AT*TW*microstate 12 11.561 <.001 .325 AT: answer type TW: time window The interaction between the answer type and the time window was not significant, F (4,96)=1.132, p =.346. The interaction between the answer type and the microstates was significant, F (6,144)=5.915, p <.001, η 2 =.198; the interaction between microstates and the time window was significant, F (6,144)=15.476, p <.001, η 2 =.392; the interaction among the answer type, the time window and the microstates was significant, F (12,288)=11.561, p <.001, η 2 =.325(Fig.2). Simple simple effects analysis indicated that in the problem presentation stage(T1), the mean duration of microstate A in the unresolved condition(miss) was significantly longer than that in the insight condition ( p =.002), and also significantly longer than that in the non-insight condition( p =.003). The mean duration of microstate B in the insight condition was significantly longer than that in the unresolved condition( p <.001), and the mean duration of microstate B in the non-insight condition was significantly longer than that in the unresolved condition( p =.002). The mean duration of microstate D in the non-insight condition was significantly longer than that in the unresolved condition( p =.040). In the solution process stage(T2), the mean duration of microstate A in both the insight condition( p <.001) and the non-insight condition( p =.026) was significantly longer than that in the unresolved condition, while the mean duration of microstate B in this stage was significantly shorter in the insight condition( p =.027) and the non-insight condition( p =.030) than that in the unresolved condition. In the response stage(T3), the mean duration of microstate A in the insight condition was longer than that in the non-insight condition( p =.016), and also longer than that of the unresolved condition( p <.001). The mean duration of the non-insight condition was longer than that in the unresolved condition ( p =.007). The mean duration of microstate B in the non-insight condition was longer than that in the insight condition( p <.001), and the mean duration of microstate B in the unresolved condition was longer than that in the insight condition( p <.001). Fig.2. The differences in the mean duration of the microstates under the answer types 3.2.2 Occurrence rate The results of repeated measurement ANOVA showed that the main effect of answer type was significant, F (2,48)=3.448, p =.039, η 2 =.127; The main effect of time window was significant, F (2,48)= 8.351, p <.001, η 2 =.258; The main effect of microstates was significant, F (3,72)= 7.756, p <.001, η 2 =.244(Table.2). AT * TW 4 1.155 .336 .046 AT *microstate 6 11.553 <.001 .325 TW *microstate 6 16.203 <.001 .403 AT*TW*microstate 12 8.279 <.001 .256 Table 2. Interaction of different conditions in frequency of occurrence rate AT: answer type TW: time window The interaction between answer type and time window is not significant, F (4,96)= 1.155, p =.336. The interaction between answer type and microstates was significant, F (6,144)= 11.553, p <.001, η 2 =.325; The interaction between time window and microstates is significant, F (6,144)= 16.203, p <.001, η 2 =.403; The interaction of answer type, time window and microstates was significant, F (12,288)= 8.279, p <.001, η 2 =.256(Fig.3). Simple simple effect analysis showed that the occurrence rate of microstate A in the unresolved condition was significantly higher than that in the insight condition( p <.001) and non-insight condition( p =.001) at the problem presentation stage. The occurrence rate of microstate B in the insight condition was significantly higher than that in the unresolved condition( p =.004), the occurrence rate of microstate B in the non-insight condition was significantly higher than that in the unresolved condition ( p =.002), and the occurrence rate of microstate C in the unresolved condition was significantly higher than that in the insight condition( p =.014) and the non-insight condition( p =.014). The occurrence rate of microstate D with insight condition was significantly higher than that in the unresolved condition( p =.009), and the occurrence rate of microstate D in the non-insight condition was significantly higher than that in the insight condition( p =.001). In the process of solution, the occurrence rate of microstate A in the insight condition was significantly higher than that in the non-insight condition ( p =.005), the occurrence rate of microstate A in non-insight condition was significantly higher than that in non-insight condition( p =.001), and the occurrence rate of microstate B in non-insight condition was higher than that in insight condition( p =.009). The occurrence rate of microstate C in the insight condition was higher than that in the non-insight condition( p =.006), and the occurrence rate of microstate C in the unresolved condition was higher than that in the non-insight condition( p <.001). In the reaction stage, the occurrence rate of microstate A in the insight condition was significantly higher than that in the non-insight condition ( p =.017), the microstate B in the non-insight condition was significantly higher than that in the insight condition( p <.001), and the microstate B in the unresolved condition was significantly higher than that in the insight condition ( p =.009). The occurrence rate of microstate C in the insight condition was higher than that in the non-insight condition( p <.001), and the occurrence rate of microstate C in the unresolved condition was higher than that in non-insight solution condition ( p <.001). Fig.3. The difference in the occurrence rate of microstates under the answer types 3.2.3 Transition probability 3.2.3.1 Transitions from microstate A The results of repeated measurement ANOVA showed that the main effect of answer type was significant, F (2,48)=4.008, p =.040, η 2 =.143; The main effect of time window was significant, F (2,48)= 27.170, p <.001, η 2 =.531; The main effect of transition