A High-Dimensional Data-Driven Modularity Analysis Reveals Five Distinct Personality Clusters with Different Psychological Profiles

A High-Dimensional Data-Driven Modularity Analysis Reveals Five Distinct Personality Clusters with Different Psychological Profiles | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article A High-Dimensional Data-Driven Modularity Analysis Reveals Five Distinct Personality Clusters with Different Psychological Profiles Yalda Shadmanesh, Amir Hossein Ghaderi This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9379725/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 6 You are reading this latest preprint version Abstract Personality is a fundamental aspect of human behavior, shaping how individuals perceive, interact with, and adapt to their environments. Despite extensive research, establishing a generalizable taxonomy of personality types remains challenging. Conventional personality assessments typically categorize individuals based on aggregated trait scores, which compress item-level variability and overlook how people with identical trait levels may express those traits through different response patterns. However, the reducing rich item-level information into a single aggregated score can obscure meaningful subgroups within the population and restrict the capacity to identify distinct personality profiles. Addressing this limitation is therefore essential for advancing theoretical models of personality structure and improving psychological assessment. Here, we leveraged 60-item NEO Five-Factor Inventory (NEO-FFI) data from the HCP dataset (N = 1206) to investigate personality clustering using an item-pattern-based modularity (IPBM) approach. This framework emphasizes personality categorization based on response styles rather than aggregated trait values, thereby offering a more nuanced characterization of personality structure. Our graph-theoretical clustering analysis identified five distinct personality profiles. While two clusters aligned with commonly reported resilient and under-controlled types, others reflected less frequently described configurations; to the best of our knowledge, one of these clusters has not been documented in prior work. We further incorporated NIMH Toolbox assessments to examine differences across clusters in negative affect, well-being, and self-efficacy. Significant differences emerged for negative affect, well-being, and self-efficacy, demonstrating that item-level personality patterns carry meaningful implications for psychological functioning. Together, these findings highlight the value of response-pattern-based approaches in refining personality taxonomies and underscore how item-level characteristics can yield precise insights into personality and its associations with key indicators of emotional health and adaptive functioning. This framework may enhance personalized assessment strategies and inform the development of more targeted psychological interventions. Physical sciences/Mathematics and computing Biological sciences/Psychology Social science/Psychology Personality types modularity data-driven Graph theoretical analysis emotion Figures Figure 1 Figure 2 Figure 3 Introduction The five-factor model of personality (also known as big five)(McCrae & Costa, 1997) has been an extensively researched model in various domains of personality psychology(Yin et al., 2021). The big five encompasses the personality traits of neuroticism (N), extraversion (E), openness to experience (O), agreeableness (A), and conscientiousness (C)(Costa Jr. & McCrae, 2008). Previous research has demonstrated that these personality traits are associated with individuals' psychological and neural characteristics(Kerber et al., 2021; Liu et al., 2019; Talaei & Ghaderi, 2022), and the configurationof these traits can influence various aspects of an individual's life, including affects(Merz & Roesch, 2011), well-being(Leikas & Salmela-Aro, 2014; Liao et al., 2025), and self-efficacy(Perera et al., 2018). To assess the configuration of personality traits, person-centered approaches (as opposed to variable-based approaches) have been proposed to provide information about individual differences in the patterns of personality traits by grouping individuals with similar values on several personality traits into one personality type(Specht et al., 2014). However, a fundamental question remains: how can meaningful and generalizable personality subgroups be reliably identified given the high dimensionality and complexity of personality assessments(Gerlach et al., 2018)? Traditional clustering methods, such as k-means or unsupervised support vector machine (SVM), are often sensitive to the number of input dimensions and may fail to capture subtle clusters when too many factors (inputs) are involved in the clustering process(Aggarwal & Reddy, 2018; Assent, 2012). Therefore, either the input data must be simplified or alternative or more sophisticated analytic approaches should be used to address this complexity and accurately uncover distinct personality profiles. Instead of developing more sophisticated methods to capture item-level complexity, most previous person-centered approaches have simplified the input data, effectively reducing the dimensionality of personality information. Consequently, individuals with identical mean trait scores can be assumed to belong to the same psychological type, even when their item-level responses differ substantially. Theses person-centered methodologies have evolved over time, from Q-factor analysis (Robins et al., 1996) to more structured approaches such as k-means clustering (Kerber et al., 2021), and more recently to model-based techniques like latent profile analysis (LPA) and latent class analysis (LCA) (Baams et al., 2014; Gerlach et al., 2018; X. Li et al., 2018; Udayar et al., 2020; Bojanić & Čolović, 2025; Liao et al., 2025; Wang, 2025; Wang & Hanges, 2011). Across these studies, three core prototypes have emerged with notable consistency: resilient, overcontrolled, and undercontrolled (Yin et al., 2021). These types are conceptually grounded in Block and Block’s (1980) foundational theory of ego-resiliency and ego-control, which posits that personality structure reflects the dynamic interplay between impulse regulation and adaptive flexibility. In this context, resilient group exhibit high ego-resiliency and demonstrate context-sensitive modulation of ego-control. Overcontrolled group display high ego-control but low ego-resiliency, leading to rigidity and reduced behavioral adaptability. In contrast, undercontrolled group are low on both ego-resiliency and ego-control, often characterized by impulsivity and limited self-regulatory flexibility (Block, J. H., & Block, J., n.d.). Compared to findings regarding the characteristics of the overcontrolled and undercontrolled groups, findings regarding the characteristics of the resilient group have been more consistent (Bojanić & Čolović, 2025; Yin et al., 2021). A recent meta-analysis study of latent profile analysis (LPA) studies quantified the trait composition of these types across large samples. The resilient group consistently showed high levels of E (82.76% of studies), A (96.55% of studies), C (86.21% of studies), and O (79.31% of studies), alongside very low N (over 93% at the lowest level). The undercontrolled group typically exhibited low C (87.5% of studies), low A (75% of studies), and low E (37.5% of studies), combined with elevated N (50% of studies). The overcontrolled group were characterized by high N (61% of studies) and low E (72% of studies), with other traits falling in the low-to-moderate range(Yin et al., 2021). Accordingly, some studies report two-profile models(Semeijn et al., 2020; Wojciechowski, 2021), while others identify three(Robins et al., 1996; Van Den Akker et al., 2013), four (Chi & Chi, 2023; Kerber et al., 2021; Rzeszutek & Gruszczyńska, 2020; Udayar et al., 2020), or five profiles (Conte et al., 2017; Kinnunen et al., 2012; X. Li et al., 2018). Notably, Gerlach et al. (2018), using a data-driven Gaussian mixture model over the large sample datasets, identified four replicable personality types, further challenging the rigidity of the three-profile model(Gerlach et al., 2018). As noted above, these findings were obtained using methods that relied on aggregated trait-level scores. It remains possible that preserving the item-level complexity of personality responses could lead to different results, potentially revealing more nuanced or distinct patterns that are obscured when data are simplified. Rather than reducing input dimensionality, which can inadvertently remove meaningful variability, it may be more advantageous to employ analytical frameworks capable of accommodating and interpreting high-dimensional structure directly. Approaches informed by modern graph-theoretic principles, among others, are well suited for capturing such complexity without collapsing the data. In particular, graph theoretical community-detection methods based on the modularity framework (Newman, 2006b, 2006a) allow one to identify naturally emerging modules (subnetworks) of highly interrelated items. Applying such modularity analyses to item-level personality response networks could reveal latent subtypes or clusters of traits that are not apparent when using aggregated scores. Here, we introduce a novel graph-theoretic, item-pattern-based modularity (IPBM) analysis to evaluate and cluster personality profiles at the item level. The first objective of this study was to characterize the structure of personality types using this new data-driven framework and to compare the resulting typology with those derived from traditional aggregation-based approaches. In contrast with previous person-centered methodologies, the presented method identifies personality types as emergent communities of individuals who share structurally similar response profiles. Specifically, the IPBM method constructs a network in which nodes represent individuals and edges reflect the similarity (Spearman’s rank correlation coefficient) of their item-level responses. Communities within this network are then identified using an optimization approach, which seeks to detect the most robust modules of individuals sharing similar item-level response patterns (as illustrated in Figure 1-a). Given that this approach retains the full complexity of the input data, we hypothesized that it may reveal distinct and potentially understudied personality groupings that remain undetected when relying on aggregated trait scores. The second objective of this study was to determine whether the personality groups identified through the IPBM framework show meaningful differences in their psychological and neural characteristics. We specifically focused on well-being, self-efficacy, and negative affect. In this context, previous research have mostly focused on the links between personality and well-being (Busseri & Erb, 2024; HAN, 2020; Henning et al., 2017; Mammadov, 2023; Mammadov & Ward, 2022; Marengo et al., 2021; Rzeszutek & Gruszczyńska, 2020; Udayar et al., 2020). These studies have shown that the resilient profile scores highest on measures of subjective well-being (Mammadov et al., 2024; Mammadov & Ward, 2023; Mammadov, 2023; Udayar et al., 2020; Rzeszutek & Gruszczyńska, 2020). Regarding self-efficacy, Perera et al. showed a cluster with low neuroticism and above-average levels of the other traits, aligns with the resilient profile, which demonstrated the highest levels of self-efficacy. In contrast, the excitable profile, characterized by slightly above-average neuroticism, higher levels of extraversion, openness, and agreeableness, and below-average conscientiousness, aligns with the undercontrolled profile, which exhibited the lowest levels of self-efficacy (Perera et al., 2018). To the best of our knowledge, no previous study has investigated the relationship between person-centered personality profiles and negative affect. Regarding the second objective of the study, which was to examine psychological and neural differences between the groups identified using the IPBM method, our hypothesis was that the different groups would exhibit distinct psychological profiles, resulting in differences in well-being, self-efficacy, negative affect, and related measures. This is particularly interesting and informative for the less commonly observed groups that may emerge from the IPBM analysis, and it could also replicate some previous findings suggesting that groups with profiles resembling resilient or undercontrolled types tend to show significant differences from other groups. If such results are obtained, they would provide validation for the clustering approach used in this study and highlight new findings that align with prior research. Results In this study, we used a graph theoretical modularity approach to cluster individuals based on their responses to the 60 single items of the NEO-FFI. Spearman’s rank correlation analysis was performed across these items, generating an adjacency matrix that captured the relationships between participants based on their personality responses. This matrix was then used for a graph theoretical modularity analysis to identify personality-based modules (groups). To explore the differences among personality clusters, big five dimensions, and NIMH toolboxes, statistical analyses were applied (figure 1). Participant Clustering via Modularity Analysis As shown in figure 1-b, modularity analysis revealed that the optimal modularity was achieved at a modularity coefficient of γ = 1.042, with a mean modularity ratio of 1.51. Higher values of γ led to excessive fragmentation of the network, resulting in an increased number of isolated nodes beyond our acceptable threshold, whereas lower γ values yielded modularity values below the optimal range. At this optimal modularity coefficient, participants were clustered into eight distinct modules based on inter-individual similarity in 60-items NEO-FFI responses. Among these, three modules were identified as an isolated cluster (with n <= 5) and excluded from subsequent analyses. Five modules, clusters 1 ( n = 398), 2 ( n = 326), 3 ( n = 312), 4 ( n = 17), and 5 ( n = 134), were selected for downstream analysis. Table 1 summarizes the mean intra-cluster similarity and dimension-specific similarity for each of the five clusters. Additionally, the scatter plot of big five factors for different individuals in different groups (identified by color codes) is presented in figure 2-a. The split-half reliability analysis confirmed the stability of our findings, as the modularity analysis consistently yielded the same five distinct personality clusters in both independent halves of the dataset. Personality Traits Differences Between Clusters Accordingly, we evaluate the mean and standard deviation (SD) of big five factors for each group (figure 2.b) and results showed: Cluster 1: cluster 1 showed the lowest mean level of C (mean = 30.74) and the highest mean level of O (mean = 34.15) among all clusters. Variability in N was markedly high (SD = 7.26), indicating substantial heterogeneity within the cluster. Similarly, E scores displayed elevated dispersion (SD = 6.13). Cluster 2: As shown in figure 2-a, cluster 2 was characterized by the highest level of A (mean = 34.14) and the lowest level of N (mean = 9.90) across all clusters. Additionally, this cluster demonstrated relatively high scores in E (mean = 33.22) and C (mean = 36.25). Cluster 3: As shown in figure 2-a, cluster 3 showed the lowest mean score in O (mean = 22.56). It ranked second in both E (mean = 30.79± = 5.30) and C (mean = 37.01± 4.84). Cluster 4 : Cluster 4 exhibited the highest mean scores in C (mean = 39.76± 5.45) and E (mean = 33.29± 4.57) among all clusters. Cluster 5 : Cluster 5 exhibited the lowest mean scores across all clusters in A (mean = 30.11± 5.64) and E (mean = 26.90 ± 5.67). Notably, it showed the highest level of N (mean = 23.55± 5.80). Then, to evaluate differences in big five personality traits across the identified clusters, we performed a series of permutation t-tests (corrected with Bonferroni correction for multiple comparisons). The analysis revealed several significant differences, as shown in Figure 2-c. Most significant differences between groups were observed in O (9 significant differences between pairs) while the analyses showed only 3 significant differences between group pairs in A. Furthermore, cluster 5 showed the most significant differences with other groups in all five traits (18 pairs of significant differences). Comprehensive statistical results for all pairwise comparisons are available in Supplementary Information (Table S1 and S2). Cluster-Specific NIMH Profiles Negative affect. As shown in figure 3-a, cluster 2 showed the lowest values of negative affect among clusters (mean = 46.77±4.69) and exhibited significant differences with all other clusters (t(1,2) = 14.20, p < 0.001; t(2,3) = -4.67, p < 0.001; t(2,4) = -3.82, p = 0.002; t(2,5) = -13.47, p < 0.001). In contrast, clusters 1 (mean = 52.20 ±5.44) and 