Test-retest reliability of resting-state functional connectivity: A methodological investigation with implications for aggression research

Test-retest reliability of resting-state functional connectivity: A methodological investigation with implications for aggression research | 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 Test-retest reliability of resting-state functional connectivity: A methodological investigation with implications for aggression research Maria Beser-Robles, Leonor Cerdá-Alberich, Fernando Aparici-Robles, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9290839/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 12 You are reading this latest preprint version Abstract The temporal stability of resting-state functional MR connectivity (rsfMRc) is a fundamental methodological consideration for longitudinal neuroimaging research, yet systematic evaluations across clinically relevant time frames remain limited. This study examined test-retest reliability of rsfMRc over short-term (1 hour) and medium-term (1 month) intervals in 34 healthy young adult males across three sessions. Intraclass correlation coefficients (ICC) and matrix-wide reliability metrics were calculated to evaluate temporal stability across major brain networks, and associations between stable connectivity patterns and aggression measures were examined. Overall reliability showed moderate values (mean ICC = .557) with significant network-specific variation. Salience Network regions, particularly the supramarginal gyrus, and lateral parietal regions of the Default Mode Network demonstrated the highest stability (ICC > .600). Matrix-wide analyses indicated high consistency in connectivity rankings (Kendall's W = .798). Among the 23 most temporally stable regions, several showed systematic associations with both reactive and proactive aggression. These findings reveal substantial heterogeneity in rsfMRc reliability across brain networks and measurement intervals, providing empirical guidance for researchers planning longitudinal connectivity studies. Network-specific considerations are critical in study design, and identifying stable connectivity patterns offers a foundation for biomarker development, while regions with poor reliability may require alternative acquisition strategies. Health sciences/Biomarkers Health sciences/Medical research Health sciences/Neurology Biological sciences/Neuroscience aggression brain networks resting-state functional connectivity test-retest reliability Figures Figure 1 Introduction The development of reliable neuroimaging biomarkers critically depends on the temporal stability of the underlying neural measurements. Resting-state functional MR connectivity (rsfMRc) has emerged as a promising approach for identifying stable neural signatures of psychological traits and clinical conditions, yet fundamental methodological questions about its test-retest reliability remain incompletely addressed. This represents a significant gap in the methodological foundation of longitudinal connectivity research, as unreliable measurements cannot serve as valid indicators of stable individual characteristics or meaningful change over time. While psychological instruments demonstrate considerable test-retest reliability over extended periods 1 , 2 , establishing comparable stability for neuroimaging measures presents unique challenges. The temporal stability of rsfMRc measurements varies significantly across brain regions, acquisition parameters, and analytical approaches 3 , 4 . In addition, sample characteristics might affect test-retest reliability, with homogeneous samples of healthy young adults typically yielding the highest reliability estimates 3 , 5 . Previous reliability studies have reported widely varying estimates, with intraclass correlation coefficients (ICCs) ranging from .20 to .80 depending on methodological factors 6 – 8 . However, many studies have focused on analysing functional connectivity between specific regions of interest (ROIs) or limited time frames using ICC approaches, leaving important gaps in our understanding of reliability patterns across comprehensive brain networks and clinically relevant intervals 8 – 11 . These analyses typically yield low to moderate reliability values (.20-.50), indicating the need for more comprehensive analytical approaches that consider global connectivity matrix stability alongside traditional regional measures. Previous research has revealed a hierarchical pattern of reliability across brain networks: primary sensory and motor regions (auditory cortex, primary visual cortex, somatomotor areas) demonstrate higher test-retest reliability (ICC > .50), as do specific default mode network (DMN) regions, particularly the posterior cingulate cortex, precuneus, and medial prefrontal cortex 6 , 7 , 12 , 13 . In contrast, subcortical regions and networks associated with higher-order cognitive functions, including the frontoparietal control and attention networks, often show lower reliability 3 , 5 . Additionally, interhemispheric connections, particularly homotopic connections between the bilateral sensorimotor and visual areas, generally display greater stability than intrahemispheric connections 14 , 15 . The methodological implications of reliability variations extend beyond measurement precision to fundamental questions about biomarker validity. Networks with poor temporal stability may require alternative analytical approaches, longer acquisition protocols, or different measurement strategies altogether. Conversely, highly stable networks offer promising targets for longitudinal research and clinical applications. However, current literature provides limited guidance for researchers who are planning studies that depend on reliable connectivity measurements. The development of neuroimaging biomarkers for psychological traits is a critical application of reliability assessment. Aggression research exemplifies this challenge: while psychological measures of aggression demonstrate high test-retest reliability (.70-.90), establishing stable neural correlates requires the identification of connectivity patterns with comparable temporal stability 16 , 17 . Previous rsfMRc studies have identified associations between aggressive behaviours and connectivity patterns in prefrontal-limbic circuits for reactive aggression and in reward networks for proactive aggression 18 , 19 . However, these findings lack systematic evaluation of whether the underlying neural measurements demonstrate sufficient temporal stability to serve as reliable biomarkers. This represents a fundamental methodological gap: brain-behaviour correlations observed at single time points may not reflect stable trait-like associations if the neural measures themselves lack reliability. This study provides a targeted evaluation of rsfMRc reliability within a protocol which is representative of typical clinical research settings. While previous large-scale efforts (e.g., Noble et al., 2019; Zuo et al., 2014a) have established guidelines for functional connectomics, there is a need to evaluate these metrics under the specific constraints of clinical protocols (e.g., standard resolution and moderate scan lengths) to guide longitudinal studies in specific populations. Our primary aim was to provide guidelines for reliability expectations across different networks and time frames under controlled conditions. Secondary aims included demonstrating the practical application of reliability-informed analysis strategies through the examination of brain-behaviour relationships with aggression measures, serving as a proof-of-concept for biomarker development. We hypothesized that reliability would follow previously reported hierarchical patterns, with sensory-motor and DMN regions demonstrating higher stability than subcortical and attention networks 3 . We predicted that one-month reliability would be reduced when compared to same-day measurements, but that specific networks would maintain adequate reliability (ICC > .50) for longitudinal applications. Additionally, we expected that matrix-wide metrics would reveal global stability patterns that complement regional ICC analyses. Regarding the biomarker application, we hypothesized that temporally stable connectivity patterns would show systematic associations with aggression measures across all three time points, demonstrating the potential for reliability-informed biomarker development. Specifically, we expected that: (1) reactive aggression would be associated with connectivity patterns involving prefrontal regulatory regions and limbic areas 20 , 21 , and (2) proactive aggression would correlate with connectivity in reward circuits involving orbitofrontal cortex and striatal regions 22 , 23 , but only when these networks demonstrated adequate temporal stability. Results Age effects and signal quality assessment Age accounted for minimal variance in connectivity patterns across the sample (mean R² = 1.8%, SD = 2.4%, range = 0.0%–15.4%). The distribution was highly right-skewed, with 89.5% of connections showing R² 10%. Age-adjusted ICCs (M = .559, SD = .221) were nearly identical to unadjusted values (M = .575, SD = .210), with a negligible mean difference of -0.016 (95% CI: -0.018 to -0.013, d = -0.07). The correlation between unadjusted and age-adjusted ICCs was r = .993 (p < .001), demonstrating that age did not alter reliability rankings. This near-perfect correlation confirms that age correction has a negligible practical impact on ICC values (Supplementary Table S3). To ensure reliability estimates reflected neural architecture rather than artifacts, we examined multiple quality control metrics. Head motion (mean framewise displacement and scrubbed volumes) remained consistent across all three sessions (all p > .05; Supplementary Figure S1), with no significant correlations between individual motion and connectivity values of stable ROIs (r .20). Signal amplitude (M = 0.345, SD = 0.091) showed a weak, non-significant correlation with regional reliability (r = 0.197, p = 0.280), explaining only 3.9% of ICC variance (Supplementary Figure S2). Critically, the Visual Network exhibited intermediate amplitude (M = 0.262) yet low reliability (ICC = 0.494), while the Salience Network showed comparable amplitude (M = 0.356), but substantially higher reliability (ICC = 0.640). This dissociation confirms that reliability differences reflect genuine biological variability rather than technical factors. Regional reliability patterns and variance decomposition The test-retest reliability analysis revealed substantial heterogeneity across brain networks and time intervals (Table 2). Overall, the mean ICC across all networks and time points was .557, indicating moderate reliability, with regional ICCs ranging from .301 to .717, highlighting the importance of network-specific considerations in study design. Exceptional stability characterized the DMN lateral parietal regions (left: ICC = .717; right: ICC = .654) and the Salience Network supramarginal gyrus (left: ICC = .699; right: ICC = .685), with minimal degradation across time intervals. Language Network regions consistently showed good stability (ICCs = .573-.628) with low percentages of non-significant connections (10-16%). The Frontoparietal Network demonstrated moderate to good reliability (ICCs = .510-.599), with consistent patterns across intervals. Conversely, several regions showed poor reliability profiles, with >55% of their connections showing non-significant results (p > .05 for ICC ≠ 0 test): the Visual Network (ICCs = .382-.526, 19-42% non-significant), the DMN medial prefrontal cortex (ICC = .466, 29% non-significant), the Dorsal Attention Network frontal eye fields (ICCs = .301-.357, 48-61% non-significant), and the cerebellar anterior region (ICC = .480, 71% non-significant at short intervals). Variance decomposition revealed that high-reliability regions are driven by stable individual differences. ICC correlated strongly with between-subject/within-subject (B/W) variance ratio (r = .805, p .70, n = 161) showed 3.7-fold larger B/W ratios (M = 1.65, SD = 0.55) compared to poor reliability connections (ICC < .40, n = 89; M = 0.45, SD = 0.07; t(248) = 20.24, p < .001, d = 2.67). The ROI-level analysis demonstrated that high-ICC regions (e.g., the DMN lateral parietal, the Salience Network supramarginal gyrus) exhibited between-subject variance proportions of 67.4-68.4%, while low-ICC regions (e.g., frontal eye fields, visual areas) showed balanced distributions where session-to-session fluctuations accounted for nearly half of the total variance (Supplementary Table S2). Temporal dynamics and multiple baseline benefit The comparison across measurement intervals revealed important patterns for study design. Same-day reliability (baseline vs. 1-hour) generally showed the highest ICC values (mean = .484), while longer intervals showed modest decreases (baseline vs. 1-month: .475; 1-hour vs. 1-month: .493). However, the differences were smaller than anticipated, suggesting that one-month intervals may be acceptable for many applications. Critically, three-time-point ICC calculations yielded higher values (mean=.557) than any pairwise comparison, supporting recommendations for multiple baseline measurements in longitudinal studies 24 . To empirically evaluate this benefit, we averaged connectivity values from the two Day 1 sessions (Baseline and 1-hour) and calculated the ICC against one-month follow-up. This aggregation produced systematic reliability improvement: global ICC increased from 0.463 (using only first baseline) to 0.543 (using averaged baseline), representing 17.2% enhancement in temporal stability. Matrix-wide reliability assessment and univariate-multivariate dissociation Global stability metrics provided complementary perspectives on connectivity reliability. Kendall's W of .798 indicated a strong preservation of connectivity rankings across sessions, suggesting that relative connectivity strengths remain stable even when absolute values show variability. Pearson correlations between connectivity matrices averaged .579 across temporal comparisons, with similar values for all pairwise comparisons (baseline vs. 1-hour: .588; baseline vs. 1-month: .574; 1-hour vs. 1-month: .574) and the Frobenius distance and Procrustes disparity showing similarly consistent values (Procrustes: 0.050-0.058; Frobenius: 25.08-27.06). These low and consistent values across all temporal comparisons confirm that the global connectivity architecture remains stable over both short-term (1-hour) and medium-term (1-month) intervals. This revealed a critical dissociation between edge-level and pattern-level reliability: while individual connections showed moderate mean reliability (ICC = .557), matrix-wide metrics revealed substantially higher consistency in global connectivity organization (Kendall's W = .798, Pearson r = .579). This superiority of multivariate reliability reflects statistical averaging—distributed patterns aggregate signals across many connections, reducing measurement noise impact in any single edge 25,26 , paralleling the psychometric principle that multi-item scales show higher reliability than individual items through error cancellation. Stable Network Identification and brain-behaviour association Based on our reliability criteria (ICC > .50 and <55% non-significant connections), we identified 23 suitable brain regions for longitudinal research applications: 7 from the Salience Network, 4 from the Language Network, 3 from the Default Mode Network, 4 from the Frontoparietal Network, 2 from the Dorsal Attention Network, and 3 from the Sensorimotor Network. This represents 72% of all examined regions. Although exclusion of the visual cortex, cerebellar areas, and medial prefrontal cortex has important implications for studies targeting these systems, this percentage suggests that most major network connections can support longitudinal research when appropriate methodological considerations are applied. The analysis of temporally stable connectivity patterns revealed systematic associations with aggression measures (Table 3, Figure 1). Six connectivity patterns showed significant correlations (|r| ≥ .50, p-FDR < .05) with aggression dimensions across the three time points. Four patterns showed associations with both reactive and proactive aggression: negative DMN lateral parietal-sensorimotor superior correlation (reactive: r = -.554; proactive: r = -.616), sensorimotor-executive control integration (SMN Superior-FPN LPFC: r = .522-.538), and DMN-Salience connectivity patterns. Two patterns were dimension specific: DMN PCC-Salience RPFC connections showed specific associations with proactive aggression (r = -.518 to .496). To validate our reliability-informed approach, we compared the temporal stability of brain-behaviour correlations between high-ICC and low-ICC connections. High-reliability connections exhibited significantly lower temporal variability in aggression associations (mean SD of r across sessions = 0.096) compared to low-reliability connections (mean SD = 0.112; t = 2.42, p = .016, d = 0.29). Furthermore, high-ICC connections showed a 45.1% directional consistency (maintaining same correlation sign across all time points), while low-ICC connections showed only a 31.9% consistency (z = -2.10, p = .036), confirming that a low-ICC signal structure is unsuitable for stable trait-level inference. The consistency of associations across all three time points strengthens confidence that the observed relationships reflect stable traits rather than measurement artifacts, thereby supporting the validity of our reliability-informed approach. Discussion This methodological investigation provides systematic guidelines for rsfMRc reliability expectations across major brain networks and clinically relevant time frames. Our findings reveal substantial heterogeneity in temporal stability that necessitates network-specific considerations in longitudinal study designs. The overall moderate stability (mean ICC = .557) masks critical regional variations that have direct implications for methodological decision-making in connectivity research. Network-Specific Reliability and Neurobiological Foundations Exceptional stability in DMN lateral parietal regions (ICCs =.654-.717 ) and Salience Network supramarginal gyrus (ICCs =.685-.699) reflects their fundamental neurobiological roles as neural "scaffolding" for stable individual differences. Our variance decomposition analysis quantified these foundations explicitly: high reliability in these networks is driven by dominant between-subject variance (exceeding 67% of total variance; Supplementary Table S2), indicating stable 'trait-like' fingerprints twice as large as the combined effect of measurement noise and state-dependent fluctuations. The DMN lateral parietal nodes support self-referential thinking, autobiographical memory, and perspective-taking—relatively stable characteristics forming the foundation of personal identity. The Salience Network supramarginal gyrus integrates external sensory information with internal states, facilitating appropriate attentional resource allocation and behavioural regulation. The temporal consistency of these networks makes them ideal candidates for longitudinal research applications requiring high reliability. In contrast, networks showing moderate reliability—Language Network (ICCs = .573-.628) and Frontoparietal Network (ICCs = .510-.599)—demonstrated acceptable but variable stability reflecting involvement in both stable and dynamic processes. The Language Network's stability likely reflects the fundamental role of verbal mediation and cognitive control in emotional regulation, while Frontoparietal Network components capture executive control mechanisms critical for behavioural regulation 27 . These networks may benefit from methodological optimization: longer acquisition protocols (12+ minutes), multiple baseline measurements, or larger sample sizes (N>50) to achieve adequate power for detecting meaningful changes, representing an intermediate category where methodological optimization can enhance reliability. In addition, using fixed-atlas ROIs may slightly underestimate peak reliability compared to subject-specific parcellations, representing a conservative estimate of temporal stability. Networks demonstrating poor reliability—Visual Network, DMN medial prefrontal cortex (ICC = .466), and Dorsal Attention Network frontal eye fields—present methodological challenges but also reveal important neurobiological insights. Variance decomposition showed within-subject variance often exceeded 50% in these networks, confirming they are more permeable to participants’ state during scanning. While this makes them less ideal for trait-biomarker development, it confirms their role as dynamic interfaces that are highly responsive to environmental factors. This distinction is critical: low ICC here reflects a genuine biological dominance of temporal variability over stable individual differentiation, not mere measurement error. Our lower Visual Network reliability compared to some reports (e.g., Noble et al., 2017) likely stems from technical factors: a 5mm slice thickness may introduce partial volume effects in thin occipital cortex, and eyes-open acquisition (chosen to maintain alertness) increases visual processing variability compared to eyes-closed protocols. These networks require alternative approaches, including specialized preprocessing or a focus on group-level rather than individual analyses. Visual networks, particularly higher-order visual areas, show considerable plasticity in response to environmental changes and learning experiences. The unexpected finding of sensorimotor network stability across all regions represents a novel finding, suggesting that motor-related connectivity patterns may serve as stable individual "fingerprints" that contribute to behavioural regulation, which supports embodied cognition theories. This challenges previous assumptions about sensorimotor variability and suggests these regions may be underutilized in biomarker development. Temporal dynamics and multivariate reliability Modest differences between short-term (1-hour; ICC = .493) and medium-term (1-month; ICC = .475) reliability provide important methodological guidance. The improved reliability observed when incorporating all three time points (mean ICC = .557) supports recommendations for multiple baseline measurements, consistent with systematic analyses demonstrating enhanced stability with repeated measurements 24 . Our findings suggest that averaging multiple baseline sessions—even within the same day—significantly enhances measurement power by stabilizing participant-specific connectivity profiles and mitigating effect size attenuation in brain-behaviour associations. Despite moderate regional ICCs, the high Kendall's W (.798) reveals fundamental network organization principles that remain preserved over clinically relevant time frames. This dissociation between edge-level and matrix-level reliability demonstrates that connectivity configurations maintain consistency even when individual connection strengths fluctuate—validating whole-brain pattern approaches while clarifying that both univariate and multivariate reliability inform analytical strategy selection. For studies prioritizing interpretability and network-specific inference, edge-level reliability informs measurement adequacy. For studies prioritizing predictive accuracy (e.g., developing clinical biomarkers), whole-brain pattern approaches that capitalize on multivariate reliability are optimal 25,28 . Rodriguez et al. (2025) demonstrated that matrix reliability can be enhanced through approaches that de-emphasize large coactivation patterns, potentially at the expense of edge reliability—suggesting reliability optimization depends on analytical goals. Our findings provide guidelines for both analysis levels, enabling informed decisions about reliability-accuracy trade-offs. Our network-based approach serves a complementary purpose: providing interpretable guidelines for the biological foundations of reliability differences. While whole brain patterns optimize predictive accuracy, network-level characterization facilitates mechanistic understanding and experimental design. The observation that the Salience and DMN regions show high univariate reliability while the Visual and Attention networks show state-dependence offers actionable guidance for researchers selecting measurement targets—­ information that complements but does not replace multivariate approaches. Comparison with Existing Reliability Literature Our overall reliability (mean ICC = .557, SD = .202) falls within ranges reported across resting-state studies (.20–.80; Noble et al., 2019), with variation that can be attributed to methodological factors: Sample Characteristics: Our homogeneous sample (ages 19-39, 94% aged 19-29, males only) represents an ideal case for establishing upper-bound reliability estimates. The minimal age effect (mean R² = 1.8%) indicates age-related variability was negligible relative to both stable individual differences and temporal fluctuations—the 'trait' component was 30-fold larger than age-related variance, justifying our focus on stable connectivity signatures. Clinical or broader age-range samples would likely show lower reliability due to symptom fluctuations or developmental processes 30 . Scan Duration and Technical Factors: Our 9-minute acquisitions align with standard protocols but are shorter than high-reliability protocols using 12-15 minutes. Our 5mm slice thickness, while maintaining signal-to-noise ratio in clinical scanners, is relatively coarse compared to high-resolution (2mm) protocols and may hinder precise cortical delineation. These factors mean our ICCs should be interpreted as context-specific guidelines rather than universal values. Test-Retest Intervals: The mean ICC decreased from .563 (1-hour) to .531 (1-month), consistent with biological variability accumulation over time. This reflects a fundamental distinction: short-interval reliability primarily captures measurement error, while long-interval reliability incorporates genuine temporal changes. Preprocessing Strategy: Our anatomical CompCor approach is more aggressive than basic global signal regression. Advanced denoising improves reliability by 0.10-0.15 31 , with our relatively high Salience (ICC = .685) and DMN (ICC = .660) reliability likely reflecting effective physiological noise removal in regions susceptible to cardiac/respiratory artifacts. Network-Specific Patterns: The highest reliability in the Salience and lateral DMN replicates prior work 6,7 . Low Visual Network reliability likely reflects eyes-open acquisition, as visual networks show greater stability during eyes-closed rest when ongoing processing is minimized 32 . Methodological Recommendations Based on these comparisons, we provide evidence-based recommendations: Preprocessing: Advanced denoising (aCompCor, ICA-AROMA) combined with rigorous motion correction substantially improves reliability. Standardized bandpass filtering (0.008-0.09 Hz) maintains consistency. Acquisition: For high-variability networks (Visual, DAN), acquisition times ≥12-15 minutes and multiple baseline sessions improve stability. Time-of-day standardization may reduce state-dependent fluctuations. Analysis Methods: Network-specific approaches optimize reliability—individualized parcellations enhance estimates for spatially variable networks (Frontoparietal, DAN), while fixed atlases suffice for stable networks (Salience, Sensorimotor). Hybrid approaches may balance reliability and biological interpretability. Network-Specific Strategies: Networks with between-subject variance >60% (Salience, DMN, Language) are optimal for trait biomarker research. For networks with dominant within-subject variance (<50%; Visual, portions of DAN), researchers should implement methodological enhancements or explicitly model state-dependence as a neurobiological feature rather than a measurement error. Statistical Approaches: Prioritizing high-ICC connections (>0.50) for brain-behaviour analyses reduces measurement error. Incorporating reliability estimates into models (measurement error models, latent variables) provides more accurate longitudinal inferences. Brain-Behaviour Relationships in Aggression The identification of 23 brain regions with adequate temporal stability (ICC > .50, <55% non-significant connections) represents a crucial methodological decision that strengthens the validity of brain-behaviour correlations. This selection criterion reflects a fundamental principle: only temporally stable neural connections can serve as reliable indicators of enduring psychological traits. The 23 stable regions span multiple functional networks (72% of examined regions), representing a neurobiologically meaningful distribution and aligning with networks involved in social cognition, emotional regulation, and executive control—processes directly relevant to psychological trait expression. The predominance of SN regions aligns with this