probability is significant, F (3,72)= 12.071, p <.001, η 2 =.335(Table.3). AT* TW 4 13.981 <.001 .368 AT * TT 4 11.999 <.001 .333 TW * TT 4 7.788 <.001 .245 AT * TW * TT 8 3.761 <.001 .135 Table 3. The interaction of different conditions on microstate A to transform other microstates AT: answer type TW: time window TT: transition type(microstate A to microstate B, microstate A to microstate C…) The interaction between answer type and time window was significant, F (4,96)= 13.981, p <.001, η 2 =.368; The interaction between answer type and transition probability is significant, F (4,96)= 11.999, p <.001, η 2 =.333; The interaction between time window and transition probability was significant, F (4,96)= 7.788, p <.001, η 2 =.245; The interaction of answer type, time window and transition probability was significant, F (8,192)= 3.761, p <.001, η 2 =.135(Fig.4). Simple simple effect analysis showed that P A-C in the unresolved condition was higher than that in the insight condition ( p <.001) and non-insight condition ( p <.001) in the problem presentation stage. In the process of solution, the non-insight condition had significantly higher P A-B ( p =.002) and P A-D ( p =.009) than that in the unresolved condition, and the insight condition had significantly higher P A-C ( p =.036) than that in the non-insight condition. In the reaction stage, P A-B in the non-insight condition was significantly higher than that in the insight condition ( p <.001) and unresolved condition ( p =.004), and P A-C in the insight condition was significantly higher than that in the non-insight condition ( p <.001) and unresolved condition ( p =.029). The insight condition has significantly higher P A-D than that in the unresolved condition ( p =.011), and the non-insight condition has significantly higher P A-D than that in the unresolved condition ( p =.013). Fig.4. Transitions from microstate A 3.2.3.2 Transitions from microstate B The results of repeated measurement ANOVA showed that the main effect of answer type was significant, F (2,48)=15.557, p <.001, η 2 =.393; The main effect of time window was significant, F (2,48)= 25.730, p <.001, η 2 =.517; The main effect of transition probability is significant, F (2,48)= 9.768, p <.001, η 2 =.289(Table.4). Table 4. The interaction of different conditions on microstate B to transform other microstates AT * TW 4 12.142 <.001 .336 AT * TT 4 7.197 <.001 .231 TW * TT 4 11.816 <.001 .330 AT * TW * TT 8 2.335 .048 .089 AT: answer type TW: time window TT: transition type The interaction between answer type and time window was significant, F (4,96)=12.142, p <.001, η 2 =.336; The interaction between answer type and transition probability is significant, F (4,96)= 7.197, p <.001, η 2 =.231; The interaction between time window and transition probability is significant, F (4,96)= 11.816, p <.001, η 2 =.330; The interaction of answer type, time window and transition probability is significant, F (8,192)= 2.335, p =.048, η 2 =.089(Fig.5), the simple simple effect analysis shows that during the problem presents stage, P B-D in the insight condition ( p =.002) and non-insight condition ( p =.004) were significantly higher than that in the unresolved condition. In the process of solution, P B-A in the non-insight condition was significantly higher than that in the unresolved condition ( p =.010), P B-C in the unresolved condition was significantly more than that in the insight condition ( p =.014) and non-insight condition ( p =.002), and P B-D in the non-insight condition was significantly higher than that in the insight condition ( p =.013). In the reaction stage, P B-A in the non-insight condition was significantly higher than that in the insight condition ( p =.003) and the unresolved condition ( p =.032), and P B-C in the unresolved condition was significantly more than that in the insight condition ( p =.010) and the non-insight condition ( p =.005). P B-D was significantly higher in the non-insight solution than that in the insight condition ( p =.003). Fig.5. Transitions from microstate B 3.2.3.3 Transitions from microstate C The results of repeated measurement ANOVA showed that the main effect of answer type was significant, F (2,48)= 22.226, p <.001, η 2 =.481; The main effect of time window was significant, F (2,48)= 13.940, p <.001, η 2 =.367; The main effect of transition probability is significant, F (2,48)= 8.220, p <.001, η 2 =.255(Table.5). Table 5. The interaction of different conditions on microstate C to transform other microstates AT * TW 4 3.508 .028 .128 AT * TT 4 .0444 .777 .018 TW * TT 4 10.961 <.001 .314 AT * TW * TT 8 6.871 <.001 .223 AT: answer type TW: time window TT: transition type The interaction between answer type and time window was significant, F (4,96)= 3.508, p =.028, η 2 =.128; The interaction between answer type and transition probability is not significant, F (4,96)=.0444, p =.777; The interaction between time window and transition probability is significant, F (4,96)= 10.961, p <.001, η 2 =.314; The interaction of answer type, time window and transition probability was significant, F (8,192)= 6.871, p <.001, η 2 =.223(Fig.6). Simple simple effect analysis showed that during the problem presents stage, the P C-A in the unresolved condition was higher than that in the insight condition( p =.001) and non-insight condition( p <.001). The P C-B in the insight condition was higher than that in the non-insight condition ( p =.043). In the process of solution, the P C-A in the insight condition was higher than that in the non-insight condition ( p =.014), P C-B in the unresolved condition was higher than that in the insight condition( p =.013) and the non-insight condition( p =.003), and P C-D in the insight condition was higher than that in the non-insight condition ( p =.040). There were higher P C-D in the unresolved condition than that in the non-insight condition ( p <.001). In the reaction stage, P C-A in the insight condition was higher than that in the non-insight condition ( p =.002), P C-D in the insight condition was higher than that in the non-insight condition ( p =.004), and P C-D in the unresolved condition was higher than that in the non-insight condition ( p <.001). Fig.6. Transitions from microstate C 3.2.3.4 Transitions from microstate D The results of repeated measurement ANOVA showed that the main effect of answer type was significant, F (2.48)= 3.395, p =.042, η 2 =.124; The main effect of time window was not significant, F (2,48)= 2.389, p =.103; The main effect of conversion probability is not significant, F(2,48)= 2.726, p =.076(Table.6). AT * TW 4 3.239 .038 .119 AT * TT 4 9.729 <.001 .288 TW * TT 4 21.973 <.001 .478 AT * TW * TT 8 8.441 <.001 .260 Table 6. The interaction of different conditions on microstate C to transform other microstates AT: answer type TW: time window TT: transition type The interaction between answer type and time window was significant, F (4,96)= 3.239, p =.038, η 2 =.119; The interaction between answer type and transition probability is significant, F (4,96)= 9.729, p <.001, η 2 =.288; The interaction between time window and transition probability is significant, F (4,96)= 21.973, p <.001, η 2 =.478; The interaction of answer type, time window and transition probability was significant, F (8,192)= 8.441, p <.001, η 2 =.260(Fig.7). The simple simple effect analysis showed that the P D-B in the insight condition was higher than that in the unresolved condition( p <.001) during the problem presents stage. There were higher P D-B in the non-insight condition than that in the unresolved condition( p <.001). In the process of solution, the P D-A in the insight condition was higher than that in the unresolved condition( p =.033), P D-A in the non-insight condition was higher than that in the unresolved condition( p =.004), and P D-B in the non-insight condition was higher than that in the insight condition( p <.001). There were higher P D-B in the unresolved condition than that in the insight condition( p =.004), higher P D-C in the insight condition than in the non-insight condition( p =.028), and higher P D-C in the unresolved condition than that in the non-insight condition( p =.040). In the reaction stage, P D-A in the insight solution condition was more than that in the unsolved condition ( p =.012), P D-A in the non-insight condition was higher than that in the unresolved condition( p =.015), and P D-B in the non-insight condition was higher than that in the insight condition( p =.003). There were higher P D-C in the insight condition than in non-insight condition( p =.002), and higher P D-C in unresolved condition than that in the non-insight condition ( p =.004). Fig.7. Transitions from microstate D 4 Discussion Fig.8. The dynamic changes of microstates during different time windows in the three conditions Note: NS: unresolved, NI: non-insight solution IS: insight solution; T1: the problem presentation stage, T2: the process of solution, T3: the reaction stage; the green outer ring of microstate means a significantly higher rate of occurrence, and the thicker the ring indicates the higher statistical significance ( p <.05, p <.01, p <.001), the red arrow represents statistically significantly higher transition probability from a microstate to the other. In order to describe the interaction of multiple cognitive processes during insight problem solving, we adopted CRA as a verbal insight problem solving task, and compared the time series under different solving conditions (answer types) by EEG microstate analysis, focusing on whether insight problem solving involves a specific cognitive process, such as unconscious processing that is detached from the current task. EEG microstate results show that successful problem solving is associated with continuous engagement of executive function, whereas insight problem solving is more unconscious processing in the middle and later stages of the process than non-insight problem solving. The present results confirm previous studies on the role of executive function network and default mode network in insight problem solving (Kounios et al., 2006; Lin et al., 2022; Beaty et al., 2015; Lloyd-Cox et al., 2022; Matheson et al., 2023), and further describes how multiple cognitive processes interact over time series of problem solving. This research identified four types of topographic microstates with the best data fitting effect, which corresponded to the microstate topographic microstates reported in several previous studies (Britz et al., 2010; Michel & Koenig, 2018; Croce et al., 2017; Pan et al., 2020; Tarailis et al., 2024) confirmed the consistency and reliability of data set results. Based on the comparison of results showing statistical differences, we provided a schematic diagram of the response patterns of microstates under the three answer types (Fig.8). The diagram includes the occurrence rate and transition probability of the four typical microstates in different time windows, in order to more clearly illustrate the differences in the neural response patterns of the three answer types. First, compared with the successfully solved condition, the microstate topography under the unresolved condition showed a significantly higher occurrence rate of microstate C in the three time windows, and increased with the time course of problem solving. Microstate C is a topographic map with relatively symmetrical front and back. Previous studies reported the activity of microstate C in the region partially overlapping with DMN, suggesting that microstate C is related to task-negative thinking, self-related thinking, and emotion and interoception processing (Bréchet et al., 2019; Custo et al., 2017; Croce et al., 2018). Therefore, according to the psychological function of microstate C, we believe that the dominant position of microstate C related to the default mode network in the three time windows indicates that one of the main factors for the failure to solve the problem is the lack of participants’ top-down