5 (mean = 53.92 ±6.17) showed the highest mean values of negative affect and exhibited significant differences with cluster 3 (t (1,3) = 9.01, p < 0.001; t (3,5) = -9.45, p < 0.001). The significant highest values of cluster 5 in negative affect dimensions, is also replicated in many pair comparisons (12 pairs of significant differences). Conversely, cluster 2 showed the significant lowest values in many pair comparisons (12 comparisons) in negative affect dimensions. Among the dimensions most significant differences were observed in anger hostility (6 significant differences), while only one significant difference was observed in anger physical aggression. Well-being. As shown in Figure 3-b, cluster 2 showed the highest values of well-being among clusters (mean=55.07 ±6.11) and exhibited significant differences with most of the other clusters (t (2,1) = -10.26, p < 0.001; t (2,3) = 4.03, p = 0.004; t (2,5) = 11.07, p < 0.001). In contrast, clusters 1 (mean=50.01± 6.99) and 5 (mean=47.86± 6.83) showed the lowest mean values of well-being and exhibited significant differences with cluster 3 (t(1,3) = -5.68, p < 0.001; t(3,5) = 7.20, p < 0.001). The significantly highest values of cluster 2 is also replicated in many pair comparisons (7 pairs of significant differences) in dimensions of well-being. Conversely, cluster 5 showed the significantly lowest values in many pair comparisons (7 pairs of significant differences) in well-being dimensions. Among the dimensions, the most significant differences were observed in meaning and purpose (6 significant differences), while only three significant differences were observed in positive affect. Self-efficiency. As shown in Figure 3-c, cluster 2 showed the highest values of self-efficiency among clusters (mean=55.78± 6.83) and exhibited significant differences with all other clusters (t(2,1) = -12.81, p < 0.001; t(2,3) = 8.22, p < 0.001; t(2,4) = 3.26, p = 0.038; t(2,5) = 13.60, p < 0.001). In contrast, clusters 1(mean=52.20 ±5.44) and 5 (mean=53.92 ±6.17) showed the lowest mean values of self-efficiency and exhibited significant differences with cluster 3 and with each other (t (1,3) = -4.40, p < 0.001; t(1,5) = 3.99, p = 0.002; t(3,5) = 7.39, p < 0.001). In dimensions of self-efficiency, the significantly highest values of cluster 2 are also replicated in many pair comparisons (7 pairs of significant differences). Conversely, cluster 5 showed the significantly lowest values in many pair comparisons (4 significant comparisons) in the dimensions. Among the dimensions, the most significant differences were observed in perceived stress (6 significant differences), while only four significant difference was observed in self-efficiency. Comprehensive statistical results for all pairwise comparisons are available in supplementary tables S3 to S6. Discussion Our findings demonstrate that the proposed IPBM approach can effectively derive a limited and interpretable set of personality groups from item-level response patterns. The resulting five clusters not only reproduced well-established personality configurations commonly described in the literature, but also revealed additional, less frequently discussed profiles, including one cluster that, to the best of our knowledge, has not been characterized previously. These distinctions were meaningfully reflected in psychological outcomes: clusters differed significantly in well-being, negative affect, and self-efficacy, indicating that item-level response structure captures consequential variation in personality expression. The Importance of the IPBM Method in Regulating Input Complexity As established in prior literature, high-dimensional spaces fundamentally erode the utility of pairwise distance metrics, the very foundation of classical clustering algorithms (Aggarwal & Reddy, 2018; Assent, 2012). A central issue is the curse of dimensionality , whereby increasing the number of features renders data points increasingly sparse and diminishes the meaningfulness of distance metrics that these algorithms rely upon. In high dimensions, objects with different underlying cluster structure may appear deceptively similar due to the concentration of distances, which complicates both cluster separation and cluster validation in unsupervised learning. Empirically, this effect has been shown to undermine the performance and interpretability of classical clustering methods as dimensionality increases, particularly when naïvely using raw feature vectors without dimensionality reduction or feature selection (Assent, 2012). Consequently, traditional clustering techniques, such as k-means or model-based profile estimation, often falter with high-dimensional inputs which may suffers from increased sensitivity to irrelevant features and poor separation between clusters, which degrades both cluster quality and reproducibility (Kadir et al., 2014). Furthermore, naïve use of distance-based clustering can result in artificially inflated numbers of clusters that primarily reflect noise and high feature count rather than meaningful subgroup distinctions (Assent, 2012). When classical clustering methods are applied to identify personality groups based on responses to the 60-item NEO-FFI personality questionnaire, these high-dimensionality–related issues inevitably arise. As a result, the derived clustering solutions can become highly unstable and strongly influenced by noise, leading to poor reproducibility and limited psychological interpretability. This analytical challenge has historically reinforced the field's reliance on aggregated trait scores, a pragmatic dimensional reduction that regrettably obscures meaningful, within-trait response-pattern variability. By adopting a graph-theoretic modularity analysis, our approach directly circumvents this limitation. The modularity optimization framework efficiently uncovers latent community structures by maximizing within-group similarity against a randomized null model, thereby enabling the discovery of a stable, parsimonious, and optimally constrained five-cluster solution. This outcome demonstrates how graph-based techniques can preserve the granular richness of item-level data while inherently resisting the over-fragmentation and instability that plague classical clustering under high-dimensional conditions. Characteristics of Our Five-Cluster Personality Structure Our analysis, with the exclusion of a very small number of participants (around 0.2% of the sample), was able to optimally cluster individuals into five groups. As described earlier, these groups were formed based on similarities and differences in individuals’ responses to each individual questionnaire item; however, when examined at the level of the five aggregated personality factors, the groups also exhibited numerous statistically significant differences. We further conducted within-group correlation analyses to identify which of the big five factors likely played a more prominent role in the formation of each group. Overall, these results indicated that two of the identified groups correspond well to previously established personality profiles, namely the resilient and under-controlled types. In contrast, the remaining three groups exhibited relatively distinct and more novel characteristics compared to those reported in many prior studies. In particular, our second cluster shows a strong correspondence with the resilient personality type reported in prior studies. This profile is typically characterized by relatively high levels of extraversion, agreeableness, conscientiousness, and openness to experience , together with low levels of neuroticism according to the big five framework (Gerlach et al., 2018; Kerber et al., 2021; Liao et al., 2025; Min & Su, 2020; Udayar et al., 2020; Wojciechowski, 2021; Yin et al., 2021; Yu & Zhang, 2021). In our analysis, this cluster demonstrated the highest degree of internal similarity across all items, indicating a highly consistent and well-defined response pattern among its members. Furthermore, their responses to openness-related items were highly correlated, suggesting that members of this cluster are highly similar in how they express openness to experience. In agreement with previous findings (Busseri & Erb, 2024; Perera et al., 2018), this cluster showed high scores across well-being and self-efficacy indices and low scores across negative affect by NIMH toolbox. Another cluster identified in our analysis was cluster 5, which can be closely aligned with the undercontrolled (Baams et al., 2014; Kerber et al., 2021; Yin et al., 2021; Yu & Zhang, 2021) or vulnerable in (De Clercq et al., 2012) or distressed in (Morgan et al., 2017) types. Previous studies suggest that individuals in this group typically exhibit high levels of neuroticism, coupled with low levels of conscientiousness and agreeableness which are closely matched with the aggregated scores of big five factor in cluster 5. In our analysis, the cluster 5 showed the highest degree of internal similarity across neuroticism which suggest individuals in this cluster share core elements of emotional distress and interpersonal challenges. Consistently, higher negative affect and lower self-efficacy and well-being scores in this cluster, compared with the other clusters, reflect difficulties in emotional regulation, impulse control, and behavioral inhibition within this group (Perera et al., 2018; Robins et al., 1996; Asendorpf et al., 2001). Among the three remaining clusters identified in our analysis, two can still be meaningfully compared with profiles reported in previous studies. In particular, cluster 3 appears to share certain characteristics with the resilient type; however, it also shows notable differences relative to cluster 2, which represented the prototypical resilient profile in our analysis. In fact, cluster 3 highlights the discriminative power of our IPBM approach by revealing a distinct profile that would typically be merged with the conventional resilient type in aggregate-based analyses. This cluster emerged from a response pattern characterized by high conscientiousness and agreeableness, and relatively low neuroticism and openness. While this aggregate profile bears resemblance to the resilient cluster identified by Van den Akker et al(Van Den Akker et al., 2013), our item-level analysis distinguished it from the classic resilient profile (cluster 2) in our own data. To our knowledge, no prior study has reported the simultaneous presence of these two distinct resilient-like profiles, and this novel finding suggests the existence of more than a single resilient prototype. This profile showed the high intra-consistency in neuroticism and the low intra-consistency in openness, in contrast to cluster 2, which exhibited low similarity in neuroticism alongside uniformly in openness. The difference between these two resilience-like clusters confirmed by NIMH questionnaire data, which showed significant differences in negative affect, well-being and life satisfaction between the two clusters. On these measures, cluster 2 demonstrated more favorable life outcomes, with significantly lower negative affect and higher well-being and life satisfaction. This suggests that life outcome measures may differ substantially between the two resilient groups, depending on whether they are primarily organized around neuroticism or, conversely, around openness. Another group that has been only partially addressed in previous studies corresponds to cluster 1 in our analysis. This cluster is characterized by a high level of openness to experience and markedly low conscientiousness. In this context, a comparable cluster, referred to as free-spirit, was previously identified by Henning et al. in a Swedish elderly cohort(Henning et al., 2017). In terms of well-being and self-efficiency, this cluster ranks second-lowest, and for negative affect ranks second highest, we attribute this pattern to the elevated neuroticism levels within the group, though this hypothesis warrants further investigation. The final cluster discussed here is cluster 4, which comprises a very small proportion of our sample yet consistently emerges as a distinct cluster. Despite its limited size, this group exhibits a set of unique and characteristic features that clearly differentiate it from the other clusters. This cluster is characterized by the high scores in conscientiousness, agreeableness, extraversion, openness and moderate scores in neuroticism. The intra-cluster analysis of this cluster revealed homogeneous response patterns for both openness and agreeableness, suggesting a high degree of similarity among individuals in this cluster with respect to these two traits. An important observation is that this group reported the highest levels of meaning and purpose, which co-occurred with elevated scores on negative affect subscales, particularly anger–hostility and anger–physical aggression. This pattern suggests that, for these individuals, a strong sense of purpose may not fully buffer against feelings of irritability or frustration. Consistent with this interpretation, this group also exhibited persistently low levels of self-efficacy. Conclusion This study shows that an item-level, graph-based personality modeling approach can reliably extract a small and interpretable set of personality prototypes while preserving meaningful individual differences. By regulating input complexity through modularity optimization, the IPBM framework overcomes key limitations of classical clustering in high-dimensional spaces and recovers both well-established and less-studied personality profiles. The identified clusters differed robustly in well-being, negative affect, and self-efficacy, indicating that item-level response structure captures psychologically consequential variation beyond aggregated trait scores. Overall, these findings highlight the value of network-based, data-driven approaches for refining personality typologies and linking them to behavioral outcomes. Limitations Despite its contributions, this study has several limitations that should be addressed in future research. First, the data were drawn exclusively from the Human Connectome Project, which primarily represents a sample from the United States. Consequently, the generalizability of the identified personality profiles across diverse cultural and demographic contexts remains uncertain. Replicating this study in different samples from various regions is essential to validate the universality or cultural specificity of these clusters. Materials and methods Participants This study used the behavioral data of the HCP Young Adult 1200 Subjects release 10. The initial dataset comprised 1206 participants but data for 8 individuals was excluded, as their questionnaire responses were not recorded. Consequently, the final analytical sample consisted of 1198 participants (650 females), with a median age of 28.7 years (age range: 22-37, SD: 3.67). Inclusion criteria required no history of major psychiatric, neurological, or cardiovascular disorders, the capacity to provide informed consent, and a Mini-Mental State Examination score of 29–30. Individuals who were asymptomatic but reported a history of smoking, overweight status, or recreational substance use were not excluded, in line with the project’s aim to capture normative variation. Exclusion criteria included diagnosed neuropsychiatric or neurological conditions, genetic disorders, use of psychoactive or hormonal medications within the past year, moderate to severe traumatic brain injury, preterm birth (<37 weeks gestation), current pregnancy, and contraindications for MRI scanning (e.g., implanted metal devices or severe claustrophobia) (Van Essen et al., 2012, 2013). This study used publicly available, de-identified data from the Human Connectome Project (HCP). All participants provided written informed consent, and all recruitment and data acquisition procedures were approved by the Washington University Institutional Review Board, in accordance with the Declaration of Helsinki and all relevant guidelines and regulations (Christova et al., 2020; Van Essen et al., 2012, 2013). The present study involved secondary analysis of anonymized data and did not require additional ethical approval. Behavioral data Personality data The 60-item version of the Costa and McCrae NEO-FFI (McCrae & Costa, 2004), which has shown excellent reliability and validity, was administered to HCP subjects(Christova et al., 2020; Van Essen et al., 2013). This measure was collected as part of the Penn Computerized Cognitive Battery [version (NEO-FFI-2, 2004)] (Costa Jr. & McCrae, 2008; Gur, 2001) The NEO-FFI is a self-report questionnaire with 60 items (abbreviated version of the 240-item inventory). For each item, participants reported their level of agreement on a 5-point Likert scale, from strongly disagree to strongly agree (strongly disagree=1; disagree=2; neither agree nor disagree =3; agree=4; strongly agree=5) (Christova et al., 2020; Van Essen et al., 2013). Openness to experience, conscientiousness, extraversion, agreeableness, and neuroticism scores are