network's role as an integrative hub that coordinates responses between different functional networks 33 , while the inclusion of language network regions reflects the importance of verbal mediation and cognitive control in emotional regulation. This 'reliability-informed' strategy maximizes signal-to-noise ratio by pre-selecting connections where between-subject variance exceeds 60% (B/W ratios >1), reducing measurement error attenuation and allowing precise estimates of relationships between functional architecture and traits. Our analysis demonstrates the practical value of reliability-informed approaches by identifying stable connectivity patterns associated with aggression. Six connectivity patterns showed significant correlations (|r| ≥ .50, p-FDR < .05) across all three time points. Shared patterns involving sensorimotor-executive control integration (SMN-FPN; r = .522-.538) supports embodied cognition theories proposing that motor system activation underlies aggressive responses 34 . Negative associations between the DMN lateral parietal regions and sensorimotor areas (r = -.554 to -.616) may reflect disrupted self-referential processing in individuals with higher aggression levels, consistent with previous findings linking DMN dysfunction to antisocial behaviour 18 . Proactive aggression showed specific DMN-SN connectivity patterns (DMN PCC with SN regions; r = -.512 to -.518), suggesting instrumental aggression may involve altered coordination between self-referential processing and salience detection systems. This pattern aligns with theoretical models proposing that proactive aggression requires different neural coordination than reactive aggression, involving more deliberate, goal-directed processes that require cognitive control and planning abilities 19 . The absence of limbic system connections likely reflects their poor temporal stability rather than lack of functional relevance —variance decomposition confirmed that these regions show within-subject fluctuations exceeding 55%, suggesting a potential relationship with aggression may be obscured by state-dependent noise in resting-state protocols. This highlights the importance of reliability-informed analytical approaches: by focusing on the 23 regions where between-subject variance was dominant, we ensured that the reported correlations with aggression represent enduring neural traits. Contextualizing Effect Sizes and Statistical Power While our results show certain networks reach ICC values of 0.6-0.7, we moderate claims regarding immediate biomarker readiness. Detecting medium effect sizes (d = 0.5) in longitudinal interventions with ICC = 0.6 requires approximately N=45-50 for 80% power 8 . Therefore, context-specific power analyses are essential—lower reliability necessitates significantly larger cohorts to avoid Type II errors. Our behavioural effect sizes (|r| ≥ .50) must be contextualized within current brain-wide association study (BWAS) landscapes. Recent large-scale evidence suggests replicable rsfMRI-behaviour associations are often considerably smaller 35 . Given our sample size (N=34), reported correlations should be considered exploratory and potentially subject to inflation. However, our repeated-measures design and reliability-informed filtering help mitigate measurement noise that requires much larger samples in single-session cross-sectional designs. The restricted aggression range in our healthy sample (well below clinical cutoffs) may limit effect sizes and generalisability to clinical populations, though observed correlations within this restricted range are notable, as range restriction typically attenuates effects. High reliability is not only required for longitudinal tracking but also a prerequisite for cross-sectional BWAS validity—without stable estimates, correlations are attenuated by measurement error, leading to replicability challenges. Limitations and future research directions Our sample size (N=34) limits the precision of ICC estimates (confidence intervals approximately ±0.20), though findings demonstrate substantial reliability patterns are unlikely explained by sampling variability. Single-site, male-only design limits the generalisability of specific ICC values—the exclusive male sample was chosen to reduce variability from sex-related differences in connectivity patterns and hormonal fluctuations 36,37 . Our one-month maximum interval provides guidelines for clinically relevant time frames but may not capture longer-term stability patterns important for trait-based research. The correlational nature of our current brain-behaviour analysis prevents direct causal inferences about directionality, requiring complementary experimental approaches for mechanistic understanding. However, the identified high temporal stability in specific circuits provides the necessary methodological foundation for future longitudinal designs (e.g., latent change score models) that can formally test directionality. Finally, the use of a predefined group-level atlas (CONN networks.32) may not capture individual-specific functional peaks, particularly in networks with high inter-individual spatial variability such as the frontoparietal and attention systems. While this atlas is a widely validated standard, future studies could complement fixed-atlas approaches with individualized parcellations (e.g., Kong et al., 2019) to maximize reliability in high-variability networks. Methods Study Design The decision to conduct a dedicated reliability study rather than analysing existing data was motivated by several factors: (1) the need for standardized acquisition parameters optimized for reliability assessment, (2) the requirement for precise timing intervals to capture both short-term and medium-term stability, and (3) the importance of controlling environmental and procedural variables that could affect reliability estimates. This prospective design allowed us to implement optimal protocols specific to reliability assessment rather than adapting analyses from studies designed for other purposes. The specific timing intervals (1 hour and 1 month) were selected based on methodological considerations critical to our research program: the 1-hour interval allows assessment of short-term measurement stability while controlling for scanner-related variability 5 , while the 1-month interval provides insight into stability over a timeframe commonly used in intervention and clinical studies 4 . These intervals directly address the practical needs of longitudinal connectivity research in our laboratory and similar research environments. The three-time-point design enables comparison of different reliability intervals within the same participants, reducing confounding variables that affect cross-study comparisons. This approach also allows the calculation of reliability estimates that incorporate all time points, potentially providing more stable estimates than pairwise comparisons alone 24 . Participants Sample size determination for reliability studies requires different considerations than traditional hypothesis testing studies. Following recommendations for ICC estimation Haga clic o pulse aquí para escribir texto. Haga clic o pulse aquí para escribir texto. , adequate precision typically requires 30-50 participants for moderate reliability coefficients. For ICC = 0.6 with a desired precision of ±0.15 (95% CI width of 0.30), approximately 45-50 participants would be optimal. Due to resource constraints and the substantial participant burden of three scanning sessions, we recruited 34 healthy male participants. Recent methodological investigations have successfully demonstrated reliability patterns with N=20 40 , N=46 for clinical populations 10 , and N=26 for multi-session designs 41 . Even intensive longitudinal studies with 10 repeated measures have used N=30 Haga clic o pulse aquí para escribir texto. Haga clic o pulse aquí para escribir texto. , demonstrating that our three-session design with N=34 represents a reasonable methodological approach for establishing reliability guidelines. The exclusive focus on male participants was chosen to reduce variability from sex-related differences in connectivity patterns and hormonal fluctuations 36,37 , maximizing our ability to detect reliability patterns within our sample size constraints. Inclusion criteria comprised participants aged 18-29 years, free from physical or mental disorders that could affect rsfMRc acquisition or introduce systematic variability. All participants were interviewed by two independent reviewers using a structured protocol validated in previous studies. Thirty-four healthy male young adults participated in this study (see Table 1 for complete demographics). Thirty-six participants were initially recruited from the general university population through institutional announcements at Universidad de Valencia and communication channels at Hospital Universitari i Politècnic La Fe. No selection criteria related to aggression levels or personality traits were applied. Online screening questionnaires assessed eligibility based on standard exclusion criteria before MRI attendance. Two participants were excluded due to incomplete image acquisition, resulting in a final sample of 34 participants. Participants' ages ranged from 19 to 39 years (M = 23.00, SD = 4.44, Mdn = 21). All participants were right-handed, native Spanish speakers with university-level education, and all of them were interviewed by trained researchers (with extensive experience in clinical psychology) to assess their mental health. Cohen’s kappa, used to assess the inter-rater agreement between qualitative interviewers, ranged from .78 to .89 in each assessed dimension. Exclusion criteria were: history of neurological disorders, head trauma with loss of consciousness, current psychiatric medication, substance abuse or MRI contraindications. All scanning sessions were conducted between 9:00 AM and 1:00 PM to control for circadian effects on brain function 42 . Participants were instructed to avoid caffeine for at least 4 hours before scanning. Substance use was assessed via self-report questionnaires (AUDIT for alcohol; Severity of Dependence Scale for cannabis); no participants reported problematic use. The study was approved by the Clinical Research Ethics Committee of Hospital Universitari i Politècnic La Fe (2023-MAG-2671408). All participants provided written informed consent and received monetary compensation. All methods were performed in accordance with the relevant guidelines and regulations. Procedure Participants completed online questionnaires and structured interviews prior to neuroimaging sessions. In the initial session, participants complemented the interview with various health, alcohol, and/or other drugs misuse, personality, and psychopathology aspects. Following the psychological evaluation, participants were scheduled for a resting-state fMRI (rs-fMRI) session within one week. For rsfMRc acquisition, participants were instructed to remain still with eyes open and fixated on a central point, while avoiding deliberate thoughts—conditions known to optimize signal reliability 3,43 . The measurement protocol involved three scanning sessions: two sessions separated by one hour on the same day and a third session 30 days later. This design enables assessment of both short-term stability (accounting for factors like positioning differences, scanner drift) and medium-term stability (incorporating biological and state-related variability). Participants were asked to maintain consistent sleep schedules and avoid caffeine on scanning days to minimize state-related variability. Aggression Assessment Participants completed the Spanish version of the Reactive-Proactive Aggression Questionnaire (RPQ; Raine et al., 2006; Spanish adaptation: Muñoz-Rivas, 2007). The RPQ comprises 23 items assessing two dimensions: Reactive aggression (11 items; impulsive, emotion-driven responses to provocation, e.g., "Gotten angry or mad when you lost at a game") and Proactive aggression (12 items; goal-directed, instrumental aggressive behaviour; e.g., "Hurt others to win a game"). Items are rated on a 3-point scale (0=never, 1=sometimes, 2=often). From the more extensive battery, we selected this instrument because its psychometric properties and temporal stability 45 allow us to distinguish levels of propensity towards violence in forensic and non-forensic populations 46 . In our sample, RPQ scores indicated low-to-moderate aggression levels typical of healthy young adults without clinical psychopathology (Reactive: M = 6.59, SD = 2.96, range = 0-13; Proactive: M = 2.15, SD = 1.89, range = 0-9). These means are substantially below published norms for clinical populations with conduct problems or antisocial traits 16 and reflect natural variability in general population sampling rather than enrichment for high/low aggression profiles. The internal consistency was adequate (Reactive: Cronbach's α = .78; Proactive: α = .72), comparable to published psychometric properties of the Spanish RPQ 44 . The RPQ has shown excellent test-retest reliability (r > .70) in its Spanish validation 44 . fMRI data acquisition and analysis The fMRI data were acquired on a 3.0 T magnet (Achieva TX, Philips Healthcare Best, The Netherlands) using an 8-channel head coil with parallel acquisition technology (SENSE). Participants were instructed to remain still during the fMRI. The acquisition protocol consisted of a T1-weighted high spatial resolution 3D gradient echo sequence (TE =3 ms, TR=6.2 s, flip angle=10º, voxel size=1×1×1 mm 3 , and 6 min duration) and a T2*-weighted 2D EPI BOLD rsfMRc sequence (TE=35 ms, TR=2000 ms, temporal dynamics=265, pixel size=1.8×1.8 mm2, slice thickness=5 mm, and 9 min total duration). Functional and anatomical data were preprocessed using a modular preprocessing pipeline 47 , based on the CONN toolbox 48 and SPM (SPM12-Statistical Parametric Mapping, n.d.), including realignment with correction of susceptibility distortion interactions, slice-timing correction, outlier detection, direct segmentation and MNI-space normalization, and smoothing. Functional data were realigned using the SPM realign & unwarp procedure 50 , where all scans were registered to a reference image (first scan of the first session) using a least squares approach and a six-parameter (rigid body) transformation 51 and resampled using b-spline interpolation to correct for motion and magnetic susceptibility interactions. Temporal misalignment between different slices of the functional data (acquired in ascending order) was corrected