attention resources investment in the problem. In addition, the unresolved condition exhibited microstate A with longer mean duration and higher rate of occurrence during the stage of problem presentation, and microstate B with longer duration during the process of problem solving and the response stage. Previous studies reported that microstate A was a topographic map with a right-frontal to left-posterior configuration (Koenig et al., 1999, 2002), and the EEG source localization mainly showed the activation of temporal cortex and auditory network (Bréchet et al., 2019; Britz et al., 2010; Custo et al., 2017), microstate B presents a topographic map with a right-frontal to left-posterior configuration (Koenig et al., 2002), and is obviously observed in activities involving visual space and image processing (Britz et al., 2010; Custo et al., 2017; Zappasodi et al., 2019). The CRA task requires participants to make long-distance associations based on three given words. microstate A, which represents auditory information processing, has a longer duration and higher rate of occurrence in the stage of problem presenting, which may mean that participants have carried out more in-depth processing of the pronunciation and semantic information of the words themselves at the auditory level. The lack of activation and diffusion according to the presentation form of the stimulus material (visual), and the excessive attention to the pronunciation and semantic information of the words themselves affect the judgment of the connections between the words, and thus affect the problem solving. In both successful problem solving, microstate D demonstrated a relatively longer mean duration and a higher rate of occurrence in the stage of problem presentation. This finding is consistent with previous research results regarding the significant role of executive functions in problem-solving (Lin et al., 2022; Beaty et al., 2015; Lloyd-Cox et al., 2022), indicating that successful problem-solving relies on the regulation of information processing by executive functions. In the middle and later stages of problem-solving, microstate D contributes to problem-solving mainly through transitions with other microstates. This might imply that executive functions do not operate in problem-solving by taking a dominant position or increasing the activation tendency. Instead, they achieve more effective regulation of information processing through multiple and frequent transitions, thereby facilitating problem-solving. Despite both being successful problem-solving, insight and non-insight conditions still exhibit very different neural representation patterns in time series (Beeman et al., 2004; Kounios et al., 2006; Yu et al., 2022; Bieth et al., 2024). In the stage of problem presentation, the response patterns of insight and non-insight conditions were similar, both showing longer microstate A mean durations and higher microstate B occurrence rate. As the course progressed, it was observed that the mean duration of microstate A decreased in the non-insight condition, while the mean duration of microstate B increased, and the occurrence rate of microstate B peaked in the response stage( p <.001). Microstate B represents cognitive processes related to visual attention and imagery; thus, it can be inferred that the gradual activation enhancement of microstate B may represent the participants’ effective activation diffusion of information based on the stimulus presentation form (visual). Through Fig.8, it can be intuitively seen that microstate B and microstate D have higher participation weights in non-insight condition, while microstate C does not show mean duration, occurrence rate or transition probability advantages in any stage of non-insight condition. These results may reflect the active regulatory role of executive function in information processing content, while unconscious processing plays an inappreciable role in non-insight problem solving. Our research results are consistent with those of previous problem-solving researches on non-insight problem solving (or analytical problem solving) [Beeman et al., 2004; Kounios et al., 2006; Yu et al. (2022)]. It is worth noting that although microstate B did not show the same level of effectiveness as microstate A, microstate A also played a certain role in non-insight problem solving. Non-insight problem solving was characterized by frequent bidirectional microstate transition patterns between microstate A, microstate B, and microstate D in the middle and later stages of the microstate, indicating that auditory semantic information processing was not dominant, but it played a significant role in problem solving by interacting with visual and executive function-related states at a higher transition probability. The significant dependence between microstate A and microstate B has also been reported in previous studies (Kleinert et al., 2023), indicating the importance of perceptual information processing in non-insight problem solving. Furthermore, since non-insight problem solving is supported throughout by perceptual information processing systems and executive function systems, and lacks the involvement of microstate C, it further indicates that non-insight problem solving occurs above the consciousness. Both microstate B and microstate D have a significantly higher rate of occurrence in the stage of problem presentation, which may mean that in the early stage of insight problem, participants rely more on visual processing to represent the words presented in the task, and executive function plays a significant regulatory role in the early stage of problem solving. However, based on the statistical analysis results of mean duration and occurrence rate, we found that in the middle and later stages of insight problem solving, compared with the unresolved and non-insight conditions, the mean duration and occurrence rate of microstate B in insight condition continued to decrease significantly in the three