derived by coding each item’s answer. Both the total score on each personality factor and subjects' responses to each of the 60 individual personality items were used. The total score was used for comparisons between groups (defined in the next section) and the 60 individual personality items scores were used as the series for correlation analysis and generating the similarity matrix. Prior to any further analysis, all reverse-keyed items were reverse-scored to ensure that the numerical values consistently aligned with the direction of the intended personality traits. NIMH Toolboxes In accordance with the standardized HCP protocol(Van Essen et al., 2012), three assessment batteries from the NIH toolbox, well-being, negative affect, and stress and self-efficacy, were administered. All scores were provided as standardized T-scores (M = 50, SD = 10), normed against a nationally representative sample. For each battery, T-scores for the constituent surveys and an overall composite score (computed as the average of those surveys) were used(Christova et al., 2020). Psychological well-being was assessed using measures of positive affect, life satisfaction, and meaning and purpose (Salsman et al., 2014). Negative affect was evaluated using six constituent surveys: sadness, fear-affect, fear–somatic arousal, anger-affect, anger–hostility, and anger–physical aggression. The first four referenced emotional experiences over the past 7 days, while the latter two assessed trait-like dispositions(Pilkonis et al., 2013). The stress and self-efficacy battery included perceived stress and self-efficacy(Kupst et al., 2015; G. Li et al., 2021). For further details regarding each of the three NIH toolbox batteries, please refer to the supplementary materials. Correlation-based personality analysis and similarity matrix We used the 60 individual personality items scores for each participant as a series for Spearman’s correlation analysis and generating a similarity matrix. As we mentioned before, all reverse-keyed items were reverse-scored. The value along each point in the series, reflects the participant’s Likert-scale response (figure1.a). Subsequently, Spearman’s rank correlation coefficient was calculated between different individual series. Then the values of correlation coefficients were arranged in a shape of a similarity matrix. This outcome 1198 by 1198 matrix was a symmetric square matrix represent the personality similarity among all individuals (figure 1.a), in which each element indicates the linear correlation (Spearman’s rank correlation coefficient) between the response patterns of a pair of participants across the entire set of items. Modularity analysis and subnetworks To reveal the modular architecture embedded within the similarity matrix, we implemented a combined approach integrating both unsupervised and supervised modularity analyses (Ghaderi et al., 2025; Newman, 2006b, 2006a). Initially, we employed the modularity_und function from the Brain Connectivity Toolbox to detect intrinsic modules based solely on the internal edge structure of the network (Rubinov & Sporns, 2010). This unsupervised algorithm includes a tunable resolution parameter, , which regulates the granularity of partitioning: lower values tend to yield larger modules, while higher values promote finer subdivisions. Importantly, this step relies purely on topological connectivity without recourse to external validation metrics. To identify the optimal , we introduced a supervised evaluation framework based on a modularity ratio (MR) (Ghaderi et al., 2025; Musa et al., 2025). Starting from a that produced at least two modules, we computed the average within-module connectivity and compared it against a null distribution derived from randomized networks with preserved degree distributions. The MR was defined as the ratio of empirical to randomized within-module connectivity. This procedure was iterated across all detected modules, and the mean MR was used as a quality index for each value. By incrementally increasing and repeating this process, we obtained a curve of mean MR values, from which the maximizing MR was selected, provided the number of isolated nodes remained below four. This optimal was then fixed for all subsequent modular analyses (fig1.b). All analyses were conducted in MATLAB 2024a, with custom scripts publicly accessible at https://github.com/AHGhaderi/Amir-Hossein-Ghaderi/commit/df636b2105e578a8969ffa88856e54f3267a40a7. Following cluster identification, we computed two types of correlation matrices for each cluster. First, an intra-cluster similarity matrix was derived from participants' responses to all 60 items. Second, domain-specific inter-individual similarity matrices were calculated based on participants' responses to the 12 individual items corresponding to each Big Five domain. The mean correlation coefficients for all matrices are reported. To evaluate the magnitude of these correlation coefficients, we followed the guidelines proposed by Gignac & Szodorai (Gignac & Szodorai, 2016), where correlations of 0.10, 0.20, and 0.30 are considered relatively small, typical, and relatively large, respectively, in the context of individual differences research. Permutation t-tests We compared personality trait scores across the five clusters identified via modularity-based analysis (Methods, section 2.4). Pairwise differences on the five NEO personality dimensions were assessed using nonparametric permutation t-tests for independent samples (5,000 iterations), with Bonferroni correction for multiple comparisons (Bonferroni, 1936). This approach preserves the between-subject structure without assuming normality. The same framework was used to assess NIMH toolboxes (integrated scales and individual domains) across the five clusters. All analyses were conducted in MATLAB using custom scripts, and the permutation test was implemented following standard guidelines((Nichols & Holmes, 2002); available at: https://github.com/AHGhaderi/Amir-Hossein-Ghaderi/commit/4f2c8f731d82707cc06dc20e04fd3ca01b9e8e02). Robustness and Reliability Analysis To evaluate the robustness and stability of the clustering solution, we performed a split-half reliability analysis. The total dataset (N=1198) was randomly partitioned into two independent halves. The identical modularity optimization pipeline used for the full dataset was then applied to each subset separately to determine whether the structure of clusters could be independently replicated. Declarations Funding The authors received no financial support for the research, authorship, and/or publication of this article. Author Contribution Y.S.: Conceptualization, Methodology, Software, Investigation, Writing - Original Draft. A.H.G.: Supervision, Conceptualization, Formal analysis, Validation, Writing - Review & Editing. Data Availability The data used in this study are available through the Human Connectome Project (HCP) at https://www.humanconnectome.org/. References Aggarwal, C. C. & Reddy, C. K. (eds) Data Clustering: Algorithms and Applications 1st edn (Chapman and Hall/CRC, 2018). https://doi.org/10.1201/9781315373515 Assent, I. Clustering high dimensional data. WIREs Data Min. Knowl. Discov. 2 (4), 340–350. https://doi.org/10.1002/widm.1062 (2012). Baams, L., Overbeek, G., Dubas, J. S. & Van Aken, M. A. G. On Early Starters and Late Bloomers: The Development of Sexual Behavior in Adolescence Across Personality Types. J. Sex. Res. 51 (7), 754–764. https://doi.org/10.1080/00224499.2013.802758 (2014). Block, J. H. & Block, J. (n.d.). The role of ego-control and ego resiliency in the organization of behavior – ScienceOpen . Retrieved July 16, from (2025). https://www.scienceopen.com/book?vid=0a410035-d7e6-4d0c-abd6-215f15fe58ee Bojanić, M. & Čolović, P. Personality profiles in SAPA data: An exploratory study. Pers. Indiv. Differ. 234 , 112957. https://doi.org/10.1016/j.paid.2024.112957 (2025). Bonferroni, C. E. Teoria statistica delle classi e calcolo delle probabilità (Seeber, 1936). Busseri, M. A. & Erb, E. M. The happy personality revisited: Re-examining associations between Big Five personality traits and subjective well‐being using meta‐analytic structural equation modeling. J. Pers. 92 (4), 968–984. https://doi.org/10.1111/jopy.12862 (2024). Chi, J. & Chi, Y. N. Cluster Analysis of Personality Types Using Respondents’ Big Five Personality Traits. Int. J. Data Sci. 4 (2), 116–135. https://doi.org/10.18517/ijods.4.2.116-135.2023 (2023). Christova, P., Joseph, J. & Georgopoulos, A. P. Behavioral–genetic associations in the Human Connectome Project. Exp. Brain Res. 238 (11), 2445–2456. https://doi.org/10.1007/s00221-020-05893-w (2020). Conte, J. M., Heffner, T. S., Roesch, S. C. & Aasen, B. A person-centric investigation of personality types, job performance, and attrition. Pers. Indiv. Differ. 104 , 554–559. https://doi.org/10.1016/j.paid.2016.09.004 (2017). Costa, P. T. Jr. & McCrae, R. R. The Revised NEO Personality Inventory (NEO-PI-R). In The SAGE handbook of personality theory and assessment, Vol 2: Personality measurement and testing (179–198). Sage Publications, Inc. https://doi.org/10.4135/9781849200479.n9 (2008). De Clercq, B., Rettew, D., Althoff, R. R. & De Bolle, M. Childhood personality types: Vulnerability and adaptation over time. J. Child Psychol. Psychiatry . 53 (6), 716–722. https://doi.org/10.1111/j.1469-7610.2011.02512.x (2012). Gerlach, M., Farb, B., Revelle, W. & Nunes Amaral, L. A. A robust data-driven approach identifies four personality types across four large data sets. Nat. Hum. Behav. 2 (10), 735–742. https://doi.org/10.1038/s41562-018-0419-z (2018). Ghaderi, A. H., Yang, X. F., Gotlieb, R., Johnson, D. L. & Immordino-Yang, M. H. Intersecting effects of social circumstances and transcendent thinking on mid-adolescents’ longitudinal functional connectome development (p. 2025.05.20.654735). bioRxiv. (2025). https://doi.org/10.1101/2025.05.20.654735 Gignac, G. E. & Szodorai, E. T. Effect size guidelines for individual differences researchers. Pers. Indiv. Differ. 102 , 74–78. https://doi.org/10.1016/j.paid.2016.06.069 (2016). Gur, R. Computerized Neurocognitive Scanning: I. Methodology and Validation in Healthy People. Neuropsychopharmacology 25 (5), 766–776. https://doi.org/10.1016/S0893-133X(01)00278-0 (2001). HAN, J. H. The Effects of Personality Traits on Subjective Well-being and Behavioral Intention Associated with Serious Leisure Experiences. J. Asian Finance Econ. Bus. 7 (5), 167–176. https://doi.org/10.13106/JAFEB.2020.VOL7.NO5.167 (2020). Henning, G., Hansson, I., Berg, A. I., Lindwall, M. & Johansson, B. The role of personality for subjective well-being in the retirement transition – Comparing variable- and person-oriented models. Pers. Indiv. Differ. 116 , 385–392. https://doi.org/10.1016/j.paid.2017.05.017 (2017). Kadir, S. N., Goodman, D. F. M. & Harris, K. D. High-Dimensional Cluster Analysis with the Masked EM Algorithm. Neural Comput. 26 (11), 2379–2394. https://doi.org/10.1162/NECO_a_00661 (2014). Kerber, A., Roth, M. & Herzberg, P. Y. Personality types revisited–a literature-informed and data-driven approach to an integration of prototypical and dimensional constructs of personality description. PLOS ONE . 16 (1), e0244849. https://doi.org/10.1371/journal.pone.0244849 (2021). Kinnunen, M. et al. Personality profiles and health: Longitudinal evidence among Finnish adults. Scand. J. Psychol. 53 (6), 512–522. https://doi.org/10.1111/j.1467-9450.2012.00969.x (2012). Kupst, M. J. et al. Assessment of stress and self-efficacy for the NIH Toolbox for Neurological and Behavioral Function. Anxiety Stress Coping . 28 (5), 531–544. https://doi.org/10.1080/10615806.2014.994204 (2015). Leikas, S. & Salmela-Aro, K. Personality types during transition to young adulthood: How are they related to life situation and well-being? J. Adolesc. 37 (5), 753–762. https://doi.org/10.1016/j.adolescence.2014.01.003 (2014). Li, G. et al. Perceived stress, self-efficacy, and the cerebral morphometric markers in binge-drinking young adults. NeuroImage: Clin. 32 , 102866. https://doi.org/10.1016/j.nicl.2021.102866 (2021). Li, X., Yao, P., Didericksen, K., Jang, J. & Olson, D. Five Personality Types Based on Big Five: A Latent Class Analysis. Family Sci. Rev. 22 (2). https://doi.org/10.26536/fsr.2018.22.02.05 (2018). Liao, Y. et al. Could social support mediate the relationship between personality trait patterns and mental health in Chinese older adults: A nationwide cross-sectional study conducted in 2022. BMC Public. Health . 25 (1). https://doi.org/10.1186/s12889-025-22449-1 (2025). Liu, W., Kohn, N. & Fernández, G. Intersubject similarity of personality is associated with intersubject similarity of brain connectivity patterns. NeuroImage 186 , 56–69. https://doi.org/10.1016/j.neuroimage.2018.10.062 (2019). Mammadov, S. Personality profiles of gifted adolescents and relations with life satisfaction, perceived social support, and academic achievement. High. Ability Stud. 34 (1), 87–108. https://doi.org/10.1080/13598139.2022.2068509 (2023). Mammadov, S., Wang, S. & Lu, Z. Personality types and their associations with psychological resilience, coping with stress, and life satisfaction among undergraduate students: A latent profile analysis approach. Pers. Indiv. Differ. 222 , 112599. https://doi.org/10.1016/j.paid.2024.112599 (2024). Mammadov, S. & Ward, T. J. Exploring the Relations Between Personality, Implicit Theories, and Subjective Well-Being Among High-Ability Undergraduate Students. Gifted Child Quarterly. (Sage CA: Los Angeles, CA). https://doi.org/10.1177/00169862221107862 (2022). Mammadov, S. & Ward, T. J. Exploring the Relations Between Personality, Implicit Theories, and Subjective Well-Being Among High-Ability Undergraduate Students. Gifted Child. Q. 67 (1), 28–43. https://doi.org/10.1177/00169862221107862 (2023). Marengo, D., Davis, K. L., Gradwohl, G. Ö. & Montag, C. A meta-analysis on individual differences in primary emotional systems and Big Five personality traits. Sci. Rep. 11 (1), 7453. https://doi.org/10.1038/s41598-021-84366-8 (2021). McCrae, R. R. & Costa, P. T. Personality trait structure as a human universal. Am. Psychol. 52 (5), 509–516. https://doi.org/10.1037/0003-066X.52.5.509 (1997). McCrae, R. R. & Costa, P. T. A contemplated revision of the NEO Five-Factor Inventory. Pers. Indiv. Differ. 36 (3), 587–596. https://doi.org/10.1016/S0191-8869(03)00118-1 (2004). Merz, E. L. & Roesch, S. C. A latent profile analysis of the Five Factor Model of personality: Modeling trait interactions. Pers. Indiv. Differ. 51 (8), 915–919. https://doi.org/10.1016/j.paid.2011.07.022 (2011). Min, H. & Su, S. Examining relationships between personality profiles and organizational health outcomes. Pers. Indiv. Differ. 164 , 110118. https://doi.org/10.1016/j.paid.2020.110118 (2020). Morgan, S. et al. Association of personality profiles with depressive, anxiety, and cancer-related symptoms in patients undergoing chemotherapy. Pers. Indiv. Differ. 117 , 130–138. https://doi.org/10.1016/j.paid.2017.05.039 (2017). Musa, L., Ghaderi, A., Chen, Y. & Crawford, J. D. Spatial task instructions and global activation trends influence functional modularity in the cortical reach network. NeuroImage 320 , 121460. https://doi.org/10.1016/j.neuroimage.2025.121460 (2025). Newman, M. E. J. Finding community structure in networks using the eigenvectors of matrices. Phys. Rev. E . 74 (3), 036104. https://doi.org/10.1103/PhysRevE.74.036104 (2006a). Newman, M. E. J. Modularity and community structure in networks. Proceedings of the National Academy of Sciences , 103 (23), 8577–8582. (2006b). https://doi.org/10.1073/pnas.0601602103 Nichols, T. E. & Holmes, A. P. Nonparametric permutation tests for functional neuroimaging: A primer with examples. Hum. Brain. Mapp. 15 (1), 1–25. https://doi.org/10.1002/hbm.1058 (2002). Perera, H. N., Granziera, H. & McIlveen, P. Profiles of teacher personality and relations with teacher self-efficacy, work engagement, and job satisfaction. Pers. Indiv. Differ. 120 , 171–178. https://doi.org/10.1016/j.paid.2017.08.034 (2018). Pilkonis, P. A. et al. Assessment of self-reported negative affect in the NIH Toolbox. Psychiatry Res. 206 (1), 88–97. https://doi.org/10.1016/j.psychres.2012.09.034 (2013). Robins, R. W., John, O. P., Caspi, A., Moffitt, T. E. & Stouthamer-Loeber, M. Resilient, overcontrolled, and undercontrolled boys: Three replicable personality types. J. Personal. Soc. Psychol. 70 (1), 157–171. https://doi.org/10.1037//0022-3514.70.1.157 (1996). Rubinov, M. & Sporns, O. Complex network measures of brain connectivity: Uses and interpretations. NeuroImage Comput. Models Brain . 52 (3), 1059–1069. https://doi.org/10.1016/j.neuroimage.2009.10.003 (2010). Rzeszutek, M. & Gruszczyńska, E. Personality types and subjective well-being among people living with HIV: A latent profile analysis. Qual. Life Res. 29 (1), 57–67. https://doi.org/10.1007/s11136-019-02288-5 (2020). Salsman, J. M. et al. Assessing psychological well-being: Self-report instruments for the NIH Toolbox. Qual. Life Research: Int. J. Qual. Life Aspects Treat. Care Rehabilitation . 