following SPM slice-timing correction (STC) procedure 52 , using sinc temporal interpolation to resample each slice BOLD time series to a common mid-acquisition time. Potential outlier scans were identified using ART 53 , as acquisitions with framewise displacement above 0.9 mm or global BOLD signal changes above 5 standard deviations 54,55 and a reference BOLD image was computed for each subject by averaging all scans and excluding outliers. Functional and anatomical data were normalized into standard MNI space, segmented into grey matter, white matter, and CSF tissue classes, and resampled to 2 mm isotropic voxels following a direct normalization procedure 54,56 using the SPM unified segmentation and normalization algorithm 57,58 . Last, functional data were smoothed using spatial convolution with a Gaussian kernel of 8 mm full width at half maximum (FWHM). Functional data were denoised using a standard denoising pipeline 47 including the regression of potential confounding effects characterized by white matter time series (5 CompCor noise components), CSF time series (5 CompCor noise components), motion parameters and their first order derivatives (12 factors) 59 , outlier scans (below 28 factors) 55 , session effects and their first order derivatives (2 factors), and linear trends (2 factors) within each functional run, followed by bandpass frequency filtering of the BOLD time series 60 between 0.008 Hz and 0.09 Hz 61,62 . Noise components within white matter and CSF were estimated by computing the average BOLD signal and the largest principal components orthogonal to the BOLD average, motion parameters, and outlier scans within each subject's eroded segmentation masks. ROI-to-ROI connectivity (RRC) matrices were estimated, characterizing the functional connectivity between each pair of ROIs in all three conditions individually. ROIs were selected using the 32-networks atlas in the CONN toolbox, which is derived from a group-ICA decomposition of 497 participants from the Human Connectome Project (HCP). This atlas provides a validated representation of canonical functional networks, offering a balance between spatial specificity, data reduction, and interpretability within established neurobiological frameworks 54 . Functional connectivity strength was represented by Fisher-transformed bivariate correlation coefficients from a general linear model (weighted-GLM 47 ), estimated separately for each pair of ROIs, characterising the association between their BOLD signal time series. The analysis focused on seven major functional brain networks and their constituent regions, selected based on established roles in attention, emotion processing, and behavioural regulation, and potential relevance to the clinical variables. The DMN included the lateral parietal regions (LP, left and right), MPFC, and PCC. The Dorsal Attention Network (DAN) comprised the frontal eye fields (FEF) and intraparietal sulcus (IPS), both bilateral. The Frontoparietal Network (FPN) included the lateral prefrontal cortex (LPFC) and posterior parietal cortex (PPC), both bilateral. The Language Network (LN) consisted of the inferior frontal gyrus (IFG) and posterior superior temporal gyrus (pSTG), both bilateral. The Salience Network (SN) encompassed the anterior cingulate cortex (ACC), anterior insula (aInsula), rostral prefrontal cortex (RPFC), and supramarginal gyrus (SMG), all bilateral except for the ACC. The Sensorimotor Network (SMN) included bilateral lateral regions and a superior region, while the Visual Network (VN) comprised bilateral lateral regions, a medial region, and an occipital region. Additionally, cerebellar regions were examined, including both anterior and posterior portions. These networks and their nodes were analysed for their connectivity patterns and stability across time points. Statistical Analysis Age effects on reliability To verify that reliability estimates were not confounded by age-related variability, we calculated age-adjusted ICCs by residualizing connectivity values for linear age effects using ordinary least squares regression separately for each connection. For each connection, we regressed connectivity values onto age across subjects, calculated residuals, and added the grand mean to maintain the original scale. ICCs were then computed on these age-adjusted connectivity values. Reliability assessment Test-retest reliability was assessed using multiple complementary approaches. Intraclass correlation coefficients (ICC model 2,k) quantified consistency across the three time points, providing the primary metric for regional reliability assessment. We calculated ICCs for all time points combined and for pairwise comparisons (baseline vs. 1-hour, baseline vs. 1-month, 1-hour vs. 1-month). For each connection, statistical significance of ICC estimates was assessed using the F-test provided by Pingouin's ICC calculation 63 , which tests the null hypothesis that ICC = 0 (no temporal consistency). The F-test compares between-subject variance to within-subject variance, with degrees of freedom df1 = n_subjects - 1 = 33 and df2 = n_subjects × (n_sessions - 1) = 68. Connections showing p > .05 were classified as having non-significant temporal stability. To characterize regional reliability patterns, we calculated the percentage of non-significant connections for each ROI (i.e., the proportion of that ROI's connections failing to show significant temporal consistency). ROIs for which >55% of connections were non-significant were classified as 'highly unstable' based on the median split of this distribution across all ROIs. To complement regional ICC analyses, we calculated several matrix-wide metrics providing global perspectives on connectivity stability. Pearson correlation was used to assess the overall similarity of connectivity matrices between sessions, with higher coefficients indicating greater consistency in global network organization 6 . Kendall’s W, a non-parametric measure, was applied to evaluate the consistency of rank-order connectivity strengths across sessions, offering robust insights into global reproducibility 64 . The Frobenius distance quantified the overall differences between connectivity matrices by computing the root mean square of element-wise differences, with smaller distances indicating greater similarity in global network structure 65 . The Procrustes disparity assessed the geometric alignment of connectivity matrices, considering both scale and orientation, in which values close to zero indicated high structural similarity and preservation of global connectivity patterns over time 66 . These metrics provide information about global network organization stability that may not be captured by regional ICC values alone. Variance decomposition and quality assessment To ensure that reliability differences were not driven by regional signal properties, we quantified signal quality for all 32 ROIs. Signal amplitude was calculated as the temporal standard deviation of pre-processed low-frequency BOLD fluctuations (0.01-0.1 Hz), reflecting neuronal signal variability after removal of motion artifacts, physiological noise, and scanner drift (equivalent to Amplitude of Low-Frequency Fluctuations [ALFF]; 15,67 . All data passed CONN toolbox's automated quality control procedures prior to analysis. To determine whether high-reliability connections reflect stable individual differences or merely low measurement noise, we decomposed the total variance into between-subject and within-subject components for all 496 connections 63 . This analysis directly quantifies the relative magnitude of stable individual differences versus temporal fluctuations, contextualizing which networks are most suitable for individual-differences research 68 . For each connection, we calculated: (1) between-subject variance (σ²_B) as the variance of subject-specific means across the three sessions, representing stable individual differences, and (2) within-subject variance (σ²_W) as the average variance within subjects across sessions, representing temporal fluctuations and measurement error. The between/within (B/W) ratio (σ²_B / σ²_W) quantifies whether individual differences exceed temporal variability: ratios > 1.0 indicate that stable differences between individuals exceed fluctuations within individuals, with larger ratios reflecting more pronounced individual differentiation. Brain-behaviour reliability analysis To examine relationships between stable connectivity patterns and aggression, we employed a reliability-informed analytical strategy, using the 23 most reliable ROIs (ICC > .50 and <55% of connections showing non-significant temporal consistency; n = 253 possible connections) from the previous analysis for the proof-of-concept biomarker analysis. In addition, we ensured that retained connections reflected stable inter-individual differences rather than measurement noise or session-specific fluctuations. For each connection among these high-reliability ROIs, we calculated functional connectivity values for each subject at each of the three scanning sessions. We then used the repeated-measures correlation (rm_corr; Bakdash & Marusich, 2017) to examine associations between session-specific connectivity values and trait aggression scores (RPQ reactive and proactive subscales). The repeated-measures correlation accounts for within-subject dependencies across sessions by computing the common within-individual association between two variables measured repeatedly. A key prerequisite for the valid application of the repeated-measures correlation ( rm_corr ) is the assumption of stationarity across repeated measurements. Bakdash & Marusich (2017) emphasize that the method may be inappropriate if systematic changes or dominant within-subject fluctuations exist across time points. To ensure the integrity of our behavioural inferences, we restricted our analysis to high-reliability connections, where measured signal should primarily reflect stable inter-individual trait differences rather than session-specific noise or fluctuations. Connections showing |r| ≥ .50 (p-FDR < .05) were considered meaningful associations, following conventional guidelines for effect size interpretation in correlational research (Cohen, 1988). False discovery rate (FDR) correction via the Benjamini-Hochberg procedure controlled for multiple comparisons across all tested connections. Conclusion This investigation establishes that rsfMRc reliability varies substantially across networks and must be considered in all phases of longitudinal study design. Network-specific reliability patterns provide empirical guidance for methodological decisions, while reliability-informed analysis strategies offer frameworks for robust biomarker development. Networks with excellent reliability offer immediate opportunities for clinical applications, while those with poor reliability require innovative methodological solutions. The identification of stable neural signatures associated with aggression traits enhances our understanding of their neurobiological basis and offers promising opportunities for developing connectivity-based biomarkers and targeted interventions. The temporal stability in brain-behaviour relationships provides a methodological framework for future research seeking to establish reliable neural markers of psychological traits, representing a significant step toward precision approaches in mental health, where stable neural signatures could inform personalized intervention strategies based on individual connectivity profiles. Declarations Funding This research was partially supported by grant number PID2022-142287OA-I00; funded by MCIN/AEI/10.13039/501100011033 and by ERDF, EU” and The Prometeo Program for research groups of excellence of the Ministry of Innovation, Universities, Science and Digital Society of the Generalitat Valenciana, grant number CIPROM/2021/46. Data availability statement The raw MRI data that support the findings of this study are not publicly available due to legal and ethical restrictions imposed by the approving Clinical Research Ethics Committee. Processed functional connectivity matrices, clinical and behavioural variables, and all analysis code are publicly available in the GitHub repository at [https://github.com/beserma/test_retest]. Requests for further information regarding data access may be directed to the corresponding author. Competing interests The author(s) declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. Author contributions M.B.-R.: Conceptualization, Data curation, Formal analysis, Methodology, Software, Visualization, Writing – original draft, Writing – review & editing. Á.R.-M.: Data curation, Funding acquisition, Project administration, Supervision, Writing – review & editing. C.S.-C.: Data curation. L.C.-A.: Writing – review & editing. F.A.: Writing – review & editing. L.M.-A.: Writing – review & editing. L.M.-B.: Writing – review & editing. References Costa, P. T. & McCrae, R. R. The Five-Factor Model of Personality and Its Relevance to Personality Disorders. J. Pers. Disord . 6 , 343–359 (1992). Watson, D. Stability versus change, dependability versus error: Issues in the assessment of personality over time. J. Res. Pers. 38 , 319–350 (2004). Noble, S. et al. Influences on the Test–Retest Reliability of Functional Connectivity MRI and its Relationship with Behavioral Utility. Cereb. Cortex . 27 , 5415–5429 (2017). Elliott, M. L. et al. What Is the Test-Retest Reliability of Common Task-Functional MRI Measures? New Empirical Evidence and a Meta-Analysis. Psychol. Sci. 31 , 792–806 (2020). Shah, L. M., Cramer, J. A., Ferguson, M. A., Birn, R. M. & Anderson, J. S. Reliability and reproducibility of individual differences in functional connectivity acquired during task and resting state. Brain Behav 6 , (2016). Shehzad, Z. et al. The Resting Brain: Unconstrained yet Reliable. Cereb. Cortex . 19 , 2209–2229 (2009). Braun, U. et al. Test–retest reliability of resting-state connectivity network characteristics using fMRI and graph theoretical measures. Neuroimage 59 , 1404–1412 (2012). Noble, S., Scheinost, D. & Constable, R. T. A decade of test-retest reliability of functional connectivity: A systematic review and meta-analysis. Neuroimage 203 , 116157 (2019). Zuo, X. N. & Xing, X. X. Test-retest reliabilities of resting-state FMRI measurements in human brain functional connectomics: A systems neuroscience perspective. Neurosci. Biobehav Rev. 45 , 100–118 (2014). Anderson, J. S., Ferguson, M. A. & Lopez-Larson, M. Yurgelun-Todd, D. Reproducibility of Single-Subject Functional Connectivity Measurements. Am. J. Neuroradiol. 