time windows. At the same time, the mean duration and occurrence rate of microstate A in the three time windows continue to increase significantly. We believe that this may reflect the change in the mode of information processing in insight problem solving, from mainly relying on visual mode to represent the problem in the early stage of the problem, to the processing mode related to auditory semantic information gradually occupying a dominant position until the problem is solved by insight. We interpret this specific change mode as indicating that after effective initial representation of the problem, processing problem information at a more abstract level (or more detached from the modality of presentation of the stimulus material) facilitates the realization of problem solving. In addition, the neural corresponding pattern of insight problem solving in time series not only relies on executive function to play a regulatory role, but also has the activation enhancement related to the default mode network. Specifically, in the initial stage of insight problem solving conditions, there is a significantly higher transition between microstate B and microstate D, and a significantly higher transition on microstate C to microstate B, while in the problem solving and reaction stage, there is a frequent transition between microstate A, microstate C and microstate D, and the occurrence rate of microstate C increases gradually within the three time windows. The default mode network is generally considered to be task-negative and is more active in the resting state of the brain or when there is no clear indication of a task (Dohmatob et al., 2020; Yeshurun et al., 2021). Previous studies have reached a unified conclusion on whether unconscious processing exists in insight problem solving, but few studies have directly explored how unconscious processing operates in the process of insight problem solving, and whether unconscious processing is mutually exclusive with other cognitive processes. For this, our findings have provided a clear, but relatively rough response. Microstate C, which is not related to the task and represents part of the default mode network, is dominant in insight-oriented problem solving. This increase in activity associated with the default mode network represents an increase in the weight of unconscious processing and is gradually enhanced over time, reaching its strongest in the reaction stage. This result is also consistent with the summary and reflection on the process of insight problem solving in previous studies (Bilali et al., 2021; Shen et al., 2018) believes that in the process of insight problem solving, after the failure to solve the problem in a conventional way, the individual will fall into a thinking impasse. At this stage, the individual may no longer intentionally continue to think about the problem, shifting from external attention to internal attention, and being in the unconscious stage of the problem. However, insight often occurs in an instant, so on the eve of problem solving, the individual may be in a very strong stage of unconscious processing and switch to conscious processing quickly (Beeman et al., 2004). According to the above results, it can be seen that microstate D also supports the solution of insight problem through frequent interaction with other Microstates in the whole process, which also echoes the previous research results on executive function and default mode network in the field of creation (including insight problem solving), indicating that in creative thinking, the default mode network and the executive function network are not completely antagonistic; instead, there is a tendency for collaboration between the two networks, with different mechanisms to support creation (Beaty et al., 2015; Yuan & Shen, 2016; Shen et al., 2018; Matheson et al., 2023). Through the comparison between the three conditions, we believe that the unconscious processing represented by the default mode network is the key to the solution of insight problem, while the unconscious processing alone is not enough to achieve insight (refer to the results of unresolved condition in this research), successful problem solving requires executive function to exert top-down regulation on information processing. 5 Conclusion This research describes how the specificity of insight problem solving is reflected in dynamic neural response patterns by comparing insight, non-insight and unresolved conditions. We found that there are at least four important cognitive processes involved in insight problem solving, and making full use of visual image information and auditory semantic information to process, retrieve and extract information is the basic condition for successful problem solving. Unconscious processing mainly appears in the middle and late stages of problem solving and plays a key role in insight problem solving. The executive function continuously supports the repeated selection and integration of information in insight problem solving in the form of regulation. 6 Limitations and prospects First, due to the data-driven characteristics of microstate analysis, the number of types of microstate topographic maps that can be formed are not necessarily, which may be 4-6 in general. It is necessary to preset multiple number of microstates for comparison and select the best fitting number of categories. The best fitting category of microstates in our research is 4, so the subsequent results are all based on the parameters of these four types of topographic maps and the cognitive significance that the topographic maps may represent, which can indicate that insight problem solving in this research may most obviously include these four types of processes. It cannot be investigated whether cognitive processes represented by more topographic maps besides these four microstates also play an important role in insight problem solving. Secondly, the results of microstate A and microstate B in the present research