23 (1), 205–215. https://doi.org/10.1007/s11136-013-0452-3 (2014). Semeijn, J. H., Van Der Heijden, B., I. J., M. & De Beuckelaer, A. Personality Traits and Types in Relation to Career Success: An Empirical Comparison Using the Big Five. Appl. Psychol. 69 (2), 538–556. https://doi.org/10.1111/apps.12174 (2020). Specht, J., Luhmann, M. & Geiser, C. On the Consistency of Personality Types Across Adulthood: Latent Profile Analyses in Two Large-Scale Panel Studies. SSRN Electron. J. https://doi.org/10.2139/ssrn.2501903 (2014). Talaei, N. & Ghaderi, A. Integration of structural brain networks is related to openness to experience: A diffusion MRI study with CSD-based tractography. Frontiers in Neuroscience , 16 . (2022). https://doi.org/10.3389/fnins.2022.1040799 Udayar, S., Urbanaviciute, I., Massoudi, K. & Rossier, J. The Role of Personality Profiles in the Longitudinal Relationship between Work–Related Well–Being and Life Satisfaction among Working Adults in Switzerland. Eur. J. Pers. 34 (1), 77–92. https://doi.org/10.1002/per.2225 (2020). Van Den Akker, A. L., Deković, M., Asscher, J. J., Shiner, R. L. & Prinzie, P. Personality types in childhood: Relations to latent trajectory classes of problem behavior and overreactive parenting across the transition into adolescence. J. Personal. Soc. Psychol. 104 (4), 750–764. https://doi.org/10.1037/a0031184 (2013). Van Essen, D. C. et al. The WU-Minn Human Connectome Project: An overview. NeuroImage 80 , 62–79. https://doi.org/10.1016/j.neuroimage.2013.05.041 (2013). Van Essen, D. C. et al. The Human Connectome Project: A data acquisition perspective. NeuroImage 62 (4), 2222–2231. https://doi.org/10.1016/j.neuroimage.2012.02.018 (2012). Wang, M. Personality profile analysis, personality-intelligence profile analysis, and the intergenerational transmission of both: Insights from Chinese evidence. Acta. Psychol. 255 , 104957. https://doi.org/10.1016/j.actpsy.2025.104957 (2025). Wang, M. & Hanges, P. J. Latent Class Procedures: Applications to Organizational Research. Organizational Res. Methods . 14 (1), 24–31. https://doi.org/10.1177/1094428110383988 (2011). Wojciechowski, T. Latent Profile Analysis of Personality Dimensions Among Juvenile Offenders: Relevance for Predicting Offending Seriousness. Crime Delinquency . 67 (2), 212–233. https://doi.org/10.1177/0011128720928920 (2021). Yin, K., Lee, P., Sheldon, O. J., Li, C. & Zhao, J. Personality profiles based on the FFM: A systematic review with a person-centered approach. Pers. Indiv. Differ. 180 , 110996. https://doi.org/10.1016/j.paid.2021.110996 (2021). Yu, Y. & Zhang, Y. Personality and Developmental Characteristics of Primary School Students’ Personality Types. Front. Psychol. 12 , 693329. https://doi.org/10.3389/fpsyg.2021.693329 (2021). Tables Table 1. Mean intra-cluster correlation coefficients for overall and domain-specific item responses. Trait Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 5 All Questions 0.3327 0.5719 0.4626 0.5574 0.3274 Agreeableness 0.3099 0.2968 0.3991 0.4871 0.4417 Conscientiousness 0.2369 0.1472 0.1820 0.1593 0.1515 Extraversion 0.1752 0.1805 0.1912 0.2097 0.1819 Neuroticism 0.1676 0.0929 0.2521 0.1968 0.2115 Openness 0.4892 0.4315 0.2715 0.6293 0.3324 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 29 Apr, 2026 Reviewers invited by journal 29 Apr, 2026 Editor invited by journal 29 Apr, 2026 Editor assigned by journal 24 Apr, 2026 Submission checks completed at journal 24 Apr, 2026 First submitted to journal 10 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9379725","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":631951306,"identity":"dc53e094-980d-4c6f-873f-1d3b48add574","order_by":0,"name":"Yalda Shadmanesh","email":"","orcid":"","institution":"University of Isfahan","correspondingAuthor":false,"prefix":"","firstName":"Yalda","middleName":"","lastName":"Shadmanesh","suffix":""},{"id":631951307,"identity":"d8e6c8d1-8af1-495c-8f5e-26bb607d9639","order_by":1,"name":"Amir Hossein Ghaderi","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAvklEQVRIiWNgGAWjYDACCSjih3CZSdAi2UCiFgYGgwPEapGPbn544+MeC3njG+kPPzBUWCc2ENJieOeYseWMZxKG227kGEswnEknQsuMBDNpngMSjEAtbAyMbYeJ0ZL+TfrPAQn7zTPSnzEw/iNCi7xEjpk0wwGJxA0SCWYMjA1EaDGQyCm27DkgkTzjzBtjiYRj6caEbZmRvvHGjwN1tv3twBD7UGMtS9iWA8i8BELKwbYQNHQUjIJRMApGAQD3yj64LXyzSAAAAABJRU5ErkJggg==","orcid":"","institution":"University of Isfahan","correspondingAuthor":true,"prefix":"","firstName":"Amir","middleName":"Hossein","lastName":"Ghaderi","suffix":""}],"badges":[],"createdAt":"2026-04-10 12:53:29","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9379725/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9379725/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108732197,"identity":"cfc80f14-8b05-4059-ae08-7c1616cb8097","added_by":"auto","created_at":"2026-05-07 19:07:13","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":200156,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003e\u0026nbsp;a. \u003c/strong\u003eStep 1: This step depicts the process of constructing a similarity matrix from personality assessments. Spearman’s rank correlation coefficients were computed between participants’ responses to the 60-item NEO-FFI, resulting in a 1198x1198 similarity matrix. Each element of the matrix represents the correlation between the personality profiles of two participants, providing a quantitative measure of similarity based on their response patterns \u003cstrong\u003eb. \u003c/strong\u003eStep 2:\u003cstrong\u003e \u003c/strong\u003eTo systematically evaluate the modular structure within the correlation matrix of NEO FFI items, a hierarchical algorithm based on the resolution parameter (γ) was employed. This algorithm enables the detection of communities within the matrix at varying topological scales, such that increasing γ leads to the identification of smaller subnetworks and, consequently, a greater number of communities. The optimal γ was defined as the value yielding the highest mean modularity ratio (Md) across subnetworks, while limiting isolated nodes to fewer than four. To achieve this, γ was systematically scanned from 1.000 to 1.100 in 0.001 increments. The results of the modularity analysis revealed eight distinct modules, which were illustrated on a 1198 × 1198 similarity matrix derived from the correlation coefficients calculated across all participants. For each resulting cluster, intra-cluster similarity matrices were computed to quantify configural consistency across both the full item set and individual trait domain \u003cstrong\u003ec. \u003c/strong\u003eStep 3: Personality clusters identified through modularity analysis were subjected to permutation t-tests with Bonferroni correction for multiple comparisons.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-9379725/v1/e7fa952992b79c9c1d90a45a.png"},{"id":108807644,"identity":"c95c0862-4abf-4cd3-8586-060cd370d52f","added_by":"auto","created_at":"2026-05-08 15:31:01","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":686159,"visible":true,"origin":"","legend":"\u003cp\u003ea) Pairwise scatter plots of the Big Five personality traits (agreeableness(A), conscientiousness(C), extraversion(E), neuroticism(N), openness to experience (O) colored by data-driven cluster assignment. Each point represents an individual participant, with color denoting cluster membership (refer to the color bar).\u003cstrong\u003e \u003c/strong\u003eb) Spider plot depicting the profile of the big five personality traits across the five clusters identified via trait-based clustering. Each axis represents one of the personality traits (Agreeableness, Conscientiousness, Extraversion, Neuroticism, Openness to Experience), with the trait scores for each cluster represented as shaded areas.\u003cstrong\u003e \u003c/strong\u003ec) Distribution of the Big Five personality traits across five clusters identified through trait-based clustering. Violin plots show the distribution and density of scores for big five traits. Pairwise comparisons between clusters, based on permutation t-tests, are indicated by distinct shape symbols (refer to the key in the right panel). Statistically significant differences at p \u0026lt; 0.001 are indicated by black symbols, while grey-filled symbols denote significance at the p \u0026lt; 0.05 level.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-9379725/v1/2677692a040a756637fdfedb.png"},{"id":108732199,"identity":"0121ef1e-4ca1-4a10-946b-30e25a1579de","added_by":"auto","created_at":"2026-05-07 19:07:13","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":830339,"visible":true,"origin":"","legend":"\u003cp\u003eResults of permutation t-tests between three NIMH toolboxes and their corresponding subscales. In each block, the larger plot (top) represents the overall score for each NIMH index, while the smaller plots (bottom) display scores for the individual questionnaires that are part of each subscale. a) Negative affect scores, b) well-being scores, c) self-efficacy scores.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-9379725/v1/b29b92d9290a29dd3baa8bca.png"},{"id":108810374,"identity":"3fb17973-3a6d-4582-b994-1e494ef8a073","added_by":"auto","created_at":"2026-05-08 15:58:35","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2064644,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9379725/v1/93126d66-f85f-435b-9f7f-c9dd0d3d94f7.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"A High-Dimensional Data-Driven Modularity Analysis Reveals Five Distinct Personality Clusters with Different Psychological Profiles","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe five-factor model of personality (also known as big five)(McCrae \u0026amp; Costa, 1997) has been an extensively researched model in various domains of personality psychology(Yin et al., 2021). The big five encompasses the personality traits of neuroticism (N), extraversion (E), openness to experience (O), agreeableness (A), and conscientiousness (C)(Costa Jr. \u0026amp; McCrae, 2008). Previous research has demonstrated that these personality traits are associated with individuals' psychological and neural characteristics(Kerber et al., 2021; Liu et al., 2019; Talaei \u0026amp; Ghaderi, 2022), and the configurationof these traits can influence various aspects of an individual's life, including affects(Merz \u0026amp; Roesch, 2011), well-being(Leikas \u0026amp; Salmela-Aro, 2014; Liao et al., 2025), and self-efficacy(Perera et al., 2018).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo assess the configuration of personality traits, person-centered approaches (as opposed to variable-based approaches) have been proposed to provide information about individual differences in the patterns of personality traits by grouping individuals with similar values on several personality traits into one personality type(Specht et al., 2014). However, a fundamental question remains: how can meaningful and generalizable personality subgroups be reliably identified given the high dimensionality and complexity of personality assessments(Gerlach et al., 2018)? Traditional clustering methods, such as k-means or unsupervised support vector machine (SVM), are often sensitive to the number of input dimensions and may fail to capture subtle clusters when too many factors (inputs) are involved in the clustering process(Aggarwal \u0026amp; Reddy, 2018; Assent, 2012). Therefore, either the input data must be simplified or alternative or more sophisticated analytic approaches should be used to address this complexity and accurately uncover distinct personality profiles.\u003c/p\u003e\n\u003cp\u003eInstead of developing more sophisticated methods to capture item-level complexity, most previous person-centered approaches have simplified the input data, effectively reducing the dimensionality of personality information. Consequently, individuals with identical mean trait scores can be assumed to belong to the same psychological type, even when their item-level responses differ substantially. Theses person-centered methodologies have evolved over time, from Q-factor analysis (Robins et al., 1996) to more structured approaches such as k-means clustering (Kerber et al., 2021), and more recently to model-based techniques like latent profile analysis (LPA) and latent class analysis (LCA)\u0026nbsp;(Baams et al., 2014; Gerlach et al., 2018; X. Li et al., 2018; Udayar et al., 2020; Bojanić \u0026amp; Čolović, 2025; Liao et al., 2025; Wang, 2025; Wang \u0026amp; Hanges, 2011).\u0026nbsp;Across these studies, three core prototypes have emerged with notable consistency: resilient, overcontrolled, and undercontrolled\u0026nbsp;(Yin et al., 2021). These types are conceptually grounded in Block and Block’s (1980) foundational theory of ego-resiliency and ego-control, which posits that personality structure reflects the dynamic interplay between impulse regulation and adaptive flexibility. In this context, resilient group exhibit high ego-resiliency and demonstrate context-sensitive modulation of ego-control. Overcontrolled group display high ego-control but low ego-resiliency, leading to rigidity and reduced behavioral adaptability. In contrast, undercontrolled group are low on both ego-resiliency and ego-control, often characterized by impulsivity and limited self-regulatory flexibility\u0026nbsp;(Block, J. H., \u0026amp; Block, J., n.d.).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eCompared to findings regarding the characteristics of the overcontrolled and undercontrolled groups, findings regarding the characteristics of the resilient group have been more consistent\u0026nbsp;(Bojanić \u0026amp; Čolović, 2025; Yin et al., 2021). A recent meta-analysis study of latent profile analysis (LPA) studies quantified the trait composition of these types across large samples. The resilient group consistently showed high levels of E (82.76% of studies), A (96.55% of studies), C (86.21% of studies), and O (79.31% of studies), alongside very low N (over 93% at the lowest level). The undercontrolled group typically exhibited low C (87.5% of studies), low A (75% of studies), and low E (37.5% of studies), combined with elevated N (50% of studies). The overcontrolled group were characterized by high N (61% of studies) and low E (72% of studies), with other traits falling in the low-to-moderate range(Yin et al., 2021). Accordingly, some studies report two-profile models(Semeijn et al., 2020; Wojciechowski, 2021), while others identify three(Robins et al., 1996; Van Den Akker et al., 2013), four\u0026nbsp;(Chi \u0026amp; Chi, 2023; Kerber et al., 2021; Rzeszutek \u0026amp; Gruszczyńska, 2020; Udayar et al., 2020), or five profiles\u0026nbsp;(Conte et al., 2017; Kinnunen et al., 2012; X. Li et al., 2018). Notably, Gerlach et al. (2018), using a data-driven Gaussian mixture model over the large sample datasets, identified four replicable personality types, further challenging the rigidity of the three-profile model(Gerlach et al., 2018).\u003c/p\u003e\n\u003cp\u003eAs noted above, these findings were obtained using methods that relied on aggregated trait-level scores. It remains possible that preserving the item-level complexity of personality responses could lead to different results, potentially revealing more nuanced or distinct patterns that are obscured when data are simplified. Rather than reducing input dimensionality, which can inadvertently remove meaningful variability, it may be more advantageous to employ analytical frameworks capable of accommodating and interpreting high-dimensional structure directly. Approaches informed by modern graph-theoretic principles, among others, are well suited for capturing such complexity without collapsing the data. In particular, graph theoretical community-detection methods based on the modularity framework (Newman, 2006b, 2006a) allow one to identify naturally emerging modules (subnetworks) of highly interrelated items. Applying such modularity analyses to item-level personality response networks could reveal latent subtypes or clusters of traits that are not apparent when using aggregated scores.\u003c/p\u003e\n\u003cp\u003eHere, we introduce a novel graph-theoretic, item-pattern-based modularity (IPBM) analysis to evaluate and cluster personality profiles at the item level. The first objective of this study was to characterize the structure of personality types using this new data-driven framework and to compare the resulting typology with those derived from traditional aggregation-based approaches. In contrast with previous person-centered methodologies, the presented method identifies personality types as emergent communities of individuals who share structurally similar response profiles. Specifically, the IPBM method constructs a network in which nodes represent individuals and edges reflect the similarity (Spearman’s rank correlation coefficient) of their item-level responses. Communities within this network are then identified using an optimization approach, which seeks to detect the most robust modules of individuals sharing similar item-level response patterns (as illustrated in Figure 1-a). Given that this approach retains the full complexity of the input data, we hypothesized that it may reveal distinct and potentially understudied personality groupings that remain undetected when relying on aggregated trait scores.