32 , 548–555 (2011). Birn, R. M. et al. The effect of scan length on the reliability of resting-state fMRI connectivity estimates. Neuroimage 83 , 550–558 (2013). Wang, J. H. et al. Graph Theoretical Analysis of Functional Brain Networks: Test-Retest Evaluation on Short- and Long-Term Resting-State Functional MRI Data. PLoS One . 6 , e21976 (2011). Blautzik, J. et al. Classifying fMRI-derived resting-state connectivity patterns according to their daily rhythmicity. Neuroimage 71 , 298–306 (2013). Chou, Y., -h., Panych, L. P., Dickey, C. C., Petrella, J. R. & Chen, N. -k. Investigation of Long-Term Reproducibility of Intrinsic Connectivity Network Mapping: A Resting-State fMRI Study. Am. J. Neuroradiol. 33 , 833–838 (2012). Zuo, X. N. et al. Reliable intrinsic connectivity networks: Test–retest evaluation using ICA and dual regression approach. Neuroimage 49 , 2163–2177 (2010). Raine, A. et al. The reactive–proactive aggression questionnaire: differential correlates of reactive and proactive aggression in adolescent boys. Aggress. Behav. 32 , 159–171 (2006). Giromini, L. et al. Validity and Reliability of the Interpersonal Competence Questionnaire: Empirical Evidence from an Italian Study. J. Psychopathol. Behav. Assess. 38 , 113–123 (2016). Romero-Martínez, Á., Sarrate-Costa, C. & Moya-Albiol, L. Reactive vs proactive aggression: A differential psychobiological profile? Conclusions derived from a systematic review. Neurosci. Biobehav Rev. 136 , 104626 (2022). Romero-Martínez, Á. et al. The Brain Resting-State Functional Connectivity Underlying Violence Proneness: Is It a Reliable Marker for Neurocriminology? A Systematic Review. Behav. Sci. 9 , 11 (2019). Davidson, R. J., Putnam, K. M. & Larson, C. L. Dysfunction in the Neural Circuitry of Emotion Regulation–A Possible Prelude to Violence. Science ( ). 289, 591–594 (2000).). 289, 591–594 (2000). (1979). Fulwiler, C. E., King, J. A. & Zhang, N. Amygdala–orbitofrontal resting-state functional connectivity is associated with trait anger. Neuroreport 23 , 606–610 (2012). Blair, R. J. R. The Neurobiology of Impulsive Aggression. J. Child. Adolesc. Psychopharmacol. 26 , 4–9 (2016). Rosell, D. R. & Siever, L. J. The neurobiology of aggression and violence. CNS Spectr. 20 , 254–279 (2015). Pannunzi, M. et al. Resting-state fMRI correlations: From link-wise unreliability to whole brain stability. Neuroimage 157 , 250–262 (2017). Taxali, A., Angstadt, M., Rutherford, S. & Sripada, C. Boost in Test–Retest Reliability in Resting State fMRI with Predictive Modeling. Cereb. Cortex . 31 , 2822–2833 (2021). Camp, C. C., Noble, S., Scheinost, D., Stringaris, A. & Nielson, D. M. Test-Retest Reliability of Functional Connectivity in Adolescents With Depression. Biol. Psychiatry Cogn. Neurosci. Neuroimaging . 9 , 21–29 (2024). Bressler, S. L. & Menon, V. Large-scale brain networks in cognition: emerging methods and principles. Trends Cogn. Sci. 14 , 277–290 (2010). Noble, S., Curtiss, J., Pessoa, L. & Scheinost, D. The tip of the iceberg: A call to embrace anti-localizationism in human neuroscience research. Imaging Neuroscience 2 , (2024). Rodriguez, R. X., Noble, S., Camp, C. C. & Scheinost, D. Connectome caricatures remove large-amplitude coactivation patterns in resting-state fMRI to emphasize individual differences. Nat. Neurosci. 28 , 2601–2611 (2025). Zuo, X. N. et al. An open science resource for establishing reliability and reproducibility in functional connectomics. Sci. Data . 1 , 140049 (2014). Parkes, L., Fulcher, B., Yücel, M. & Fornito, A. An evaluation of the efficacy, reliability, and sensitivity of motion correction strategies for resting-state functional MRI. Neuroimage 171 , 415–436 (2018). Patriat, R. et al. The effect of resting condition on resting-state fMRI reliability and consistency: A comparison between resting with eyes open, closed, and fixated. Neuroimage 78 , 463–473 (2013). Menon, V. Large-scale brain networks and psychopathology: a unifying triple network model. Trends Cogn. Sci. 15 , 483–506 (2011). Yankouskaya, A. & Sui, J. Self-prioritization is supported by interactions between large‐scale brain networks. Eur. J. Neurosci. 55 , 1244–1261 (2022). Marek, S. et al. Reproducible brain-wide association studies require thousands of individuals. Nature 603 , 654–660 (2022). Engman, J., Linnman, C., Van Dijk, K. R. A. & Milad, M. R. Amygdala subnuclei resting-state functional connectivity sex and estrogen differences. Psychoneuroendocrinology 63 , 34–42 (2016). Ritchie, S. J. et al. Sex Differences in the Adult Human Brain: Evidence from 5216 UK Biobank Participants. Cereb. Cortex . 28 , 2959–2975 (2018). Kong, R. et al. Spatial Topography of Individual-Specific Cortical Networks Predicts Human Cognition, Personality, and Emotion. Cereb. Cortex . 29 , 2533–2551 (2019). Koo, T. K. & Li, M. Y. A Guideline of Selecting and Reporting Intraclass Correlation Coefficients for Reliability Research. J. Chiropr. Med. 15 , 155–163 (2016). Guo, C. C. et al. One-year test–retest reliability of intrinsic connectivity network fMRI in older adults. Neuroimage 61 , 1471–1483 (2012). Sheu, L. K., Jennings, J. R. & Gianaros, P. J. Test–retest reliability of an fMRI paradigm for studies of cardiovascular reactivity. Psychophysiology 49 , 873–884 (2012). Jiang, C. et al. Diurnal Variations in Neural Activity of Healthy Human Brain Decoded with Resting-State Blood Oxygen Level Dependent fMRI. Front. Hum. Neurosci. 10 , 634 (2016). Zou, Q. et al. Reliability comparison of spontaneous brain activities between BOLD and CBF contrasts in eyes-open and eyes-closed resting states. Neuroimage 121 , 91–105 (2015). Muñoz-Rivas, M. J. M. D. P. Validación de la versión modificada de la Conflicts Tactics Scale (M-CTS) en población juvenil española. Psicothema 19 , 693–698 (2007). McAuliffe, M. D., Hubbard, J. A., Rubin, R. M., Morrow, M. T. & Dearing, K. F. Reactive and Proactive Aggression: Stability of Constructs and Relations to Correlates. J. Genet. Psychol. 167 , 365–382 (2007). Brugman, S. et al. Examining the reactive proactive questionnaire in adults in forensic and non-forensic settings: A variable‐ and person‐based approach. Aggress. Behav. 43 , 155–162 (2017). Nieto-Castanon, A. Handbook of Functional Connectivity Magnetic Resonance Imaging Methods in CONN (Hilbert, 2020). 10.56441/hilbertpress.2207.6598 Whitfield-Gabrieli, S., Nieto-Castanon, A. & Conn A Functional Connectivity Toolbox for Correlated and Anticorrelated Brain Networks. Brain Connect. 2 , 125–141 (2012). SPM12 -Statistical Parametric Mapping. https://www.fil.ion.ucl.ac.uk/spm/software/spm12/ Andersson, J. L. R., Hutton, C., Ashburner, J., Turner, R. & Friston, K. Modeling Geometric Deformations in EPI Time Series. Neuroimage 13 , 903–919 (2001). Friston, K. J. et al. Spatial registration and normalization of images. Hum. Brain Mapp. 3 , 165–189 (1995). Sladky, R. et al. Slice-timing effects and their correction in functional MRI. Neuroimage 58 , 588–594 (2011). Whitfield-Gabrieli, S., Nieto-Castanon, A. & Ghosh, S. Artifact detection tools (ART). Cambridge, MA : (2011). Nieto-Castanon, A. Preparing fMRI Data for Statistical Analysis. (2022). Power, J. D. et al. Methods to detect, characterize, and remove motion artifact in resting state fMRI. Neuroimage 84 , 320–341 (2014). Calhoun, V. D. et al. The impact of T1 versus EPI spatial normalization templates for fMRI data analyses. Hum. Brain Mapp. 38 , 5331–5342 (2017). Ashburner, J. A fast diffeomorphic image registration algorithm. Neuroimage 38 , 95–113 (2007). Ashburner, J. & Friston, K. J. Unified segmentation. Neuroimage 26 , 839–851 (2005). Friston, K. J., Williams, S., Howard, R., Frackowiak, R. S. J. & Turner, R. Movement-Related effects in fMRI time‐series. Magn. Reson. Med. 35 , 346–355 (1996). Hallquist, M. N., Hwang, K. & Luna, B. The nuisance of nuisance regression: Spectral misspecification in a common approach to resting-state fMRI preprocessing reintroduces noise and obscures functional connectivity. Neuroimage 82 , 208–225 (2013). Chai, X. J., Castañón, A. N., Öngür, D. & Whitfield-Gabrieli, S. Anticorrelations in resting state networks without global signal regression. Neuroimage 59 , 1420–1428 (2012). Behzadi, Y., Restom, K., Liau, J. & Liu, T. T. A component based noise correction method (CompCor) for BOLD and perfusion based fMRI. Neuroimage 37 , 90–101 (2007). Shrout, P. E. & Fleiss, J. L. Intraclass correlations: Uses in assessing rater reliability. Psychol. Bull. 86 , 420–428 (1979). Legendre, P. Species associations: the Kendall coefficient of concordance revisited. J. Agric. Biol. Environ. Stat. 10 , 226–245 (2005). Loan, C. F. V. The ubiquitous Kronecker product. J. Comput. Appl. Math. 123 , 85–100 (2000). Gower, J. C. & Dijksterhuis, G. B. Procrustes Problems (Oxford University Press, 2004). 10.1093/acprof:oso/9780198510581.001.0001 Yang, H. et al. Amplitude of low frequency fluctuation within visual areas revealed by resting-state functional MRI. Neuroimage 36 , 144–152 (2007). Gratton, C. et al. Functional Brain Networks Are Dominated by Stable Group and Individual Factors, Not Cognitive or Daily Variation. Neuron 98 , 439–452e5 (2018). Bakdash, J. Z. & Marusich, L. R. Repeated Measures Correlation. Front Psychol 8 , (2017). Tables Table 1. Participant Demographic and Clinical Characteristics Variable n = 34 Age (years) Mean (SD) 23.00 (4.44) Median 21 Range 19-39 Sex Male 34 (100%) Marital Status Single 34 (100%) Education Level University 34 (100%) RPQ Reactive Mean (SD) 6.59 (2.96) Median 7.0 Range 0-13 RPQ Proactive Mean (SD) 2.15 (1.89) Median 1.5 Range 0-9 Note: All participants were male university students recruited from the general population. RPQ = Reactive-Proactive Aggression Questionnaire. Age outliers defined as >1.5 IQR above Q3 (>30 years). Main analyses reported for n=34; sensitivity analyses excluding age outliers (n=32) showed highly similar reliability patterns (see Supplementary Results). Table 2. Test-retest reliability of functional connectivity across brain networks. ROIs All time points Baseline vs. 1 Hour Baseline vs. 1 Month 1 Hour vs. 1 Month Network Region Mean ICC % Non-significant Mean ICC % Non-significant Mean ICC % Non-significant Mean ICC % Non-significant Cerebellar Anterior .480 10% .348 71%* .320 84%* .501 26% Posterior .527 13% .444 39% .404 55% .488 48% DMN LP (L) .717 6% .635 19% .617 16% .649 10% LP (R) .654 10% .601 19% .557 29% .609 13% MPFC .466 29% .335 68%* .395 61%* .455 55% PCC .531 12% .512 32% .466 52%* .440 55% DAN FEF (L) .357 48% .346 81%* .373 61%* .305 74%* FEF (R) .301 61%* .296 77%* .311 77%* .317 77%* IPS (L) .601 6% .547 26% .447 35% .517 26% IPS (R) .615 3% .570 19% .504 32% .504 29% FPN LPFC (L) .510 23% .435 48% .427 42% .443 45% LPFC (R) .569 13% .464 42% .470 42% .513 32% PPC (L) .599 13% .507 35% .516 42% .518 35% PPC (R) .589 16% .487 42% .515 26% .498 26% LN IFG (L) .578 16% .477 45% .566 26% .525 39% IFG (R) .628 16% .528 29% .567 23% .542 23% pSTG (L) .624 14% .553 26% .569 23% .512 35% pSTG (R) .573 10% .506 35% .526 32% .535 32% SN ACC .605 10% .511 29% .498 35% .531 35% aInsula (L) .601 23% .479 39% .467 42% .592 26% aInsula (R) .601 20% .486 35% .504 39% .565 29% RPFC (L) .590 10% .564 23% .476 35% .490 32% RPFC (R) .636 10% .592 19% .558 32% .549 29% SMG (L) .699 13% .616 19% .643 13% .657 10% SMG (R) .685 10% .637 16% .568 23% .615 19% SMN Lateral (L) .499 29% .403 55% .440 45% .398 52% Lateral (R) .561 10% .441 55% .459 45% .473 42% Superior .537 19% .383 55% .497 35% .475 26% VN Lateral (L) .526 19% .491 42% .398 65%* .448 52%* Lateral (R) .471 29% .462 42% .389 58%* .382 58%* Medial .505 19% .431 45% .400 61%* .431 52%* Occipital .382 42%* .412* 55% .364 61%* .312 84%* Total ROIs .557 .484 .475 .493 Note. Mean intraclass correlation coefficients (ICC) and percentage of non-significant connections for each functional network region across different temporal comparisons. Asterisks (*) show regions with more than 55% non-significant connections or | r| < 0.5, classified as highly unstable. Abbreviations: DMN = Default Mode Network; DAN = Dorsal Attention Network; FPN = Frontoparietal Network; LN = Language Network; SN = Salience Network; SMN = Sensorimotor Network; VN = Visual Network; LP = Lateral Parietal; MPFC = Medial Prefrontal Cortex; PCC = Posterior Cingulate Cortex; FEF = Frontal Eye Fields; IPS = Intraparietal Sulcus; LPFC = Lateral Prefrontal Cortex; PPC = Posterior Parietal Cortex; IFG = Inferior Frontal Gyrus; pSTG = posterior Superior Temporal Gyrus; ACC = Anterior Cingulate Cortex; RPFC = Rostral Prefrontal Cortex; SMG = Supramarginal Gyrus; L = Left; R = Right. Table 3. Stable connectivity patterns associated with aggression dimensions (RPQ proactive and reactive). Region 1 Region 2 PROACTIVE REACTIVE DMN LP (L) SN RPFC (L) -0.441* -0.467* DMN LP (R) FPN LPFC (L) -0.467* -0.212 SMN Superior -0.616* -0.554* DMN PCC SN RPFC (L) -0.518* -0.072 SN RPFC (R) 0.496* 0.454* SN SMG (R) -0.512* -0.396 DAN IPS (L) FPN LPFC (L) -0.127* -0.053 LN IFG (L) -0.467* -0.231* LN pSTG (R) 0.186* 0.345 FPN LPFC (L) 0.391* 0.040 LN IFG (L) 0.227* 0.493* LN pSTG (R) -0.435* -0.344 FPN LPFC (L) LN pSTG (R) -0.505* -0.502* FPN PPC (R) LN pSTG (L) -0.525* -0.545* SN SMG (R) DAN IPS (L) 0.498* 0.471* SMN Lateral (L) SN ACC 0.383* 0.459* SN RPFC (L) -0.292* -0.440* SMN Lateral (R) DAN IPS (L) -0.410* -0.445* SMN Superior FPN LPFC (L) 0.538* 0.522* LN pSTG (L) 0.501* 0.407 Note: Values represent Pearson correlation coefficients. * Indicates statistically significant correlations (p-FDR < 0.005) after false discovery rate correction for multiple comparisons (253 comparisons). Bold values indicate both statistical significance (p-FDR 0.5). Region 1 labels are shown only at the beginning of each grouping. DMN = Default Mode Network; DAN = Dorsal Attention Network; FPN = Frontoparietal Network; LN = Language Network; SN = Salience Network; SMN = Sensorimotor Network; LP = Lateral Parietal; PCC = Posterior Cingulate Cortex; IPS = Intraparietal Sulcus; LPFC = Lateral Prefrontal Cortex; PPC = Posterior Parietal Cortex; IFG = Inferior Frontal Gyrus; pSTG = Posterior Superior Temporal Gyrus; ACC = Anterior Cingulate Cortex; AIns = Anterior Insula; RPFC = Rostral Prefrontal Cortex; SMG = Supramarginal Gyrus; L = Left; R = Right. Additional Declarations No competing interests reported. Supplementary Files SupplementarymaterialFigureS2.docx SupplementaryMaterialFigureS3.docx SupplementarymaterialTable1.docx SupplementarymaterialTableS2.docx SupplementarymaterialFigureS1.docx SupplementaryMaterialTableS3.docx SupplementarymaterialFigureS1.docx SupplementaryMaterialTableS3.docx SupplementarymaterialTable1.docx SupplementarymaterialFigureS2.docx SupplementaryMaterialFigureS3.docx SupplementarymaterialTableS2.