are based on our task (CRA), which contains multiple groups of words. Therefore, it can be imagined that the participants will see the words presented parallelly on the screen when completing the task, and execute the auditory and semantic processing to the words. Although CRA has been widely used to observe creativity, especially insight problem solving, in the existing studies, more than one type of task has been used to explore insight problem solving, such as matchstick problem, nine-point problem and other image tasks, and it is reasonable to speculate that there may be different information processing patterns in such tasks. Therefore, it is necessary to rigorously show that the possibility of microstate A and microstate B in this research is more related to the characteristics of the task. The generalizability of this result needs to be verified in more types of insight problem solving tasks. In the future, researchers may try to verify the stability of the microstate found in this research by using more diverse insight tasks, explore the possibility of other microstates, and explore the role of related cognitive processes on insight problem solving. In addition, although microstates can give an overview of how multiple cognitive processes might operate on a time series, their greatest significance is to provide a possible direction for continuing to purposefully explore the neural response patterns of several cognitive processes in a task. Future research can combine machine learning or multimodal analysis to verify some indicators that have been found to play a key role in insight problem solving, so as to further clarify the cognitive meaning represented by multiple microstates and the key processes involved in insight problem solving. Data Availability Statement In accordance with the ethical guidelines, the data sets generated and processed in this study will be uploaded to a permanent repository (OSF) following publication. These data sets will be made available upon request to the authors. Author Contributions Statement Yan Chen: Conceptualization, Methodology, Investigation, Data Analysis, Writing the original draft. Ying Li: Data curation, Investigation. Guanxiong Liu: Methodology, Investigation. Quanlei Yu: Visualization, Conceptualization. Zheng Liang: Visualization, Investigation, Validation. Shi Chen: Supervision, Validation, Data Analysis. Qingbai Zhao: Conceptualization, Writing & Editing the original draft, Supervision, Funding Acquisition Acknowledgement This work was supported by National Natural Science Foundation of China (Grant No. 32471108), the Humanities and Social Science Planning Fund Project of the Ministry of Education (Grant No.22YJA190013), and the Fundamental Research Funds for the Central Universities (CCNU) (Grant No. XJ2024003701). Reference Aziz‐Zadeh, L., Kaplan, J. T., & Iacoboni, M. (2009). “Aha!”: The neural correlates of verbal insight solutions. Human brain mapping , 30 (3), 908-916. https://doi.org/10.1002/hbm.20554 Baird, B., Smallwood, J., Mrazek, M. D., Kam, J. W., Franklin, M. S., & Schooler, J. W. (2012). Inspired by distraction: Mind wandering facilitates creative incubation. Psychological science , 23 (10), 1117-1122. DOI: 10.1177/0956797612446024 Beaty, R. E., Benedek, M., Barry Kaufman, S., & Silvia, P. J. (2015). Default and executive network coupling supports creative idea production. Scientific reports , 5 (1), 10964. DOI: 10.1038/srep10964 Benedek, M., Christensen, A. P., Fink, A., & Beaty, R. E. (2019). Creativity assessment in neuroscience research. Psychology of Aesthetics, Creativity, and the Arts , 13 (2), 218. DOI: 10.1037/aca0000215 Betzel, R. F., Erickson, M. A., Abell, M., O’Donnell, B. F., Hetrick, W. P., & Sporns, O. (2012). Synchronization dynamics and evidence for a repertoire of network states in resting EEG. Frontiers in computational neuroscience , 6 , 74. DOI: 10.3389/fncom.2012.00074 Bieth, T., Ovando‐Tellez, M., Lopez‐Persem, A., Garcin, B., Hugueville, L., Lehongre, K., … & Volle, E. (2024). Time course of EEG power during creative problem‐solving with insight or remote thinking. Human Brain Mapping , 45 (1), e26547. https://doi.org/10.1002/hbm.26547 Bilalić, M., Graf, M., Vaci, N., & Danek, A. H. (2021). The temporal dynamics of insight problem solving–restructuring might not always be sudden. Thinking & Reasoning , 27 (1), 1-37. https://doi.org/10.1080/13546783.2019.1705912 Bowden, E. M., & Jung-Beeman, M. (2003). Normative data for 144 compound remote associate problems. Behavior research methods, instruments, & computers , 35 , 634-639. DOI: 10.3758/BF03195543 Bréchet, L., Brunet, D., Birot, G., Gruetter, R., Michel, C. M., & Jorge, J. (2019). Capturing the spatiotemporal dynamics of self-generated, task-initiated thoughts with EEG and fMRI. Neuroimage , 194 , 82-92. DOI: 10.1016/j.neuroimage.2019.03.029 Britz, J., Van De Ville, D., & Michel, C. M. (2010). BOLD correlates of EEG topography reveal rapid resting-state network dynamics. Neuroimage , 52 (4), 1162-1170. DOI: 10.1016/j.neuroimage.2010.02.052 Brouwer, H., & Hoeks, J. C. (2013). A time and place for language comprehension: mapping the N400 and the P600 to a minimal cortical network. Frontiers in human neuroscience , 7 , 758. DOI: 10.3389/fnhum.2013.00758 Chen, S., Liang, Z., Li, X., Chen, Y., Zhao, Q., Yu, Q., … & Liu, L. (2021). The role of novel semantic association in the promoting effect of insight on memory. Acta Psychologica Sinica , 53 (8), 837. DOI:10.3724/SP.J.1041.2021.00837 Custo, A., Van De Ville, D., Wells, W. M., Tomescu, M. I., Brunet, D., & Michel, C. M. (2017). Electroencephalographic resting-state networks: source localization of microstates. Brain connectivity , 7 (10), 671-682. DOI: 10.1089/brain.2016.0476 Darsaud, A., Wagner, U., Balteau, E., Desseilles, M., Sterpenich, V., Vandewalle, G., … & Maquet, P. (2011). Neural precursors of delayed insight. Journal of cognitive neuroscience , 23 (8), 1900-1910. DOI:10.1162/jocn.2010.21550 