\u003c/p\u003e\n\u003cp\u003eThe second objective of this study was to determine whether the personality groups identified through the IPBM framework show meaningful differences in their psychological and neural characteristics.\u0026nbsp;We specifically focused on well-being, self-efficacy, and negative affect. In this context, previous research have mostly focused on the links between personality and well-being\u0026nbsp;(Busseri \u0026amp; Erb, 2024; HAN, 2020; Henning et al., 2017; Mammadov, 2023; Mammadov \u0026amp; Ward, 2022; Marengo et al., 2021; Rzeszutek \u0026amp; Gruszczyńska, 2020; Udayar et al., 2020). These studies have shown that the resilient profile scores highest on measures of subjective well-being\u0026nbsp;(Mammadov et al., 2024; Mammadov \u0026amp; Ward, 2023; Mammadov, 2023; Udayar et al., 2020; Rzeszutek \u0026amp; Gruszczyńska, 2020).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eRegarding self-efficacy, Perera et al. showed a cluster with low neuroticism and above-average levels of the other traits, aligns with the resilient profile, which demonstrated the highest levels of self-efficacy. In contrast, the excitable profile, characterized by slightly above-average neuroticism, higher levels of extraversion, openness, and agreeableness, and below-average conscientiousness, aligns with the undercontrolled profile, which exhibited the lowest levels of self-efficacy (Perera et al., 2018). To the best of our knowledge, no previous study has investigated the relationship between person-centered personality profiles and negative affect.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eRegarding the second objective of the study, which was to examine psychological and neural differences between the groups identified using the IPBM method, our hypothesis was that the different groups would exhibit distinct psychological profiles, resulting in differences in well-being, self-efficacy, negative affect, and related measures. This is particularly interesting and informative for the less commonly observed groups that may emerge from the IPBM analysis, and it could also replicate some previous findings suggesting that groups with profiles resembling resilient or undercontrolled types tend to show significant differences from other groups. If such results are obtained, they would provide validation for the clustering approach used in this study and highlight new findings that align with prior research.\u0026nbsp;\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eIn this study, we used a graph theoretical modularity approach to cluster individuals based on their responses to the 60 single items of the NEO-FFI.\u0026nbsp;Spearman’s rank correlation\u0026nbsp;analysis was performed across these items, generating an adjacency matrix that captured the relationships between participants based on their personality responses. This matrix was then used for a graph theoretical modularity analysis to identify personality-based modules (groups). To explore the differences among personality clusters, big five dimensions, and NIMH toolboxes, statistical analyses were applied (figure 1).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eParticipant Clustering via Modularity Analysis\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAs shown in figure 1-b, modularity analysis revealed that the optimal modularity\u0026nbsp;was achieved at a modularity coefficient of γ = 1.042, with a mean modularity ratio of 1.51. Higher values of γ led to excessive fragmentation of the network, resulting in an increased number of isolated nodes beyond our acceptable threshold, whereas lower γ values yielded modularity values below the optimal range.\u003c/p\u003e\n\u003cp\u003eAt this optimal modularity coefficient, participants were clustered into eight distinct modules based on inter-individual similarity in 60-items NEO-FFI responses. Among these, three modules were identified as an isolated cluster (with \u003cem\u003en\u003c/em\u003e \u0026lt;= 5) and excluded from subsequent analyses. Five modules, clusters 1 (\u003cem\u003en\u003c/em\u003e = 398), 2 (\u003cem\u003en\u003c/em\u003e = 326), 3 (\u003cem\u003en\u003c/em\u003e = 312), 4 (\u003cem\u003en\u003c/em\u003e = 17), and 5 (\u003cem\u003en\u003c/em\u003e = 134), were selected for downstream analysis.\u0026nbsp;Table 1 summarizes the mean intra-cluster similarity and dimension-specific similarity for each of the five clusters. Additionally, the scatter plot of big five factors for different individuals in different groups (identified by color codes) is presented in figure 2-a.\u003c/p\u003e\n\u003cp\u003eThe split-half reliability analysis confirmed the stability of our findings, as the modularity analysis consistently yielded the same five distinct personality clusters in both independent halves of the dataset.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePersonality Traits Differences Between Clusters\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAccordingly, we evaluate the mean and standard deviation (SD) of big five factors for each group (figure 2.b) and results showed:\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCluster 1:\u003c/em\u003e cluster 1 showed the lowest mean level of C (mean = 30.74) and the highest mean level of O (mean = 34.15) among all clusters. Variability in N was markedly high (SD = 7.26), indicating substantial heterogeneity within the cluster. Similarly, E scores displayed elevated dispersion (SD = 6.13).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCluster 2:\u003c/em\u003e As shown in figure 2-a, cluster 2 was characterized by the highest level of A (mean = 34.14) and the lowest level of N (mean = 9.90) across all clusters. Additionally, this cluster demonstrated relatively high scores in E (mean = 33.22) and C (mean = 36.25).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCluster 3:\u003c/em\u003e As shown in figure 2-a, cluster 3 showed the lowest mean score in O (mean = 22.56). It ranked second in both E (mean = 30.79± = 5.30) and C (mean = 37.01± 4.84).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCluster 4\u003c/em\u003e: Cluster 4 exhibited the highest mean scores in C (mean = 39.76± 5.45) and E (mean = 33.29± 4.57) among all clusters.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cem\u003eCluster 5\u003c/em\u003e: Cluster 5 exhibited the lowest mean scores across all clusters in A (mean = 30.11± 5.64) and E (mean = 26.90 ± 5.67). Notably, it showed the highest level of N (mean = 23.55± 5.80).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThen, to evaluate differences in big five personality traits across the identified clusters, we performed a series of permutation t-tests (corrected with Bonferroni correction for multiple comparisons). The analysis revealed several significant differences, as shown in Figure 2-c. Most significant differences between groups were observed in O (9 significant differences between pairs) while the analyses showed only 3 significant differences between group pairs in A. Furthermore, cluster 5 showed the most significant differences with other groups in all five traits (18 pairs of significant differences). Comprehensive statistical results for all pairwise comparisons are available in Supplementary Information (Table S1 and S2).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCluster-Specific NIMH Profiles\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eNegative affect.\u0026nbsp;\u003c/em\u003eAs shown in figure 3-a, cluster 2 showed the lowest values of negative affect among clusters\u0026nbsp;(mean = 46.77±4.69) and exhibited significant differences with all other clusters (t(1,2) = 14.20, p \u0026lt; 0.001; t(2,3) = -4.67, p \u0026lt; 0.001; t(2,4) = -3.82, p = 0.002; t(2,5) = -13.47, p \u0026lt; 0.001). In contrast, clusters 1 (mean = 52.20 ±5.44) and 5 (mean = 53.92 ±6.17) showed the highest mean values of negative affect and exhibited significant differences with cluster 3 (t (1,3) = 9.01, p \u0026lt; 0.001; t (3,5) = -9.45, p \u0026lt; 0.001). The significant highest values of cluster 5 in negative affect dimensions, is also replicated in many pair comparisons (12 pairs of significant differences). Conversely, cluster 2 showed the significant lowest values in many pair comparisons (12 comparisons) in negative affect dimensions. Among the dimensions most significant differences were observed in anger hostility (6 significant differences), while only one significant difference was observed in anger physical aggression.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eWell-being.\u003c/em\u003e As shown in Figure 3-b, cluster 2 showed the highest values of well-being among clusters (mean=55.07 ±6.11) and exhibited significant differences with most of the other clusters\u0026nbsp;(t\u0026nbsp;(2,1) = -10.26, p \u0026lt; 0.001; t\u0026nbsp;(2,3) = 4.03, p = 0.004; t\u0026nbsp;(2,5) = 11.07, p \u0026lt; 0.001). In contrast, clusters 1 (mean=50.01± 6.99) and 5 (mean=47.86± 6.83) showed the lowest mean values of well-being and exhibited significant differences with cluster 3 (t(1,3) = -5.68, p \u0026lt; 0.001; t(3,5) = 7.20, p \u0026lt; 0.001). The significantly highest values of cluster 2 is also replicated in many pair comparisons (7 pairs of significant differences) in dimensions of well-being. Conversely, cluster 5 showed the significantly lowest values in many pair comparisons (7 pairs of significant differences) in well-being dimensions. Among the dimensions, the most significant differences were observed in meaning and purpose (6 significant differences), while only three significant differences were observed in positive affect.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eSelf-efficiency.\u003c/em\u003e As shown in Figure 3-c, cluster 2 showed the highest values of self-efficiency among clusters (mean=55.78± 6.83) and exhibited significant differences with all other clusters (t(2,1) = -12.81, p \u0026lt; 0.001; t(2,3) = 8.22, p \u0026lt; 0.001; t(2,4) = 3.26, p = 0.038; t(2,5) = 13.60, p \u0026lt; 0.001). In contrast, clusters 1(mean=52.20 ±5.44) and 5 (mean=53.92 ±6.17) showed the lowest mean values of self-efficiency and exhibited significant differences with cluster 3 and with each other (t (1,3) = -4.40, p \u0026lt; 0.001; t(1,5) = 3.99, p = 0.002; t(3,5) = 7.39, p \u0026lt; 0.001). In dimensions of self-efficiency, the significantly highest values of cluster 2 are also replicated in many pair comparisons (7 pairs of significant differences). Conversely, cluster 5 showed the significantly lowest values in many pair comparisons (4 significant comparisons) in the dimensions. Among the dimensions, the most significant differences were observed in perceived stress (6 significant differences), while only four significant difference was observed in self-efficiency. Comprehensive statistical results for all pairwise comparisons are available in supplementary tables S3 to S6.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eOur findings demonstrate that the proposed IPBM approach can effectively derive a limited and interpretable set of personality groups from item-level response patterns. The resulting five clusters not only reproduced well-established personality configurations commonly described in the literature, but also revealed additional, less frequently discussed profiles, including one cluster that, to the best of our knowledge, has not been characterized previously. These distinctions were meaningfully reflected in psychological outcomes: clusters differed significantly in well-being, negative affect, and self-efficacy, indicating that item-level response structure captures consequential variation in personality expression.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eThe Importance of the IPBM Method in Regulating Input Complexity\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAs established in prior literature, high-dimensional spaces fundamentally erode the utility of pairwise distance metrics, the very foundation of classical clustering algorithms (Aggarwal \u0026amp; Reddy, 2018; Assent, 2012).\u0026nbsp;A central issue is the \u003cem\u003ecurse of dimensionality\u003c/em\u003e, whereby increasing the number of features renders data points increasingly sparse and diminishes the meaningfulness of distance metrics that these algorithms rely upon. In high dimensions, objects with different underlying cluster structure may appear deceptively similar due to the concentration of distances, which complicates both cluster separation and cluster validation in unsupervised learning. Empirically, this effect has been shown to undermine the performance and interpretability of classical clustering methods as dimensionality increases, particularly when naïvely using raw feature vectors without dimensionality reduction or feature selection\u0026nbsp;(Assent, 2012).\u0026nbsp;Consequently, traditional clustering techniques, such as k-means or model-based profile estimation, often falter with high-dimensional inputs which may\u0026nbsp;suffers from increased sensitivity to irrelevant features and poor separation between clusters, which degrades both cluster quality and reproducibility\u0026nbsp;(Kadir et al., 2014). Furthermore,\u0026nbsp;naïve use of distance-based clustering can result in artificially inflated numbers of clusters that primarily reflect noise and high feature count rather than meaningful subgroup distinctions\u0026nbsp;(Assent, 2012).\u003c/p\u003e\n\u003cp\u003eWhen classical clustering methods are applied to identify personality groups based on responses to the 60-item NEO-FFI personality questionnaire, these high-dimensionality–related issues inevitably arise. As a result, the derived clustering solutions can become highly unstable and strongly influenced by noise, leading to poor reproducibility and limited psychological interpretability. This analytical challenge has historically reinforced the field's reliance on aggregated trait scores, a pragmatic dimensional reduction that regrettably obscures meaningful, within-trait response-pattern variability.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBy adopting a graph-theoretic modularity analysis, our approach directly circumvents this limitation. The modularity optimization framework efficiently uncovers latent community structures by maximizing within-group similarity against a randomized null model, thereby enabling the discovery of a stable, parsimonious, and optimally constrained five-cluster solution. This outcome demonstrates how graph-based techniques can preserve the granular richness of item-level data while inherently resisting the over-fragmentation and instability that plague classical clustering under high-dimensional conditions.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cem\u003eCharacteristics of Our\u0026nbsp;\u003c/em\u003e\u003cem\u003eFive-Cluster Personality Structure\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eOur analysis, with the exclusion of a very small number of participants (around 0.2% of the sample), was able to optimally cluster individuals into five groups. As described earlier, these groups were formed based on similarities and differences in individuals’ responses to each individual questionnaire item; however, when examined at the level of the five aggregated personality factors, the groups also exhibited numerous statistically significant differences. We further conducted within-group correlation analyses to identify which of the big five factors likely played a more prominent role in the formation of each group. Overall, these results indicated that two of the identified groups correspond well to previously established personality profiles, namely the resilient and under-controlled types. In contrast, the remaining three groups exhibited relatively distinct and more novel characteristics compared to those reported in many prior studies.\u003c/p\u003e\n\u003cp\u003eIn particular, our second cluster shows a strong correspondence with the \u003cstrong\u003eresilient\u003c/strong\u003e personality type reported in prior studies. This profile is typically characterized by relatively \u003cstrong\u003ehigh levels of extraversion, agreeableness, conscientiousness, and openness to experience\u003c/strong\u003e\u003cstrong\u003e,\u003c/strong\u003e together with \u003cstrong\u003elow levels of neuroticism\u003c/strong\u003e according to the big five framework\u0026nbsp;(Gerlach et al., 2018; Kerber et al., 2021; Liao et al., 2025; Min \u0026amp; Su, 2020; Udayar et al., 2020; Wojciechowski, 2021; Yin et al., 2021; Yu \u0026amp; Zhang, 2021). In our analysis, this cluster demonstrated the highest degree of internal similarity across all items, indicating a highly consistent and well-defined response pattern among its members. Furthermore, their responses to openness-related items were highly correlated, suggesting that members of this cluster are highly similar in how they express openness to experience. In agreement with previous findings (Busseri \u0026amp; Erb, 2024; Perera et al., 2018), this cluster showed high scores across well-being and self-efficacy indices and low scores across negative affect by NIMH toolbox.