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 28 Apr, 2026 Reviews received at journal 22 Apr, 2026 Reviews received at journal 19 Apr, 2026 Reviewers agreed at journal 15 Apr, 2026 Reviewers agreed at journal 12 Apr, 2026 Reviewers agreed at journal 10 Apr, 2026 Reviewers agreed at journal 10 Apr, 2026 Reviewers invited by journal 10 Apr, 2026 Editor assigned by journal 10 Apr, 2026 Editor invited by journal 08 Apr, 2026 Submission checks completed at journal 07 Apr, 2026 First submitted to journal 07 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-9290839","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":625722137,"identity":"ba5e2465-e84e-4690-8e76-c886e4007acc","order_by":0,"name":"Maria Beser-Robles","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA5klEQVRIiWNgGAWjYBACAxiDjYGx8QGQ5uEjQQtzM4jDw0a0FgYG9jYJiF4CwJy9/eHjyhy7fD72g22VX3PsZIDWPXx0A48Wy54zxoZntyVbtvEktt2W3ZYMdBibsXEOPofdyGGTbNzGbMDGANQiuY0ZqIWHTRqvlvvPn/9s3FZvwMb/sK1Ycls9EVpuMJgxNm47bMAmkdjG+HHbYcJaLHtyjIEOOw7U8rBZmnHbcR42ZgJ+MWc//vBj47ZqA/n+9Icff26rtudnb374GJ8WFMDMAyaJVQ4CjD9IUT0KRsEoGAUjBgAAIR1DXZCr7EQAAAAASUVORK5CYII=","orcid":"","institution":"Hospital Universitari i Politècnic La Fe","correspondingAuthor":true,"prefix":"","firstName":"Maria","middleName":"","lastName":"Beser-Robles","suffix":""},{"id":625722138,"identity":"e321b09a-382d-4dd3-a064-72de04716d75","order_by":1,"name":"Leonor Cerdá-Alberich","email":"","orcid":"","institution":"Hospital Universitari i Politècnic La Fe","correspondingAuthor":false,"prefix":"","firstName":"Leonor","middleName":"","lastName":"Cerdá-Alberich","suffix":""},{"id":625722139,"identity":"2beea933-0a9d-4ca1-801f-0a8f4fad4994","order_by":2,"name":"Fernando Aparici-Robles","email":"","orcid":"","institution":"Hospital Universitari i Politècnic La Fe","correspondingAuthor":false,"prefix":"","firstName":"Fernando","middleName":"","lastName":"Aparici-Robles","suffix":""},{"id":625722140,"identity":"f8090c33-2ad9-4f53-981d-cb8471765295","order_by":3,"name":"Carolina Sarrate-Costa","email":"","orcid":"","institution":"University of Valencia","correspondingAuthor":false,"prefix":"","firstName":"Carolina","middleName":"","lastName":"Sarrate-Costa","suffix":""},{"id":625722141,"identity":"9e581d0b-6b71-4b83-8d9f-82b398ebf3b8","order_by":4,"name":"Luis Moya-Albiol","email":"","orcid":"","institution":"University of Valencia","correspondingAuthor":false,"prefix":"","firstName":"Luis","middleName":"","lastName":"Moya-Albiol","suffix":""},{"id":625722142,"identity":"d638fec1-8c63-47bf-add2-4826cd650626","order_by":5,"name":"Luis Martí-Bonmatí","email":"","orcid":"","institution":"Hospital Universitari i Politècnic La Fe","correspondingAuthor":false,"prefix":"","firstName":"Luis","middleName":"","lastName":"Martí-Bonmatí","suffix":""},{"id":625722143,"identity":"d81b3c2e-d6e5-4f4d-bf1d-4ab276228d92","order_by":6,"name":"Ángel Romero-Martínez","email":"","orcid":"","institution":"University of Valencia","correspondingAuthor":false,"prefix":"","firstName":"Ángel","middleName":"","lastName":"Romero-Martínez","suffix":""}],"badges":[],"createdAt":"2026-04-01 11:00:56","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9290839/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9290839/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107349386,"identity":"3b22bc4e-d764-4436-9d89-459290764f23","added_by":"auto","created_at":"2026-04-20 15:49:13","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":713004,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSignificant brain network correlations for RPQ Proactive and Reactive subscales. \u003c/strong\u003e\u003cem\u003eNote. Network visualization shows statistically significant correlations (p-FDR \u0026lt; 0.005) between brain regions for RPQ Reactive (upper panel) and RPQ Proactive (lower panel) scores. Node representation: Circles represent brain regions, color-coded by functional network: DMN (red), LN (blue), DAN (yellow), SN (green), FPN (purple), SMN (gray). Node positioning is optimized using force-directed layout algorithm to minimize edge crossings while grouping functionally related regions. Edge representation: Lines connecting nodes indicate significant correlations. Blue edges represent positive correlations; red edges represent negative correlations. Edge thickness reflects correlation magnitude (thicker = stronger |r|). Only correlations surviving FDR correction (q \u0026lt; 0.05) are displayed. Complete correlation matrix available in Table 3 and Supplementary Table S1. Note. Network visualization shows statistically significant correlations (p-FDR \u0026lt; 0.005) between brain regions for RPQ Reactive (upper panel) and RPQ Proactive (lower panel) scores. Node representation: Circles represent brain regions, color-coded by functional network: DMN (red), LN (blue), DAN (yellow), SN (green), FPN (purple), SMN (gray). Node positioning is optimized using force-directed layout algorithm\u003c/em\u003e\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-9290839/v1/ced3aed13246cca04a330ead.png"},{"id":107490071,"identity":"e342f807-3c1b-4a6e-8d94-a2cffcb18e19","added_by":"auto","created_at":"2026-04-22 02:49:56","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1822908,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9290839/v1/60678afd-9bc6-406c-aa79-89fbd6c8484e.pdf"},{"id":107349385,"identity":"4a6b77dd-1860-4a58-b56b-930a9937fac0","added_by":"auto","created_at":"2026-04-20 15:49:13","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":168537,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementarymaterialFigureS2.docx","url":"https://assets-eu.researchsquare.com/files/rs-9290839/v1/e27a5fbfc3cfa36657a2a617.docx"},{"id":107349388,"identity":"6d0459e0-f12e-411f-b25e-ebd90d6cd16e","added_by":"auto","created_at":"2026-04-20 15:49:13","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":168026,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterialFigureS3.docx","url":"https://assets-eu.researchsquare.com/files/rs-9290839/v1/0425e9f22c040e3e544e3c03.docx"},{"id":107487253,"identity":"414dfdd9-ec70-47b9-b6a7-6f7eb5b26079","added_by":"auto","created_at":"2026-04-22 02:40:13","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":34312,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementarymaterialTable1.docx","url":"https://assets-eu.researchsquare.com/files/rs-9290839/v1/01f06f6ee6c63de0ad5627a0.docx"},{"id":107485967,"identity":"fd14bfd2-719e-44d4-b46b-c545fdf3c321","added_by":"auto","created_at":"2026-04-22 02:37:02","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":15203,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementarymaterialTableS2.docx","url":"https://assets-eu.researchsquare.com/files/rs-9290839/v1/f80070d20f2b67b4529b595d.docx"},{"id":107487254,"identity":"d9492de6-e5e2-476a-ac1f-962caab8395d","added_by":"auto","created_at":"2026-04-22 02:40:14","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":264478,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementarymaterialFigureS1.docx","url":"https://assets-eu.researchsquare.com/files/rs-9290839/v1/0f308f579567f15a5a44a18b.docx"},{"id":107349391,"identity":"044537b9-0545-44a2-a876-983121d0e627","added_by":"auto","created_at":"2026-04-20 15:49:13","extension":"docx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":16427,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterialTableS3.docx","url":"https://assets-eu.researchsquare.com/files/rs-9290839/v1/759df2b3158cf0c5a44e1d44.docx"},{"id":107489469,"identity":"b35f5304-956b-4446-a692-05ab3b353ac8","added_by":"auto","created_at":"2026-04-22 02:47:46","extension":"docx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":264478,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementarymaterialFigureS1.docx","url":"https://assets-eu.researchsquare.com/files/rs-9290839/v1/4160236584f41a358cfd8878.docx"},{"id":107349393,"identity":"73068593-60c7-4400-b47d-423532419fb8","added_by":"auto","created_at":"2026-04-20 15:49:13","extension":"docx","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":16427,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterialTableS3.docx","url":"https://assets-eu.researchsquare.com/files/rs-9290839/v1/cada937bfe73c9778f95a54f.docx"},{"id":107486061,"identity":"f311e56a-db77-4e7f-97d4-5d2c1e1637f5","added_by":"auto","created_at":"2026-04-22 02:37:20","extension":"docx","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":34312,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementarymaterialTable1.docx","url":"https://assets-eu.researchsquare.com/files/rs-9290839/v1/5fe93ae821ca8d85955053a7.docx"},{"id":107484654,"identity":"c05222c1-cca0-4650-bb9e-62a5fb0d72e3","added_by":"auto","created_at":"2026-04-22 02:32:37","extension":"docx","order_by":9,"title":"","display":"","copyAsset":false,"role":"supplement","size":168537,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementarymaterialFigureS2.docx","url":"https://assets-eu.researchsquare.com/files/rs-9290839/v1/2d4e817f58b91bb2ee280ede.docx"},{"id":107486068,"identity":"32a7f6d3-3819-4e86-bb36-3c516f807de9","added_by":"auto","created_at":"2026-04-22 02:37:21","extension":"docx","order_by":10,"title":"","display":"","copyAsset":false,"role":"supplement","size":168026,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterialFigureS3.docx","url":"https://assets-eu.researchsquare.com/files/rs-9290839/v1/0e14104b6f328853dc81693d.docx"},{"id":107349397,"identity":"ba2a4dd7-e367-4c1f-a544-fb98b5e0f929","added_by":"auto","created_at":"2026-04-20 15:49:13","extension":"docx","order_by":11,"title":"","display":"","copyAsset":false,"role":"supplement","size":15203,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementarymaterialTableS2.docx","url":"https://assets-eu.researchsquare.com/files/rs-9290839/v1/72719894fb60bec0d63cd65e.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Test-retest reliability of resting-state functional connectivity: A methodological investigation with implications for aggression research","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe development of reliable neuroimaging biomarkers critically depends on the temporal stability of the underlying neural measurements. Resting-state functional MR connectivity (rsfMRc) has emerged as a promising approach for identifying stable neural signatures of psychological traits and clinical conditions, yet fundamental methodological questions about its test-retest reliability remain incompletely addressed. This represents a significant gap in the methodological foundation of longitudinal connectivity research, as unreliable measurements cannot serve as valid indicators of stable individual characteristics or meaningful change over time.\u003c/p\u003e \u003cp\u003eWhile psychological instruments demonstrate considerable test-retest reliability over extended periods \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, establishing comparable stability for neuroimaging measures presents unique challenges. The temporal stability of rsfMRc measurements varies significantly across brain regions, acquisition parameters, and analytical approaches \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. In addition, sample characteristics might affect test-retest reliability, with homogeneous samples of healthy young adults typically yielding the highest reliability estimates \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Previous reliability studies have reported widely varying estimates, with intraclass correlation coefficients (ICCs) ranging from .20 to .80 depending on methodological factors \u003csup\u003e\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHowever, many studies have focused on analysing functional connectivity between specific regions of interest (ROIs) or limited time frames using ICC approaches, leaving important gaps in our understanding of reliability patterns across comprehensive brain networks and clinically relevant intervals \u003csup\u003e\u003cspan additionalcitationids=\"CR9 CR10\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. These analyses typically yield low to moderate reliability values (.20-.50), indicating the need for more comprehensive analytical approaches that consider global connectivity matrix stability alongside traditional regional measures. Previous research has revealed a hierarchical pattern of reliability across brain networks: primary sensory and motor regions (auditory cortex, primary visual cortex, somatomotor areas) demonstrate higher test-retest reliability (ICC \u0026gt; .50), as do specific default mode network (DMN) regions, particularly the posterior cingulate cortex, precuneus, and medial prefrontal cortex \u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. In contrast, subcortical regions and networks associated with higher-order cognitive functions, including the frontoparietal control and attention networks, often show lower reliability \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Additionally, interhemispheric connections, particularly homotopic connections between the bilateral sensorimotor and visual areas, generally display greater stability than intrahemispheric connections \u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe methodological implications of reliability variations extend beyond measurement precision to fundamental questions about biomarker validity. Networks with poor temporal stability may require alternative analytical approaches, longer acquisition protocols, or different measurement strategies altogether. Conversely, highly stable networks offer promising targets for longitudinal research and clinical applications. However, current literature provides limited guidance for researchers who are planning studies that depend on reliable connectivity measurements.\u003c/p\u003e \u003cp\u003eThe development of neuroimaging biomarkers for psychological traits is a critical application of reliability assessment. Aggression research exemplifies this challenge: while psychological measures of aggression demonstrate high test-retest reliability (.70-.90), establishing stable neural correlates requires the identification of connectivity patterns with comparable temporal stability \u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Previous rsfMRc studies have identified associations between aggressive behaviours and connectivity patterns in prefrontal-limbic circuits for reactive aggression and in reward networks for proactive aggression \u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. However, these findings lack systematic evaluation of whether the underlying neural measurements demonstrate sufficient temporal stability to serve as reliable biomarkers. This represents a fundamental methodological gap: brain-behaviour correlations observed at single time points may not reflect stable trait-like associations if the neural measures themselves lack reliability.