Delorme, A., & Makeig, S. (2004). EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. Journal of neuroscience methods , 134 (1), 9-21. https://doi.org/10.1016/j.jneumeth.2003.10.009 Dohmatob, E., Dumas, G., & Bzdok, D. (2020). Dark control: The default mode network as a reinforcement learning agent. Human brain map** , 41 (12), 3318-3341. https://doi.org/10.1002/hbm.25019 Erickson, B., Truelove-Hill, M., Oh, Y., Anderson, J., Zhang, F. Z., & Kounios, J. (2018). Resting-state brain oscillations predict trait-like cognitive styles. Neuropsychologia , 120 , 1-8. https://doi.org/10.1016/j.neuropsychologia.2018.09.014 Fink, A., Graif, B., & Neubauer, A. C. (2009). Brain correlates underlying creative thinking: EEG alpha activity in professional vs. novice dancers. NeuroImage , 46 (3), 854-862. https://doi.org/10.1016/j.neuroimage.2009.02.036 Haavold, P. Ø., & Sriraman, B. (2022). Creativity in problem solving: integrating two different views of insight. ZDM–Mathematics Education , 54 (1), 83-96. DOI:10.1007/s11858-021-01304-8 Huang, F., Zhao, Q., Zhou, Z., & Luo, J. (2019). People got lost in solving a set of similar problems. NeuroImage , 186 , 192-199. https://doi.org/10.1016/j.neuroimage.2018.10.063 Jung-Beeman, M., Bowden, E. M., Haberman, J., Frymiare, J. L., Arambel-Liu, S., Greenblatt, R., … & Kounios, J. (2004). Neural activity when people solve verbal problems with insight. PLoS biology , 2 (4), e97. https://doi.org/10.1371/journal.pbio.0020097 Khanna, A., Pascual-Leone, A., Michel, C. M., & Farzan, F. (2015). Microstates in resting-state EEG: current status and future directions. Neuroscience & Biobehavioral Reviews , 49 , 105-113. https://doi.org/10.1016/j.neubiorev.2014.12.010 Koenig, T., Lehmann, D., Merlo, M. C., Kochi, K., Hell, D., & Koukkou, M. (1999). A deviant EEG brain microstate in acute, neuroleptic-naive schizophrenics at rest. European archives of psychiatry and clinical neuroscience , 249 , 205-211. DOI:10.1007/s004060050088 Koenig, T., Prichep, L., Lehmann, D., Sosa, P. V., Braeker, E., Kleinlogel, H., … & John, E. R. (2002). Millisecond by millisecond, year by year: normative EEG microstates and developmental stages. Neuroimage , 16 (1), 41-48. DOI:10.1006/nimg.2002.1070 Kohler, W. (1985). The mentality of apes . Routledge. Kounios, J., Frymiare, J. L., Bowden, E. M., Fleck, J. I., Subramaniam, K., Parrish, T. B., & Jung-Beeman, M. (2006). The prepared mind: Neural activity prior to problem presentation predicts subsequent solution by sudden insight. Psychological science , 17 (10), 882-890. DOI:10.1111/j.1467-9280.2006.01798.x Lehmann, D., Ozaki, H., & Pál, I. (1987). EEG alpha map series: brain micro-states by space-oriented adaptive segmentation. Electroencephalography and clinical neurophysiology , 67 (3), 271-288. DOI:10.1016/0013-4694(87)90025-3 Lehmann, D., Pascual-Marqui, R. D., & Michel, C. (2009). EEG microstates. Scholarpedia , 4 (3), 7632. doi:10.4249/scholarpedia.7632 Leszczynski, M., Chaieb, L., Reber, T. P., Derner, M., Axmacher, N., & Fell, J. (2017). Mind wandering simultaneously prolongs reactions and promotes creative incubation. Scientific reports , 7 (1), 10197. DOI:10.1038/s41598-017-10616-3 Lin, J., Chen, Y., **e, J., & Mo, L. (2022). Altered brain connectivity patterns of individual differences in insightful problem solving. Frontiers in Behavioral Neuroscience , 16 , 905806. DOI:10.3389/fnbeh.2022.905806 Lloyd-Cox, J., Chen, Q., & Beaty, R. E. (2022). The time course of creativity: Multivariate classification of default and executive network contributions to creative cognition over time. Cortex , 156 , 90-105. DOI:10.1016/j.cortex.2022.08.008 Lopez, K. L., Monachino, A. D., Morales, S., Leach, S. C., Bowers, M. E., & Gabard-Durnam, L. J. (2022). HAPPILEE: HAPPE In Low Electrode Electroencephalography, a standardized pre-processing software for lower density recordings. NeuroImage , 260 , 119390. DOI: 10.1016/j.neuroimage.2022.119390 Lu, Y., & Singer, W. (2023). Dynamic signatures of the Eureka effect: an EEG study. Cerebral Cortex , 33 (13), 8679-8692. DOI:10.1093/cercor/bhad150 Luo, J. (2004). Neural correlates of insight. Acta Psychologica Sinica , 36 (2), 219-234. Mai, X. Q., Luo, J., Wu, J. H., & Luo, Y. J. (2004). “Aha!” effects in a guessing riddle task: An ERP study. Human brain mapping , 23 (2), 128-128. https://doi.org/10.1002/hbm.20066 Matheson, H. E., Kenett, Y. N., Gerver, C., & Beaty, R. E. (2023). Representing creative thought: A representational similarity analysis of creative idea generation and evaluation. Neuropsychologia , 187 , 108587. https://doi.org/10.1016/j.neuropsychologia.2023.108587 Michel, C. M., & Koenig, T. (2018). EEG microstates as a tool for studying the temporal dynamics of whole-brain neuronal networks: a review. Neuroimage , 180 , 577-593. https://doi.org/10.1016/j.neuroimage.2017.11.062 Michel, C. M., Bréchet, L., Schiller, B., & Koenig, T. (2024). Current state of EEG/ERP microstate research. Brain topography , 37 (2), 169-180. DOI:10.1007/s10548-024-01037-3 Musso, F., Brinkmeyer, J., Mobascher, A., Warbrick, T., & Winterer, G. (2010). Spontaneous brain activity and EEG microstates. A novel EEG/fMRI analysis approach to explore resting-state networks. Neuroimage , 52 (4), 1149-1161. DOI:10.1016/j.neuroimage.2010.01.093 Ohlsson, S. (2011). Deep learning: How the mind overrides experience . Cambridge University Press. Pan, D. N., Hoid, D., Gu, R. L., & Li, X. (2020). Emotional working memory training reduces rumination and alters the EEG microstate in anxious individuals. NeuroImage: Clinical , 28 , 102488. https://doi.org/10.1016/j.nicl.2020.102488 Poulsen, A. T., Pedroni, A., Langer, N., & Hansen, L. K. (2018). Microstate EEGlab toolbox: An introductory guide. BioRxiv , 289850. doi: https://doi.org/10.1101/289850 Ptak, R., Schnider, A., & Fellrath, J. (2017). The dorsal frontoparietal network: a core system for emulated action. Trends in cognitive sciences , 21 (8), 589-599. DOI: 10.1016/j.tics.2017.05.002 Qiu, J., Li, H., Jou, J., Liu, J., Luo, Y., Feng, T., … & Zhang, Q. (2010). Neural