\u003c/p\u003e\n\u003cp\u003eAnother cluster identified in our analysis was cluster 5, which can be closely aligned with the undercontrolled\u0026nbsp;(Baams et al., 2014; Kerber et al., 2021; Yin et al., 2021; Yu \u0026amp; Zhang, 2021) or vulnerable in (De Clercq et al., 2012) or distressed in (Morgan et al., 2017) types. Previous\u0026nbsp;studies suggest that individuals in this group typically exhibit high levels of neuroticism, coupled with low levels of conscientiousness and agreeableness which are closely matched with the aggregated scores of big five factor in cluster 5.\u0026nbsp;In our analysis, the cluster 5 showed the highest degree of internal similarity across neuroticism which suggest individuals in this cluster share core elements of emotional distress and interpersonal challenges.\u0026nbsp;Consistently, higher negative affect and lower self-efficacy and well-being scores in this cluster, compared with the other clusters, reflect difficulties in emotional regulation, impulse control, and behavioral inhibition within this group\u0026nbsp;(Perera et al., 2018; Robins et al., 1996; Asendorpf et al., 2001).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAmong the three remaining clusters identified in our analysis, two can still be meaningfully compared with profiles reported in previous studies. In particular, cluster 3 appears to share certain characteristics with the resilient type; however, it also shows notable differences relative to cluster 2, which represented the prototypical resilient profile in our analysis.\u0026nbsp;In fact, cluster 3 highlights the discriminative power of our IPBM approach by revealing a distinct profile that would typically be merged with the conventional resilient type in aggregate-based analyses. This cluster emerged from a response pattern characterized by high conscientiousness and agreeableness, and relatively low neuroticism and openness. While this aggregate profile bears resemblance to the resilient cluster identified by Van den Akker et al(Van Den Akker et al., 2013), our item-level analysis distinguished it from the classic resilient profile (cluster 2) in our own data. To our knowledge, no prior study has reported the simultaneous presence of these two distinct resilient-like profiles, and this novel finding suggests the existence of more than a single resilient prototype. This profile showed the high intra-consistency in neuroticism and the low intra-consistency in openness, in contrast to cluster 2, which exhibited low similarity in neuroticism alongside uniformly in openness. The difference between these two resilience-like clusters confirmed by NIMH questionnaire data, which showed significant differences in negative affect, well-being and life satisfaction between the two clusters. On these measures, cluster 2 demonstrated more favorable life outcomes, with significantly lower negative affect and higher well-being and life satisfaction. This suggests that life outcome measures may differ substantially between the two resilient groups, depending on whether they are primarily organized around neuroticism or, conversely, around openness.\u003c/p\u003e\n\u003cp\u003eAnother group that has been only partially addressed in previous studies corresponds to cluster 1 in our analysis.\u0026nbsp;This cluster is characterized by a high level of openness to experience and markedly low conscientiousness. In this context, a comparable cluster, referred to as free-spirit, was previously identified by Henning et al. in a Swedish elderly cohort(Henning et al., 2017). In terms of well-being and self-efficiency, this cluster ranks second-lowest, and for negative affect ranks second highest, we attribute this pattern to the elevated neuroticism levels within the group, though this hypothesis warrants further investigation.\u003c/p\u003e\n\u003cp\u003eThe final cluster discussed here is cluster 4, which comprises a very small proportion of our sample yet consistently emerges as a distinct cluster. Despite its limited size, this group exhibits a set of unique and characteristic features that clearly differentiate it from the other clusters. This cluster is characterized by the high scores in conscientiousness, agreeableness, extraversion, openness and moderate scores in neuroticism.\u0026nbsp;The intra-cluster analysis of this cluster revealed homogeneous response patterns for both openness and agreeableness, suggesting a high degree of similarity among individuals in this cluster with respect to these two traits.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAn important observation is that this group reported the highest levels of meaning and purpose, which co-occurred with elevated scores on negative affect subscales, particularly anger–hostility and anger–physical aggression. This pattern suggests that, for these individuals, a strong sense of purpose may not fully buffer against feelings of irritability or frustration. Consistent with this interpretation, this group also exhibited persistently low levels of self-efficacy.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study shows that an item-level, graph-based personality modeling approach can reliably extract a small and interpretable set of personality prototypes while preserving meaningful individual differences. By regulating input complexity through modularity optimization, the IPBM framework overcomes key limitations of classical clustering in high-dimensional spaces and recovers both well-established and less-studied personality profiles. The identified clusters differed robustly in well-being, negative affect, and self-efficacy, indicating that item-level response structure captures psychologically consequential variation beyond aggregated trait scores. Overall, these findings highlight the value of network-based, data-driven approaches for refining personality typologies and linking them to behavioral outcomes.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eLimitations\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eDespite its contributions, this study has several limitations that should be addressed in future research. First, the data were drawn exclusively from the Human Connectome Project, which primarily represents a sample from the United States. Consequently, the generalizability of the identified personality profiles across diverse cultural and demographic contexts remains uncertain. Replicating this study in different samples from various regions is essential to validate the universality or cultural specificity of these clusters.\u0026nbsp;\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cp\u003e\u003cem\u003eParticipants\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThis study used the behavioral data of the HCP Young Adult 1200 Subjects release 10. The initial dataset comprised 1206 participants but data for 8 individuals was excluded, as their questionnaire responses were not recorded. Consequently, the final analytical sample consisted of 1198 participants (650 females), with a median age of 28.7 years (age range: 22-37, SD: 3.67). Inclusion criteria required no history of major psychiatric, neurological, or cardiovascular disorders, the capacity to provide informed consent, and a Mini-Mental State Examination score of 29\u0026ndash;30. Individuals who were asymptomatic but reported a history of smoking, overweight status, or recreational substance use were not excluded, in line with the project\u0026rsquo;s aim to capture normative variation. Exclusion criteria included diagnosed neuropsychiatric or neurological conditions, genetic disorders, use of psychoactive or hormonal medications within the past year, moderate to severe traumatic brain injury, preterm birth (\u0026lt;37 weeks gestation), current pregnancy, and contraindications for MRI scanning (e.g., implanted metal devices or severe claustrophobia) (Van Essen et al., 2012, 2013).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis study used publicly available, de-identified data from the Human Connectome Project (HCP). All participants provided written informed consent, and all recruitment and data acquisition procedures were approved by the Washington University Institutional Review Board, in accordance with the Declaration of Helsinki and all relevant guidelines and regulations (Christova et al., 2020; Van Essen et al., 2012, 2013). The present study involved secondary analysis of anonymized data and did not require additional ethical approval.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eBehavioral data\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePersonality data\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe 60-item version of the Costa and McCrae NEO-FFI (McCrae \u0026amp; Costa, 2004), which has shown excellent reliability and validity, was administered to HCP subjects(Christova et al., 2020; Van Essen et al., 2013). This measure was collected as part of the Penn Computerized Cognitive Battery [version (NEO-FFI-2, 2004)] (Costa Jr. \u0026amp; McCrae, 2008; Gur, 2001) The NEO-FFI is a self-report questionnaire with 60 items (abbreviated version of the 240-item inventory). For each item, participants reported their level of agreement on a 5-point Likert scale, from strongly disagree to strongly agree (strongly disagree=1; disagree=2; neither agree nor disagree =3; agree=4; strongly agree=5) (Christova et al., 2020; Van Essen et al., 2013). Openness to experience, conscientiousness, extraversion, agreeableness, and neuroticism scores are derived by coding each item\u0026rsquo;s answer. Both the total score on each personality factor and subjects\u0026apos; responses to each of the 60 individual personality items were used. The total score was used for comparisons between groups (defined in the next section) and the 60 individual personality items scores were used as the series for correlation analysis and generating the similarity matrix.\u0026nbsp;Prior to any further analysis, all reverse-keyed items were reverse-scored to ensure that the numerical values consistently aligned with the direction of the intended personality traits.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eNIMH Toolboxes\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eIn accordance with the standardized HCP protocol(Van Essen et al., 2012), three assessment batteries from the NIH toolbox, well-being, negative affect, and stress and self-efficacy, were administered. All scores were provided as standardized T-scores (M = 50, SD = 10), normed against a nationally representative sample. For each battery, T-scores for the constituent surveys and an overall composite score (computed as the average of those surveys) were used(Christova et al., 2020). Psychological well-being was assessed using measures of positive affect, life satisfaction, and meaning and purpose\u0026nbsp;(Salsman et al., 2014). Negative affect was evaluated using six constituent surveys: sadness, fear-affect, fear\u0026ndash;somatic arousal, anger-affect, anger\u0026ndash;hostility, and anger\u0026ndash;physical aggression. The first four referenced emotional experiences over the past 7 days, while the latter two assessed trait-like dispositions(Pilkonis et al., 2013). The stress and self-efficacy battery included perceived stress and self-efficacy(Kupst et al., 2015; G. Li et al., 2021). For further details regarding each of the three NIH toolbox batteries, please refer to the supplementary materials.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCorrelation-based personality analysis and similarity matrix\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eWe used the 60 individual personality items scores for each participant as a series for Spearman\u0026rsquo;s correlation analysis and generating a similarity matrix. As we mentioned before, all reverse-keyed items were reverse-scored.\u0026nbsp;The value along each point in the series, reflects the participant\u0026rsquo;s Likert-scale response (figure1.a). Subsequently,\u0026nbsp;Spearman\u0026rsquo;s rank correlation coefficient\u0026nbsp;was calculated between different individual series. Then the values of correlation coefficients were arranged in a shape of a similarity matrix. This outcome 1198 by 1198 matrix was a symmetric square matrix represent the personality similarity among all individuals (figure 1.a), in which each element indicates the linear correlation (Spearman\u0026rsquo;s rank correlation coefficient) between the response patterns of a pair of participants across the entire set of items.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eModularity analysis and subnetworks\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eTo reveal the modular architecture embedded within the similarity matrix, we implemented a combined approach integrating both unsupervised and supervised modularity analyses\u0026nbsp;(Ghaderi et al., 2025; Newman, 2006b, 2006a). Initially, we employed the modularity_und function from the Brain Connectivity Toolbox to detect intrinsic modules based solely on the internal edge structure of the network\u0026nbsp;(Rubinov \u0026amp; Sporns, 2010). This unsupervised algorithm includes a tunable resolution parameter,\u0026nbsp;\u003cimg width=\"8\" height=\"17\" src=\"https://myfiles.space/user_files/69519_bce2c0439cd956a6/69519_custom_files/img1778180639.png\" v:shapes=\"_x0000_i1025\" alt=\"image\"\u003e, which regulates the granularity of partitioning: lower\u0026nbsp;\u003cimg width=\"8\" height=\"17\" src=\"https://myfiles.space/user_files/69519_bce2c0439cd956a6/69519_custom_files/img177818063957.png\" v:shapes=\"_x0000_i1025\" alt=\"image\"\u003e\u0026nbsp;values tend to yield larger modules, while higher values promote finer subdivisions. Importantly, this step relies purely on topological connectivity without recourse to external validation metrics.\u003c/p\u003e\n\u003cp\u003eTo identify the optimal\u0026nbsp;\u003cimg width=\"8\" height=\"17\" src=\"https://myfiles.space/user_files/69519_bce2c0439cd956a6/69519_custom_files/img177818063949.png\" v:shapes=\"_x0000_i1025\" alt=\"image\"\u003e, we introduced a supervised evaluation framework based on a modularity ratio (MR)\u0026nbsp;(Ghaderi et al., 2025; Musa et al., 2025). Starting from a\u0026nbsp;\u003cimg width=\"8\" height=\"17\" src=\"https://myfiles.space/user_files/69519_bce2c0439cd956a6/69519_custom_files/img177818063926.png\" v:shapes=\"_x0000_i1025\" alt=\"image\"\u003e\u0026nbsp;that produced at least two modules, we computed the average within-module connectivity and compared it against a null distribution derived from randomized networks with preserved degree distributions. The MR was defined as the ratio of empirical to randomized within-module connectivity. This procedure was iterated across all detected modules, and the mean MR was used as a quality index for each\u0026nbsp;\u003cimg width=\"8\" height=\"17\" src=\"https://myfiles.space/user_files/69519_bce2c0439cd956a6/69519_custom_files/img1778180640.png\" v:shapes=\"_x0000_i1025\" alt=\"image\"\u003e\u0026nbsp;value.\u003c/p\u003e\n\u003cp\u003eBy incrementally increasing\u0026nbsp;\u003cimg width=\"8\" height=\"17\" src=\"https://myfiles.space/user_files/69519_bce2c0439cd956a6/69519_custom_files/img177818063985.png\" v:shapes=\"_x0000_i1025\" alt=\"image\"\u003e\u0026nbsp;and repeating this process, we obtained a curve of mean MR values, from which the\u0026nbsp;\u003cimg width=\"8\" height=\"17\" src=\"https://myfiles.space/user_files/69519_bce2c0439cd956a6/69519_custom_files/img177818063939.png\" v:shapes=\"_x0000_i1025\" alt=\"image\"\u003e\u0026nbsp;maximizing MR was selected, provided the number of isolated nodes remained below four. This optimal\u0026nbsp;\u003cimg width=\"8\" height=\"17\" src=\"https://myfiles.space/user_files/69519_bce2c0439cd956a6/69519_custom_files/img177818064021.png\" v:shapes=\"_x0000_i1025\" alt=\"image\"\u003e\u0026nbsp;was then fixed for all subsequent modular analyses (fig1.b). All analyses were conducted in MATLAB 2024a, with custom scripts publicly accessible at https://github.com/AHGhaderi/Amir-Hossein-Ghaderi/commit/df636b2105e578a8969ffa88856e54f3267a40a7.\u003c/p\u003e\n\u003cp\u003eFollowing cluster identification, we computed two types of correlation matrices for each cluster. First, an intra-cluster similarity matrix was derived from participants\u0026apos; responses to all 60 items. Second, domain-specific inter-individual similarity matrices were calculated based on participants\u0026apos; responses\u0026nbsp;to the 12 individual items\u0026nbsp;corresponding to each Big Five domain. The mean correlation coefficients for all matrices are reported.\u0026nbsp;To evaluate the magnitude of\u0026nbsp;these\u0026nbsp;correlation coefficients, we followed the guidelines proposed by Gignac \u0026amp; Szodorai\u0026nbsp;(Gignac \u0026amp; Szodorai, 2016), where correlations of 0.10, 0.20, and 0.30 are considered relatively small, typical, and relatively large, respectively, in the context of individual differences research.