\u003c/p\u003e \u003cp\u003eThis study provides a targeted evaluation of rsfMRc reliability within a protocol which is representative of typical clinical research settings. While previous large-scale efforts (e.g., Noble et al., 2019; Zuo et al., 2014a) have established guidelines for functional connectomics, there is a need to evaluate these metrics under the specific constraints of clinical protocols (e.g., standard resolution and moderate scan lengths) to guide longitudinal studies in specific populations. Our primary aim was to provide guidelines for reliability expectations across different networks and time frames under controlled conditions. Secondary aims included demonstrating the practical application of reliability-informed analysis strategies through the examination of brain-behaviour relationships with aggression measures, serving as a proof-of-concept for biomarker development.\u003c/p\u003e \u003cp\u003eWe hypothesized that reliability would follow previously reported hierarchical patterns, with sensory-motor and DMN regions demonstrating higher stability than subcortical and attention networks \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. We predicted that one-month reliability would be reduced when compared to same-day measurements, but that specific networks would maintain adequate reliability (ICC \u0026gt; .50) for longitudinal applications. Additionally, we expected that matrix-wide metrics would reveal global stability patterns that complement regional ICC analyses.\u003c/p\u003e \u003cp\u003eRegarding the biomarker application, we hypothesized that temporally stable connectivity patterns would show systematic associations with aggression measures across all three time points, demonstrating the potential for reliability-informed biomarker development. Specifically, we expected that: (1) reactive aggression would be associated with connectivity patterns involving prefrontal regulatory regions and limbic areas \u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e, and (2) proactive aggression would correlate with connectivity in reward circuits involving orbitofrontal cortex and striatal regions \u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e,\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e, but only when these networks demonstrated adequate temporal stability.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003e\u003cem\u003eAge effects and signal quality assessment\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAge accounted for minimal variance in connectivity patterns across the sample (mean R\u0026sup2; = 1.8%, SD = 2.4%, range = 0.0%\u0026ndash;15.4%). The distribution was highly right-skewed, with 89.5% of connections showing R\u0026sup2; \u0026lt; 5% and only 1.2% (6 connections) showing R\u0026sup2; \u0026gt; 10%. Age-adjusted ICCs (M = .559, SD = .221) were nearly identical to unadjusted values (M = .575, SD = .210), with a negligible mean difference of -0.016 (95% CI: -0.018 to -0.013, d = -0.07). The correlation between unadjusted and age-adjusted ICCs was r = .993 (p \u0026lt; .001), demonstrating that age did not alter reliability rankings. This near-perfect correlation confirms that age correction has a negligible practical impact on ICC values (Supplementary Table S3).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo ensure reliability estimates reflected neural architecture rather than artifacts, we examined multiple quality control metrics. Head motion (mean framewise displacement and scrubbed volumes) remained consistent across all three sessions (all p \u0026gt; .05; Supplementary Figure S1), with no significant correlations between individual motion and connectivity values of stable ROIs (r \u0026lt; .15, p \u0026gt; .20). Signal amplitude (M = 0.345, SD = 0.091) showed a weak, non-significant correlation with regional reliability (r = 0.197, p = 0.280), explaining only 3.9% of ICC variance (Supplementary Figure S2). Critically, the Visual Network exhibited intermediate amplitude (M = 0.262) yet low reliability (ICC = 0.494), while the Salience Network showed comparable amplitude (M = 0.356), but substantially higher reliability (ICC = 0.640). This dissociation confirms that reliability differences reflect genuine biological variability rather than technical factors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eRegional reliability patterns\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u003cem\u003eand variance decomposition\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe test-retest reliability analysis revealed substantial heterogeneity across brain networks and time intervals (Table 2). Overall, the mean ICC across all networks and time points was .557, indicating moderate reliability, with regional ICCs ranging from .301 to .717, highlighting the importance of network-specific considerations in study design.\u003c/p\u003e\n\u003cp\u003eExceptional stability characterized the DMN lateral parietal regions (left: ICC = .717; right: ICC = .654) and the Salience Network supramarginal gyrus (left: ICC = .699; right: ICC = .685), with minimal degradation across time intervals. Language Network regions consistently showed good stability (ICCs = .573-.628) with low percentages of non-significant connections (10-16%). The Frontoparietal Network demonstrated moderate to good reliability (ICCs = .510-.599), with consistent patterns across intervals.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eConversely, several regions showed poor reliability profiles, with \u0026gt;55% of their connections showing non-significant results (p \u0026gt; .05 for ICC \u0026ne; 0 test): the Visual Network (ICCs = .382-.526, 19-42% non-significant), the DMN medial prefrontal cortex (ICC = .466, 29% non-significant), the Dorsal Attention Network frontal eye fields (ICCs = .301-.357, 48-61% non-significant), and the cerebellar anterior region (ICC = .480, 71% non-significant at short intervals).\u003c/p\u003e\n\u003cp\u003eVariance decomposition revealed that high-reliability regions are driven by stable individual differences. ICC correlated strongly with between-subject/within-subject (B/W) variance ratio (r = .805, p \u0026lt; .001; Supplementary Figure S3). Connections with excellent reliability (ICC \u0026gt; .70, n = 161) showed 3.7-fold larger B/W ratios (M = 1.65, SD = 0.55) compared to poor reliability connections (ICC \u0026lt; .40, n = 89; M = 0.45, SD = 0.07; t(248) = 20.24, p \u0026lt; .001, d = 2.67). The ROI-level analysis demonstrated that high-ICC regions (e.g., the DMN lateral parietal, the Salience Network supramarginal gyrus) exhibited between-subject variance proportions of 67.4-68.4%, while low-ICC regions (e.g., frontal eye fields, visual areas) showed balanced distributions where session-to-session fluctuations accounted for nearly half of the total variance (Supplementary Table S2).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eTemporal dynamics and multiple baseline benefit\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe comparison across measurement intervals revealed important patterns for study design. Same-day reliability (baseline vs. 1-hour) generally showed the highest ICC values (mean = .484), while longer intervals showed modest decreases (baseline vs. 1-month: .475; 1-hour vs. 1-month: .493). However, the differences were smaller than anticipated, suggesting that one-month intervals may be acceptable for many applications.\u003c/p\u003e\n\u003cp\u003eCritically, three-time-point ICC calculations yielded higher values (mean=.557) than any pairwise comparison, supporting recommendations for multiple baseline measurements in longitudinal studies \u003csup\u003e24\u003c/sup\u003e. To empirically evaluate this benefit, we averaged connectivity values from the two Day 1 sessions (Baseline and 1-hour) and calculated the ICC against one-month follow-up. This aggregation produced systematic reliability improvement: global ICC increased from 0.463 (using only first baseline) to 0.543 (using averaged baseline), representing 17.2% enhancement in temporal stability.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMatrix-wide reliability assessment and univariate-multivariate dissociation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGlobal stability metrics provided complementary perspectives on connectivity reliability. Kendall\u0026apos;s W of .798 indicated a strong preservation of connectivity rankings across sessions, suggesting that relative connectivity strengths remain stable even when absolute values show variability. Pearson correlations between connectivity matrices averaged .579 across temporal comparisons, with similar values for all pairwise comparisons (baseline vs. 1-hour: .588; baseline vs. 1-month: .574; 1-hour vs. 1-month: .574) and the Frobenius distance and Procrustes disparity showing similarly consistent values (Procrustes: 0.050-0.058; Frobenius: 25.08-27.06). These low and consistent values across all temporal comparisons confirm that the global connectivity architecture remains stable over both short-term (1-hour) and medium-term (1-month) intervals.\u003c/p\u003e\n\u003cp\u003eThis revealed a critical dissociation between edge-level and pattern-level reliability: while individual connections showed moderate mean reliability (ICC = .557), matrix-wide metrics revealed substantially higher consistency in global connectivity organization (Kendall\u0026apos;s W = .798, Pearson r = .579). This superiority of multivariate reliability reflects statistical averaging\u0026mdash;distributed patterns aggregate signals across many connections, reducing measurement noise impact in any single edge \u003csup\u003e25,26\u003c/sup\u003e, paralleling the psychometric principle that multi-item scales show higher reliability than individual items through error cancellation.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eStable Network Identification and brain-behaviour association\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBased on our reliability criteria (ICC \u0026gt; .50 and \u0026lt;55% non-significant connections), we identified 23 suitable brain regions for longitudinal research applications: 7 from the Salience Network, 4 from the Language Network, 3 from the Default Mode Network, 4 from the Frontoparietal Network, 2 from the Dorsal Attention Network, and 3 from the Sensorimotor Network. This represents 72% of all examined regions. Although exclusion of the visual cortex, cerebellar areas, and medial prefrontal cortex has important implications for studies targeting these systems, this percentage suggests that most major network connections can support longitudinal research when appropriate methodological considerations are applied.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe analysis of temporally stable connectivity patterns revealed systematic associations with aggression measures (Table 3, Figure 1). Six connectivity patterns showed significant correlations (|r| \u0026ge; .50, p-FDR \u0026lt; .05) with aggression dimensions across the three time points.\u003c/p\u003e\n\u003cp\u003eFour patterns showed associations with both reactive and proactive aggression: negative DMN lateral parietal-sensorimotor superior correlation (reactive: r = -.554; proactive: r = -.616), sensorimotor-executive control integration (SMN Superior-FPN LPFC: r = .522-.538), and DMN-Salience connectivity patterns. Two patterns were dimension specific: DMN PCC-Salience RPFC connections showed specific associations with proactive aggression (r = -.518 to .496).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo validate our reliability-informed approach, we compared the temporal stability of brain-behaviour correlations between high-ICC and low-ICC connections. High-reliability connections exhibited significantly lower temporal variability in aggression associations (mean SD of r across sessions = 0.096) compared to low-reliability connections (mean SD = 0.112; t = 2.42, p = .016, d = 0.29). Furthermore, high-ICC connections showed a 45.1% directional consistency (maintaining same correlation sign across all time points), while low-ICC connections showed only a 31.9% consistency (z = -2.10, p = .036), confirming that a low-ICC signal structure is unsuitable for stable trait-level inference. The consistency of associations across all three time points strengthens confidence that the observed relationships reflect stable traits rather than measurement artifacts, thereby supporting the validity of our reliability-informed approach.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis methodological investigation provides systematic guidelines for rsfMRc reliability expectations across major brain networks and clinically relevant time frames. Our findings reveal substantial heterogeneity in temporal stability that necessitates network-specific considerations in longitudinal study designs. The overall moderate stability (mean ICC = .557) masks critical regional variations that have direct implications for methodological decision-making in connectivity research.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eNetwork-Specific Reliability and Neurobiological Foundations\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eExceptional stability in DMN lateral parietal regions (ICCs =.654-.717 ) and Salience Network supramarginal gyrus (ICCs =.685-.699) reflects their fundamental neurobiological roles as neural \u0026quot;scaffolding\u0026quot; for stable individual differences. Our variance decomposition analysis quantified these foundations explicitly: high reliability in these networks is driven by dominant between-subject variance (exceeding 67% of total variance; Supplementary Table S2), indicating stable \u0026apos;trait-like\u0026apos; fingerprints twice as large as the combined effect of measurement noise and state-dependent fluctuations. The DMN lateral parietal nodes support self-referential thinking, autobiographical memory, and perspective-taking\u0026mdash;relatively stable characteristics forming the foundation of personal identity.