correlates of the “Aha” experiences: evidence from an fMRI study of insight problem solving. cortex , 46 (3), 397-403. DOI: 10.1016/j.cortex.2009.06.006 Qiu, J., Li, H., Yang, D., Luo, Y., Li, Y., Wu, Z., & Zhang, Q. (2008). The neural basis of insight problem solving: An event-related potential study. Brain and cognition , 68 (1), 100-106. DOI: 10.1016/j.bandc.2008.03.004 Rokach, L., & Maimon, O. (2010). Data mining and knowledge discovery handbook . Springer New York. DOI: 10.1007/978-0-387-09823-4 Rominger, C., Papousek, I., Perchtold, C. M., Weber, B., Weiss, E. M., & Fink, A. (2018). The creative brain in the figural domain: Distinct patterns of EEG alpha power during idea generation and idea elaboration. Neuropsychologia , 118 , 13-19. https://doi.org/10.1016/j.neuropsychologia.2018.02.013 Saggar, M., Quintin, E. M., Kienitz, E., Bott, N. T., Sun, Z., Hong, W. C., … & Reiss, A. L. (2015). Pictionary-based fMRI paradigm to study the neural correlates of spontaneous improvisation and figural creativity. Scientific reports , 5 (1), 1-11. DOI:10.1038/srep10894 Scheerer, M. (1963). Problem-solving. Scientific American , 208 (4), 118-131. https://www.jstor.org/stable/24936537 Schiller, B., Gianotti, L. R., Baumgartner, T., Nash, K., Koenig, T., & Knoch, D. (2016). Clocking the social mind by identifying mental processes in the IAT with electrical neuroimaging. Proceedings of the National Academy of Sciences , 113 (10), 2786-2791. https://doi.org/10.1073/pnas.1515828113 Shen, W. B., Liu, C., Luo, J., & Yu, J. (2012). Brain perceived intuitively mental impasses in insight problem solving: An ERP study. Acta Psychologica Sinica , 44 (7), 924-935. https://doi.org/10.3724/SP.J.1041.2012.00924 Shen, W., Tong, Y., Li, F., Yuan, Y., Hommel, B., Liu, C., & Luo, J. (2018). Tracking the neurodynamics of insight: A meta-analysis of neuroimaging studies. Biological psychology , 138 , 189-198. https://doi.org/10.1016/j.biopsycho.2018.08.018 Shen, W., Yuan, Y., Liu, C., & Luo, J. (2016). In search of the ‘Aha!’experience: Elucidating the emotionality of insight problem‐solving. British Journal of Psychology , 107 (2), 281-298. https://doi.org/10.1111/bjop.12142 Sio, U. N., & Ormerod, T. C. (2009). Does incubation enhance problem solving? A meta-analytic review. Psychological bulletin , 135 (1), 94. DOI:10.1037/a0014212 Tan, T., Zou, H., Chen, C., & Luo, J. (2015). Mind wandering and the incubation effect in insight problem solving. Creativity Research Journal , 27 (4), 375-382. DOI:10.1177/0033294119896039 Tarailis, P., Koenig, T., Michel, C. M., & Griškova-Bulanova, I. (2024). The functional aspects of resting EEG microstates: a systematic review. Brain topography , 37 (2), 181-217. DOI:10.1007/s10548-023-00958-9 Tarailis, P., Šimkutė, D., Koenig, T., & Griškova-Bulanova, I. (2021). Relationship between spatiotemporal dynamics of the brain at rest and self-reported spontaneous thoughts: an EEG microstate approach. Journal of personalized medicine , 11 (11), 1216. DOI:10.3390/jpm11111216 Wang, T., Zhang, Q., Li, H., Qiu, J., Tu, S., & Yu, C. (2009). The time course of Chinese riddles solving: Evidence from an ERP study. Behavioural Brain Research , 199 (2), 278-282. DOI: 10.1016/j.bbr.2008.12.002 Weisberg, R. W. (2019). Toward an integrated theory of insight in problem solving. In Insight and Creativity in Problem Solving (pp. 5-39). Routledge. DOI:10.1080/13546783.2014.886625 Yeshurun, Y., Nguyen, M., & Hasson, U. (2021). The default mode network: where the idiosyncratic self meets the shared social world. Nature reviews neuroscience , 22 (3), 181-192. DOI:10.1038/s41583-020-00420-w Yu, Y., Oh, Y., Kounios, J., & Beeman, M. (2022). Dynamics of hidden brain states when people solve verbal puzzles. NeuroImage , 255 , 119202. DOI: 10.1016/j.neuroimage.2022.119202 Zappasodi, F., Croce, P., Giordani, A., Assenza, G., Giannantoni, N. M., Profice, P., … & Tecchio, F. (2017). Prognostic value of EEG microstates in acute stroke. Brain topography , 30 , 698-710. DOI:10.1007/s10548-017-0572-0 Zhao, Q., Li, S., Chen, S., Zhou, Z., & Cheng, L. (2015). Dynamic neural processing mode of creative problem solving. Advances in Psychological Science , 23 (3), 375-384. doi: 10.3724/SP.J.1042.2015.00375 Zhao, Q., Wei, L., Li, Y., Zhou, Z., Zhao, L., & Tang, L. (2017). Right hemispheric dominance in forming novel semantic associations. Acta Psychologica Sinica , 49 (11), 1370-1382. doi: 10.3724/SP.J.1041.2017.01370 Zhou, S., Chen, S., Wang, S., Zhao, Q., Zhou, Z., & Lu, C. (2018). Temporal and spatial patterns of neural activity associated with information selection in open-ended creativity. Neuroscience , 371 , 268-276. https://doi.org/10.1016/j.neuroscience.2017.12.006 Information & Authors Information Version history V1 Version 1 09 January 2025 Copyright This work is licensed under a Non Exclusive No Reuse License. Authors Affiliations Yan Chen 0009-0002-6273-1485 Central China Normal University View all articles by this author Ying Li Central China Normal University View all articles by this author Guanxiong Liu Central China Normal University View all articles by this author Quanlei Yu Central China Normal University View all articles by this author Zheng Liang Central China Normal University View all articles by this author Shi Chen Hubei University of Traditional Chinese Medicine View all articles by this author Qingbai Zhao [email protected] Central China Normal University View all articles by this author Metrics & Citations Metrics Article Usage 500 views 169 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Yan Chen, Ying Li, Guanxiong Liu, et al. Insight problem solving may have specific neural response patterns. Authorea . 09 January 2025. DOI: https://doi.org/10.22541/au.173639730.03579860/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.173639730.03579860/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9ff65358884adfa9',t:'MTc3OTM5NTI4NQ=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();