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePermutation t-tests\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eWe compared personality trait scores across the five clusters identified via modularity-based analysis (Methods, section 2.4). Pairwise differences on the five NEO personality dimensions were assessed using nonparametric permutation t-tests for independent samples (5,000 iterations), with Bonferroni correction for multiple comparisons (Bonferroni, 1936). This approach preserves the between-subject structure without assuming normality. The same framework was used to assess NIMH toolboxes (integrated scales and individual domains) across the five clusters. All analyses were conducted in MATLAB using custom scripts, and the permutation test was implemented following standard guidelines((Nichols \u0026amp; Holmes, 2002); available at: https://github.com/AHGhaderi/Amir-Hossein-Ghaderi/commit/4f2c8f731d82707cc06dc20e04fd3ca01b9e8e02).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eRobustness and Reliability Analysis\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eTo evaluate the robustness and stability of the clustering solution, we performed a split-half reliability analysis. The total dataset (N=1198) was randomly partitioned into two independent halves. The identical modularity optimization pipeline used for the full dataset was then applied to each subset separately to determine whether the structure of clusters could be independently replicated.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThe authors received no financial support for the research, authorship, and/or publication of this article.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eY.S.: Conceptualization, Methodology, Software, Investigation, Writing - Original Draft. A.H.G.: Supervision, Conceptualization, Formal analysis, Validation, Writing - Review \u0026amp; Editing.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe data used in this study are available through the Human Connectome Project (HCP) at https://www.humanconnectome.org/.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAggarwal, C. C. \u0026amp; Reddy, C. K. (eds) \u003cem\u003eData Clustering: Algorithms and Applications\u003c/em\u003e 1st edn (Chapman and Hall/CRC, 2018). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1201/9781315373515\u003c/span\u003e\u003cspan address=\"10.1201/9781315373515\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAssent, I. Clustering high dimensional data. \u003cem\u003eWIREs Data Min. Knowl. Discov.\u003c/em\u003e \u003cb\u003e2\u003c/b\u003e (4), 340\u0026ndash;350. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/widm.1062\u003c/span\u003e\u003cspan address=\"10.1002/widm.1062\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaams, L., Overbeek, G., Dubas, J. S. \u0026amp; Van Aken, M. A. G. On Early Starters and Late Bloomers: The Development of Sexual Behavior in Adolescence Across Personality Types. \u003cem\u003eJ. Sex. Res.\u003c/em\u003e \u003cb\u003e51\u003c/b\u003e (7), 754\u0026ndash;764. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/00224499.2013.802758\u003c/span\u003e\u003cspan address=\"10.1080/00224499.2013.802758\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBlock, J. H. \u0026amp; Block, J. (n.d.). \u003cem\u003eThe role of ego-control and ego resiliency in the organization of behavior\u003c/em\u003e \u0026ndash; \u003cem\u003eScienceOpen\u003c/em\u003e. Retrieved July 16, from (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.scienceopen.com/book?vid=0a410035-d7e6-4d0c-abd6-215f15fe58ee\u003c/span\u003e\u003cspan address=\"https://www.scienceopen.com/book?vid=0a410035-d7e6-4d0c-abd6-215f15fe58ee\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBojanić, M. \u0026amp; Čolović, P. Personality profiles in SAPA data: An exploratory study. \u003cem\u003ePers. Indiv. Differ.\u003c/em\u003e \u003cb\u003e234\u003c/b\u003e, 112957. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.paid.2024.112957\u003c/span\u003e\u003cspan address=\"10.1016/j.paid.2024.112957\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBonferroni, C. E. \u003cem\u003eTeoria statistica delle classi e calcolo delle probabilit\u0026agrave;\u003c/em\u003e (Seeber, 1936).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBusseri, M. A. \u0026amp; Erb, E. M. The happy personality revisited: Re-examining associations between Big Five personality traits and subjective well‐being using meta‐analytic structural equation modeling. \u003cem\u003eJ. Pers.\u003c/em\u003e \u003cb\u003e92\u003c/b\u003e (4), 968\u0026ndash;984. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/jopy.12862\u003c/span\u003e\u003cspan address=\"10.1111/jopy.12862\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChi, J. \u0026amp; Chi, Y. N. Cluster Analysis of Personality Types Using Respondents\u0026rsquo; Big Five Personality Traits. \u003cem\u003eInt. J. Data Sci.\u003c/em\u003e \u003cb\u003e4\u003c/b\u003e (2), 116\u0026ndash;135. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.18517/ijods.4.2.116-135.2023\u003c/span\u003e\u003cspan address=\"10.18517/ijods.4.2.116-135.2023\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChristova, P., Joseph, J. \u0026amp; Georgopoulos, A. P. Behavioral\u0026ndash;genetic associations in the Human Connectome Project. \u003cem\u003eExp. Brain Res.\u003c/em\u003e \u003cb\u003e238\u003c/b\u003e (11), 2445\u0026ndash;2456. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00221-020-05893-w\u003c/span\u003e\u003cspan address=\"10.1007/s00221-020-05893-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eConte, J. M., Heffner, T. S., Roesch, S. C. \u0026amp; Aasen, B. A person-centric investigation of personality types, job performance, and attrition. \u003cem\u003ePers. Indiv. Differ.\u003c/em\u003e \u003cb\u003e104\u003c/b\u003e, 554\u0026ndash;559. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.paid.2016.09.004\u003c/span\u003e\u003cspan address=\"10.1016/j.paid.2016.09.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCosta, P. T. Jr. \u0026amp; McCrae, R. R. The Revised NEO Personality Inventory (NEO-PI-R). In The SAGE handbook of personality theory and assessment, Vol 2: Personality measurement and testing (179\u0026ndash;198). Sage Publications, Inc. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.4135/9781849200479.n9\u003c/span\u003e\u003cspan address=\"10.4135/9781849200479.n9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDe Clercq, B., Rettew, D., Althoff, R. R. \u0026amp; De Bolle, M. Childhood personality types: Vulnerability and adaptation over time. \u003cem\u003eJ. Child Psychol. Psychiatry\u003c/em\u003e. \u003cb\u003e53\u003c/b\u003e (6), 716\u0026ndash;722. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1469-7610.2011.02512.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1469-7610.2011.02512.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGerlach, M., Farb, B., Revelle, W. \u0026amp; Nunes Amaral, L. A. A robust data-driven approach identifies four personality types across four large data sets. \u003cem\u003eNat. Hum. Behav.\u003c/em\u003e \u003cb\u003e2\u003c/b\u003e (10), 735\u0026ndash;742. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41562-018-0419-z\u003c/span\u003e\u003cspan address=\"10.1038/s41562-018-0419-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGhaderi, A. H., Yang, X. F., Gotlieb, R., Johnson, D. L. \u0026amp; Immordino-Yang, M. H. \u003cem\u003eIntersecting effects of social circumstances and transcendent thinking on mid-adolescents\u0026rsquo; longitudinal functional connectome development\u003c/em\u003e (p. 2025.05.20.654735). bioRxiv. (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1101/2025.05.20.654735\u003c/span\u003e\u003cspan address=\"10.1101/2025.05.20.654735\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGignac, G. E. \u0026amp; Szodorai, E. T. Effect size guidelines for individual differences researchers. \u003cem\u003ePers. Indiv. Differ.\u003c/em\u003e \u003cb\u003e102\u003c/b\u003e, 74\u0026ndash;78. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.paid.2016.06.069\u003c/span\u003e\u003cspan address=\"10.1016/j.paid.2016.06.069\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGur, R. Computerized Neurocognitive Scanning: I. Methodology and Validation in Healthy People. \u003cem\u003eNeuropsychopharmacology\u003c/em\u003e \u003cb\u003e25\u003c/b\u003e (5), 766\u0026ndash;776. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0893-133X(01)00278-0\u003c/span\u003e\u003cspan address=\"10.1016/S0893-133X(01)00278-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHAN, J. H. The Effects of Personality Traits on Subjective Well-being and Behavioral Intention Associated with Serious Leisure Experiences. \u003cem\u003eJ. Asian Finance Econ. Bus.\u003c/em\u003e \u003cb\u003e7\u003c/b\u003e (5), 167\u0026ndash;176. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.13106/JAFEB.2020.VOL7.NO5.167\u003c/span\u003e\u003cspan address=\"10.13106/JAFEB.2020.VOL7.NO5.167\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHenning, G., Hansson, I., Berg, A. I., Lindwall, M. \u0026amp; Johansson, B. The role of personality for subjective well-being in the retirement transition \u0026ndash; Comparing variable- and person-oriented models. \u003cem\u003ePers. Indiv. Differ.\u003c/em\u003e \u003cb\u003e116\u003c/b\u003e, 385\u0026ndash;392. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.paid.2017.05.017\u003c/span\u003e\u003cspan address=\"10.1016/j.paid.2017.05.017\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKadir, S. N., Goodman, D. F. M. \u0026amp; Harris, K. D. High-Dimensional Cluster Analysis with the Masked EM Algorithm. \u003cem\u003eNeural Comput.\u003c/em\u003e \u003cb\u003e26\u003c/b\u003e (11), 2379\u0026ndash;2394. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1162/NECO_a_00661\u003c/span\u003e\u003cspan address=\"10.1162/NECO_a_00661\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKerber, A., Roth, M. \u0026amp; Herzberg, P. Y. Personality types revisited\u0026ndash;a literature-informed and data-driven approach to an integration of prototypical and dimensional constructs of personality description. \u003cem\u003ePLOS ONE\u003c/em\u003e. \u003cb\u003e16\u003c/b\u003e (1), e0244849. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0244849\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0244849\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKinnunen, M. et al. Personality profiles and health: Longitudinal evidence among Finnish adults. \u003cem\u003eScand. J. Psychol.\u003c/em\u003e \u003cb\u003e53\u003c/b\u003e (6), 512\u0026ndash;522. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1467-9450.2012.00969.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1467-9450.2012.00969.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKupst, M. J. et al. Assessment of stress and self-efficacy for the NIH Toolbox for Neurological and Behavioral Function. \u003cem\u003eAnxiety Stress Coping\u003c/em\u003e. \u003cb\u003e28\u003c/b\u003e (5), 531\u0026ndash;544. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/10615806.2014.994204\u003c/span\u003e\u003cspan address=\"10.1080/10615806.2014.994204\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeikas, S. \u0026amp; Salmela-Aro, K. Personality types during transition to young adulthood: How are they related to life situation and well-being? \u003cem\u003eJ. Adolesc.\u003c/em\u003e \u003cb\u003e37\u003c/b\u003e (5), 753\u0026ndash;762. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.adolescence.2014.01.003\u003c/span\u003e\u003cspan address=\"10.1016/j.adolescence.2014.01.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, G. et al. Perceived stress, self-efficacy, and the cerebral morphometric markers in binge-drinking young adults. \u003cem\u003eNeuroImage: Clin.\u003c/em\u003e \u003cb\u003e32\u003c/b\u003e, 102866. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.nicl.2021.102866\u003c/span\u003e\u003cspan address=\"10.1016/j.nicl.2021.102866\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, X., Yao, P., Didericksen, K., Jang, J. \u0026amp; Olson, D. Five Personality Types Based on Big Five: A Latent Class Analysis. \u003cem\u003eFamily Sci. Rev.\u003c/em\u003e \u003cb\u003e22\u003c/b\u003e (2). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.26536/fsr.2018.22.02.05\u003c/span\u003e\u003cspan address=\"10.26536/fsr.2018.22.02.05\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiao, Y. et al. Could social support mediate the relationship between personality trait patterns and mental health in Chinese older adults: A nationwide cross-sectional study conducted in 2022. \u003cem\u003eBMC Public. Health\u003c/em\u003e. \u003cb\u003e25\u003c/b\u003e (1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s12889-025-22449-1\u003c/span\u003e\u003cspan address=\"10.1186/s12889-025-22449-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, W., Kohn, N. \u0026amp; Fern\u0026aacute;ndez, G. Intersubject similarity of personality is associated with intersubject similarity of brain connectivity patterns. \u003cem\u003eNeuroImage\u003c/em\u003e \u003cb\u003e186\u003c/b\u003e, 56\u0026ndash;69. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.neuroimage.2018.10.062\u003c/span\u003e\u003cspan address=\"10.1016/j.neuroimage.2018.10.062\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMammadov, S. Personality profiles of gifted adolescents and relations with life satisfaction, perceived social support, and academic achievement. \u003cem\u003eHigh. Ability Stud.\u003c/em\u003e \u003cb\u003e34\u003c/b\u003e (1), 87\u0026ndash;108. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/13598139.2022.2068509\u003c/span\u003e\u003cspan address=\"10.1080/13598139.2022.2068509\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMammadov, S., Wang, S. \u0026amp; Lu, Z. Personality types and their associations with psychological resilience, coping with stress, and life satisfaction among undergraduate students: A latent profile analysis approach. \u003cem\u003ePers. Indiv. Differ.\u003c/em\u003e \u003cb\u003e222\u003c/b\u003e, 112599. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.paid.2024.112599\u003c/span\u003e\u003cspan address=\"10.1016/j.paid.2024.112599\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMammadov, S. \u0026amp; Ward, T. J. Exploring the Relations Between Personality, Implicit Theories, and Subjective Well-Being Among High-Ability Undergraduate Students. Gifted Child Quarterly. (Sage CA: Los Angeles, CA). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/00169862221107862\u003c/span\u003e\u003cspan address=\"10.1177/00169862221107862\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMammadov, S. \u0026amp; Ward, T. J. Exploring the Relations Between Personality, Implicit Theories, and Subjective Well-Being Among High-Ability Undergraduate Students. \u003cem\u003eGifted Child. Q.\u003c/em\u003e \u003cb\u003e67\u003c/b\u003e (1), 28\u0026ndash;43. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/00169862221107862\u003c/span\u003e\u003cspan address=\"10.1177/00169862221107862\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarengo, D., Davis, K. L., Gradwohl, G. \u0026Ouml;. \u0026amp; Montag, C. A meta-analysis on individual differences in primary emotional systems and Big Five personality traits. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e (1), 7453. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41598-021-84366-8\u003c/span\u003e\u003cspan address=\"10.1038/s41598-021-84366-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcCrae, R. R. \u0026amp; Costa, P. T. Personality trait structure as a human universal. \u003cem\u003eAm. Psychol.\u003c/em\u003e \u003cb\u003e52\u003c/b\u003e (5), 509\u0026ndash;516. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1037/0003-066X.52.5.509\u003c/span\u003e\u003cspan address=\"10.1037/0003-066X.52.5.509\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1997).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcCrae, R. R. \u0026amp; Costa, P. T. A contemplated revision of the NEO Five-Factor Inventory. \u003cem\u003ePers. Indiv. Differ.\u003c/em\u003e \u003cb\u003e36\u003c/b\u003e (3), 587\u0026ndash;596. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0191-8869(03)00118-1\u003c/span\u003e\u003cspan address=\"10.1016/S0191-8869(03)00118-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2004).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMerz, E. L. \u0026amp; Roesch, S. C. A latent profile analysis of the Five Factor Model of personality: Modeling trait interactions. \u003cem\u003ePers. Indiv. Differ.\u003c/em\u003e \u003cb\u003e51\u003c/b\u003e (8), 915\u0026ndash;919. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.paid.2011.07.022\u003c/span\u003e\u003cspan address=\"10.1016/j.paid.2011.07.022\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMin, H. \u0026amp; Su, S. Examining relationships between personality profiles and organizational health outcomes. \u003cem\u003ePers. Indiv. Differ.\u003c/em\u003e \u003cb\u003e164\u003c/b\u003e, 110118. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.paid.2020.110118\u003c/span\u003e\u003cspan address=\"10.1016/j.paid.2020.110118\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMorgan, S. et al. Association of personality profiles with depressive, anxiety, and cancer-related symptoms in patients undergoing chemotherapy. \u003cem\u003ePers. Indiv. Differ.\u003c/em\u003e \u003cb\u003e117\u003c/b\u003e, 130\u0026ndash;138. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.paid.2017.05.039\u003c/span\u003e\u003cspan address=\"10.1016/j.paid.2017.05.039\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMusa, L., Ghaderi, A., Chen, Y. \u0026amp; Crawford, J. D. Spatial task instructions and global activation trends influence functional modularity in the cortical reach network. \u003cem\u003eNeuroImage\u003c/em\u003e \u003cb\u003e320\u003c/b\u003e, 121460. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.neuroimage.2025.121460\u003c/span\u003e\u003cspan address=\"10.1016/j.neuroimage.2025.121460\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNewman, M. E. J. Finding community structure in networks using the eigenvectors of matrices. \u003cem\u003ePhys. Rev. E\u003c/em\u003e. \u003cb\u003e74\u003c/b\u003e (3), 036104. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1103/PhysRevE.74.036104\u003c/span\u003e\u003cspan address=\"10.1103/PhysRevE.74.036104\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2006a).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNewman, M. E. J. Modularity and community structure in networks. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e, \u003cem\u003e103\u003c/em\u003e(23), 8577\u0026ndash;8582. (2006b). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1073/pnas.0601602103\u003c/span\u003e\u003cspan address=\"10.1073/pnas.0601602103\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNichols, T. E. \u0026amp; Holmes, A. P. Nonparametric permutation tests for functional neuroimaging: A primer with examples. \u003cem\u003eHum. Brain. Mapp.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e (1), 1\u0026ndash;25. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/hbm.1058\u003c/span\u003e\u003cspan address=\"10.1002/hbm.1058\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2002).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePerera, H. N., Granziera, H. \u0026amp; McIlveen, P. Profiles of teacher personality and relations with teacher self-efficacy, work engagement, and job satisfaction. \u003cem\u003ePers. Indiv. Differ.\u003c/em\u003e \u003cb\u003e120\u003c/b\u003e, 171\u0026ndash;178. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.paid.2017.08.034\u003c/span\u003e\u003cspan address=\"10.1016/j.paid.2017.08.034\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePilkonis, P. A. et al. Assessment of self-reported negative affect in the NIH Toolbox. \u003cem\u003ePsychiatry Res.\u003c/em\u003e \u003cb\u003e206\u003c/b\u003e (1), 88\u0026ndash;97. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.psychres.2012.09.034\u003c/span\u003e\u003cspan address=\"10.1016/j.psychres.2012.09.034\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRobins, R. W., John, O. P., Caspi, A., Moffitt, T. E. \u0026amp; Stouthamer-Loeber, M. Resilient, overcontrolled, and undercontrolled boys: Three replicable personality types. \u003cem\u003eJ. Personal. Soc. Psychol.\u003c/em\u003e \u003cb\u003e70\u003c/b\u003e (1), 157\u0026ndash;171. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1037//0022-3514.70.1.157\u003c/span\u003e\u003cspan address=\"10.1037//0022-3514.70.1.157\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1996).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRubinov, M. \u0026amp; Sporns, O. Complex network measures of brain connectivity: Uses and interpretations. \u003cem\u003eNeuroImage Comput. Models Brain\u003c/em\u003e. \u003cb\u003e52\u003c/b\u003e (3), 1059\u0026ndash;1069. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.neuroimage.2009.10.003\u003c/span\u003e\u003cspan address=\"10.1016/j.neuroimage.2009.10.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRzeszutek, M. \u0026amp; Gruszczyńska, E. Personality types and subjective well-being among people living with HIV: A latent profile analysis. \u003cem\u003eQual. Life Res.\u003c/em\u003e \u003cb\u003e29\u003c/b\u003e (1), 57\u0026ndash;67. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11136-019-02288-5\u003c/span\u003e\u003cspan address=\"10.1007/s11136-019-02288-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSalsman, J. M. et al. Assessing psychological well-being: Self-report instruments for the NIH Toolbox. \u003cem\u003eQual. Life Research: Int. J. Qual. Life Aspects Treat. Care Rehabilitation\u003c/em\u003e. \u003cb\u003e23\u003c/b\u003e (1), 205\u0026ndash;215. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11136-013-0452-3\u003c/span\u003e\u003cspan address=\"10.1007/s11136-013-0452-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSemeijn, J. H., Van Der Heijden, B., I. J., M. \u0026amp; De Beuckelaer, A. Personality Traits and Types in Relation to Career Success: An Empirical Comparison Using the Big Five. \u003cem\u003eAppl. Psychol.\u003c/em\u003e \u003cb\u003e69\u003c/b\u003e (2), 538\u0026ndash;556. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/apps.12174\u003c/span\u003e\u003cspan address=\"10.1111/apps.12174\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSpecht, J., Luhmann, M. \u0026amp; Geiser, C. On the Consistency of Personality Types Across Adulthood: Latent Profile Analyses in Two Large-Scale Panel Studies. \u003cem\u003eSSRN Electron. J.\u003c/em\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2139/ssrn.2501903\u003c/span\u003e\u003cspan address=\"10.2139/ssrn.2501903\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTalaei, N. \u0026amp; Ghaderi, A. Integration of structural brain networks is related to openness to experience: A diffusion MRI study with CSD-based tractography. \u003cem\u003eFrontiers in Neuroscience\u003c/em\u003e, \u003cem\u003e16\u003c/em\u003e. (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fnins.2022.1040799\u003c/span\u003e\u003cspan address=\"10.3389/fnins.2022.1040799\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUdayar, S., Urbanaviciute, I., Massoudi, K. \u0026amp; Rossier, J. The Role of Personality Profiles in the Longitudinal Relationship between Work\u0026ndash;Related Well\u0026ndash;Being and Life Satisfaction among Working Adults in Switzerland. \u003cem\u003eEur. J. Pers.\u003c/em\u003e \u003cb\u003e34\u003c/b\u003e (1), 77\u0026ndash;92. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/per.2225\u003c/span\u003e\u003cspan address=\"10.1002/per.2225\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Den Akker, A. L., Deković, M., Asscher, J. J., Shiner, R. L. \u0026amp; Prinzie, P. Personality types in childhood: Relations to latent trajectory classes of problem behavior and overreactive parenting across the transition into adolescence. \u003cem\u003eJ. Personal. Soc. Psychol.\u003c/em\u003e \u003cb\u003e104\u003c/b\u003e (4), 750\u0026ndash;764. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1037/a0031184\u003c/span\u003e\u003cspan address=\"10.1037/a0031184\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Essen, D. C. et al. The WU-Minn Human Connectome Project: An overview. \u003cem\u003eNeuroImage\u003c/em\u003e \u003cb\u003e80\u003c/b\u003e, 62\u0026ndash;79. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.neuroimage.2013.05.041\u003c/span\u003e\u003cspan address=\"10.1016/j.neuroimage.2013.05.041\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Essen, D. C. et al. The Human Connectome Project: A data acquisition perspective. \u003cem\u003eNeuroImage\u003c/em\u003e \u003cb\u003e62\u003c/b\u003e (4), 2222\u0026ndash;2231. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.neuroimage.2012.02.018\u003c/span\u003e\u003cspan address=\"10.1016/j.neuroimage.2012.02.018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, M. Personality profile analysis, personality-intelligence profile analysis, and the intergenerational transmission of both: Insights from Chinese evidence. \u003cem\u003eActa. Psychol.\u003c/em\u003e \u003cb\u003e255\u003c/b\u003e, 104957. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.actpsy.2025.104957\u003c/span\u003e\u003cspan address=\"10.1016/j.actpsy.2025.104957\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, M. \u0026amp; Hanges, P. J. Latent Class Procedures: Applications to Organizational Research. \u003cem\u003eOrganizational Res. Methods\u003c/em\u003e. \u003cb\u003e14\u003c/b\u003e (1), 24\u0026ndash;31. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/1094428110383988\u003c/span\u003e\u003cspan address=\"10.1177/1094428110383988\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWojciechowski, T. Latent Profile Analysis of Personality Dimensions Among Juvenile Offenders: Relevance for Predicting Offending Seriousness. \u003cem\u003eCrime Delinquency\u003c/em\u003e. \u003cb\u003e67\u003c/b\u003e (2), 212\u0026ndash;233. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/0011128720928920\u003c/span\u003e\u003cspan address=\"10.1177/0011128720928920\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYin, K., Lee, P., Sheldon, O. J., Li, C. \u0026amp; Zhao, J. Personality profiles based on the FFM: A systematic review with a person-centered approach. \u003cem\u003ePers. Indiv. Differ.\u003c/em\u003e \u003cb\u003e180\u003c/b\u003e, 110996. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.paid.2021.110996\u003c/span\u003e\u003cspan address=\"10.1016/j.paid.2021.110996\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu, Y. \u0026amp; Zhang, Y. Personality and Developmental Characteristics of Primary School Students\u0026rsquo; Personality Types. \u003cem\u003eFront. Psychol.\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e, 693329. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fpsyg.2021.693329\u003c/span\u003e\u003cspan address=\"10.3389/fpsyg.2021.693329\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003e\u003cstrong\u003eTable 1.\u0026nbsp;Mean intra-cluster correlation coefficients for overall and domain-specific item responses.\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eTrait\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eCluster 1\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eCluster 2\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eCluster 3\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eCluster 4\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eCluster 5\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eAll Questions\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.3327\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.5719\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.4626\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.5574\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.3274\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eAgreeableness\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.3099\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.2968\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.3991\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.4871\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.4417\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eConscientiousness\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.2369\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.1472\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.1820\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.1593\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.1515\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eExtraversion\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.1752\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.1805\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.1912\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.2097\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.1819\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eNeuroticism\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.1676\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.0929\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.2521\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.1968\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.2115\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eOpenness\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.4892\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.4315\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.2715\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.6293\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.3324\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Personality types, modularity, data-driven, Graph theoretical analysis, emotion","lastPublishedDoi":"10.21203/rs.3.rs-9379725/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9379725/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003ePersonality is a fundamental aspect of human behavior, shaping how individuals perceive, interact with, and adapt to their environments. Despite extensive research, establishing a generalizable taxonomy of personality types remains challenging. Conventional personality assessments typically categorize individuals based on aggregated trait scores, which compress item-level variability and overlook how people with identical trait levels may express those traits through different response patterns. However, the reducing rich item-level information into a single aggregated score can obscure meaningful subgroups within the population and restrict the capacity to identify distinct personality profiles. Addressing this limitation is therefore essential for advancing theoretical models of personality structure and improving psychological assessment.\u003c/p\u003e \u003cp\u003eHere, we leveraged 60-item NEO Five-Factor Inventory (NEO-FFI) data from the HCP dataset (N\u0026thinsp;=\u0026thinsp;1206) to investigate personality clustering using an item-pattern-based modularity (IPBM) approach. This framework emphasizes personality categorization based on response styles rather than aggregated trait values, thereby offering a more nuanced characterization of personality structure. Our graph-theoretical clustering analysis identified five distinct personality profiles. While two clusters aligned with commonly reported resilient and under-controlled types, others reflected less frequently described configurations; to the best of our knowledge, one of these clusters has not been documented in prior work.\u003c/p\u003e \u003cp\u003eWe further incorporated NIMH Toolbox assessments to examine differences across clusters in negative affect, well-being, and self-efficacy. Significant differences emerged for negative affect, well-being, and self-efficacy, demonstrating that item-level personality patterns carry meaningful implications for psychological functioning.\u003c/p\u003e \u003cp\u003eTogether, these findings highlight the value of response-pattern-based approaches in refining personality taxonomies and underscore how item-level characteristics can yield precise insights into personality and its associations with key indicators of emotional health and adaptive functioning. This framework may enhance personalized assessment strategies and inform the development of more targeted psychological interventions.\u003c/p\u003e","manuscriptTitle":"A High-Dimensional Data-Driven Modularity Analysis Reveals Five Distinct Personality Clusters with Different Psychological Profiles","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-07 19:07:09","doi":"10.21203/rs.3.rs-9379725/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"53042042583108366293628986491314841084","date":"2026-04-29T16:27:14+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-29T13:04:05+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-04-29T05:21:36+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-24T09:18:21+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-24T09:17:29+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-04-10T12:42:26+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"70f08497-cbc3-4940-88a6-417e9897e10e","owner":[],"postedDate":"May 7th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":67265255,"name":"Physical sciences/Mathematics and computing"},{"id":67265256,"name":"Biological sciences/Psychology"},{"id":67265257,"name":"Social science/Psychology"}],"tags":[],"updatedAt":"2026-05-07T19:07:09+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-07 19:07:09","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9379725","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9379725","identity":"rs-9379725","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}