\u0026nbsp;The Salience Network supramarginal gyrus integrates external sensory information with internal states, facilitating appropriate attentional resource allocation and behavioural regulation. The temporal consistency of these networks makes them ideal candidates for longitudinal research applications requiring high reliability.\u003c/p\u003e\n\u003cp\u003eIn contrast, networks showing moderate reliability\u0026mdash;Language Network (ICCs = .573-.628) and Frontoparietal Network (ICCs = .510-.599)\u0026mdash;demonstrated acceptable but variable stability reflecting involvement in both stable and dynamic processes. The Language Network\u0026apos;s stability likely reflects the fundamental role of verbal mediation and cognitive control in emotional regulation, while Frontoparietal Network components capture executive control mechanisms critical for behavioural regulation \u003csup\u003e27\u003c/sup\u003e. These networks may benefit from methodological optimization: \u0026nbsp;longer acquisition protocols (12+ minutes), multiple baseline measurements, or larger sample sizes (N\u0026gt;50) to achieve adequate power for detecting meaningful changes, representing an intermediate category where methodological optimization can enhance reliability. In addition, using fixed-atlas ROIs may slightly underestimate peak reliability compared to subject-specific parcellations, representing a conservative estimate of temporal stability.\u003c/p\u003e\n\u003cp\u003eNetworks demonstrating poor reliability\u0026mdash;Visual Network, DMN medial prefrontal cortex (ICC = .466), and Dorsal Attention Network frontal eye fields\u0026mdash;present methodological challenges but also reveal important neurobiological insights. Variance decomposition showed within-subject variance often exceeded 50% in these networks, confirming they are more permeable to participants\u0026rsquo; state during scanning. While this makes them less ideal for trait-biomarker development, it confirms their role as dynamic interfaces that are highly responsive to environmental factors. This distinction is critical: low ICC here reflects a genuine biological dominance of temporal variability over stable individual differentiation, not mere measurement error. Our lower Visual Network reliability compared to some reports (e.g., Noble et al., 2017) likely stems from technical factors: a 5mm slice thickness may introduce partial volume effects in thin occipital cortex, and eyes-open acquisition (chosen to maintain alertness) increases visual processing variability compared to eyes-closed protocols. These networks require alternative approaches, including specialized preprocessing or a focus on group-level rather than individual analyses. Visual networks, particularly higher-order visual areas, show considerable plasticity in response to environmental changes and learning experiences.\u003c/p\u003e\n\u003cp\u003eThe unexpected finding of sensorimotor network stability across all regions represents a novel finding, suggesting that motor-related connectivity patterns may serve as stable individual \u0026quot;fingerprints\u0026quot; that contribute to behavioural regulation, which supports embodied cognition theories. This challenges previous assumptions about sensorimotor variability and suggests these regions may be underutilized in biomarker development.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eTemporal dynamics and multivariate reliability\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eModest differences between short-term (1-hour; ICC = .493) and medium-term (1-month; ICC = .475) reliability provide important methodological guidance. The improved reliability observed when incorporating all three time points (mean ICC = .557) supports recommendations for multiple baseline measurements, consistent with systematic analyses demonstrating enhanced stability with repeated measurements \u003csup\u003e24\u003c/sup\u003e. Our findings suggest that averaging multiple baseline sessions\u0026mdash;even within the same day\u0026mdash;significantly enhances measurement power by stabilizing participant-specific connectivity profiles and mitigating effect size attenuation in brain-behaviour associations.\u003c/p\u003e\n\u003cp\u003eDespite moderate regional ICCs, the high Kendall\u0026apos;s W (.798) reveals fundamental network organization principles that remain preserved over clinically relevant time frames. \u0026nbsp;This dissociation between edge-level and matrix-level reliability demonstrates that connectivity configurations maintain consistency even when individual connection strengths fluctuate\u0026mdash;validating whole-brain pattern approaches while clarifying that both univariate and multivariate reliability inform analytical strategy selection. For studies prioritizing interpretability and network-specific inference, edge-level reliability informs measurement adequacy. For studies prioritizing predictive accuracy (e.g., developing clinical biomarkers), whole-brain pattern approaches that capitalize on multivariate reliability are optimal \u003csup\u003e25,28\u003c/sup\u003e. Rodriguez et al. (2025) demonstrated that matrix reliability can be enhanced through approaches that de-emphasize large coactivation patterns, potentially at the expense of edge reliability\u0026mdash;suggesting reliability optimization depends on analytical goals. Our findings provide guidelines for both analysis levels, enabling informed decisions about reliability-accuracy trade-offs.\u003c/p\u003e\n\u003cp\u003eOur network-based approach serves a complementary purpose: providing interpretable guidelines for the biological foundations of reliability differences. While whole brain patterns optimize predictive accuracy, network-level characterization facilitates mechanistic understanding and experimental design. The observation that the Salience and DMN regions show high univariate reliability while the Visual and Attention networks show state-dependence offers actionable guidance for researchers selecting measurement targets\u0026mdash;\u0026shy; information that complements but does not replace multivariate approaches.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eComparison with Existing Reliability Literature\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOur overall reliability (mean ICC = .557, SD = .202) falls within ranges reported across resting-state studies (.20\u0026ndash;.80; Noble et al., 2019), with variation that can be attributed to methodological factors:\u003c/p\u003e\n\u003cp\u003eSample Characteristics: Our homogeneous sample (ages 19-39, 94% aged 19-29, males only) represents an ideal case for establishing upper-bound reliability estimates. The minimal age effect (mean R\u0026sup2; = 1.8%) indicates age-related variability was negligible relative to both stable individual differences and temporal fluctuations\u0026mdash;the \u0026apos;trait\u0026apos; component was 30-fold larger than age-related variance, justifying our focus on stable connectivity signatures. Clinical or broader age-range samples would likely show lower reliability due to symptom fluctuations or developmental processes \u003csup\u003e30\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eScan Duration and Technical Factors: Our 9-minute acquisitions align with standard protocols but are shorter than high-reliability protocols using 12-15 minutes. Our 5mm slice thickness, while maintaining signal-to-noise ratio in clinical scanners, is relatively coarse compared to high-resolution (2mm) protocols and may hinder precise cortical delineation. These factors mean our ICCs should be interpreted as context-specific guidelines rather than universal values.\u003c/p\u003e\n\u003cp\u003eTest-Retest Intervals: The mean ICC decreased from .563 (1-hour) to .531 (1-month), consistent with biological variability accumulation over time. This reflects a fundamental distinction: short-interval reliability primarily captures measurement error, while long-interval reliability incorporates genuine temporal changes.\u003c/p\u003e\n\u003cp\u003ePreprocessing Strategy: Our anatomical CompCor approach is more aggressive than basic global signal regression. Advanced denoising improves reliability by 0.10-0.15 \u003csup\u003e31\u003c/sup\u003e, with our relatively high Salience (ICC = .685) and DMN (ICC = .660) reliability likely reflecting effective physiological noise removal in regions susceptible to cardiac/respiratory artifacts.\u003c/p\u003e\n\u003cp\u003eNetwork-Specific Patterns: The highest reliability in the Salience and lateral DMN replicates prior work \u003csup\u003e6,7\u003c/sup\u003e. Low Visual Network reliability likely reflects eyes-open acquisition, as visual networks show greater stability during eyes-closed rest when ongoing processing is minimized \u003csup\u003e32\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eMethodological Recommendations\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBased on these comparisons, we provide evidence-based recommendations:\u003c/p\u003e\n\u003cp\u003ePreprocessing: Advanced denoising (aCompCor, ICA-AROMA) combined with rigorous motion correction substantially improves reliability. Standardized bandpass filtering (0.008-0.09 Hz) maintains consistency.\u003c/p\u003e\n\u003cp\u003eAcquisition: For high-variability networks (Visual, DAN), acquisition times \u0026ge;12-15 minutes and multiple baseline sessions improve stability. Time-of-day standardization may reduce state-dependent fluctuations.\u003c/p\u003e\n\u003cp\u003eAnalysis Methods: Network-specific approaches optimize reliability\u0026mdash;individualized parcellations enhance estimates for spatially variable networks (Frontoparietal, DAN), while fixed atlases suffice for stable networks (Salience, Sensorimotor). Hybrid approaches may balance reliability and biological interpretability.\u003c/p\u003e\n\u003cp\u003eNetwork-Specific Strategies: Networks with between-subject variance \u0026gt;60% (Salience, DMN, Language) are optimal for trait biomarker research. For networks with dominant within-subject variance (\u0026lt;50%; Visual, portions of DAN), researchers should implement methodological enhancements or explicitly model state-dependence as a neurobiological feature rather than a measurement error.\u003c/p\u003e\n\u003cp\u003eStatistical Approaches: Prioritizing high-ICC connections (\u0026gt;0.50) for brain-behaviour analyses reduces measurement error. Incorporating reliability estimates into models (measurement error models, latent variables) provides more accurate longitudinal inferences.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBrain-Behaviour Relationships in Aggression\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe identification of 23 brain regions with adequate temporal stability (ICC \u0026gt; .50, \u0026lt;55% non-significant connections) represents a crucial methodological decision that strengthens the validity of brain-behaviour correlations. This selection criterion reflects a fundamental principle: only temporally stable neural connections can serve as reliable indicators of enduring psychological traits. The 23 stable regions span multiple functional networks (72% of examined regions), representing a \u0026nbsp;neurobiologically meaningful distribution and aligning with networks involved in social cognition, emotional regulation, and executive control\u0026mdash;processes directly relevant to psychological trait expression. The predominance of SN regions aligns with this network\u0026apos;s role as an integrative hub that coordinates responses between different functional networks \u003csup\u003e33\u003c/sup\u003e, while the inclusion of language network regions reflects the importance of verbal mediation and cognitive control in emotional regulation. This \u0026apos;reliability-informed\u0026apos; strategy maximizes signal-to-noise ratio by pre-selecting connections where between-subject variance exceeds 60% (B/W ratios \u0026gt;1), reducing measurement error attenuation and allowing precise estimates of relationships between functional architecture and traits.\u003c/p\u003e\n\u003cp\u003eOur analysis demonstrates the practical value of reliability-informed approaches by identifying stable connectivity patterns associated with aggression. Six connectivity patterns showed significant correlations (|r| \u0026ge; .50, p-FDR \u0026lt; .05) across all three time points. Shared patterns involving sensorimotor-executive control integration (SMN-FPN; r = .522-.538) supports embodied cognition theories proposing that motor system activation underlies aggressive responses \u003csup\u003e34\u003c/sup\u003e. Negative associations between the DMN lateral parietal regions and sensorimotor areas (r = -.554 to -.616) may reflect disrupted self-referential processing in individuals with higher aggression levels, consistent with previous findings linking DMN dysfunction to antisocial behaviour \u003csup\u003e18\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eProactive aggression showed specific DMN-SN connectivity patterns (DMN PCC with SN regions; r = -.512 to -.518), suggesting instrumental aggression may involve altered coordination between self-referential processing and salience detection systems. This pattern aligns with theoretical models proposing that proactive aggression requires different neural coordination than reactive aggression, involving more deliberate, goal-directed processes that require cognitive control and planning abilities \u003csup\u003e19\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eThe absence of limbic system connections likely reflects their poor temporal stability rather than lack of functional relevance\u0026nbsp;\u0026mdash;variance decomposition confirmed that these regions show within-subject fluctuations exceeding 55%, suggesting a potential relationship with aggression may be obscured by state-dependent noise in resting-state protocols. This highlights the importance of reliability-informed analytical approaches: by focusing on the 23 regions where between-subject variance was dominant, we ensured that the reported correlations with aggression represent enduring neural traits.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eContextualizing Effect Sizes and Statistical Power\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWhile our results show certain networks reach ICC values of 0.6-0.7, we moderate claims regarding immediate biomarker readiness. Detecting medium effect sizes (d = 0.5) in longitudinal interventions with ICC = 0.6 requires approximately N=45-50 for 80% power \u003csup\u003e8\u003c/sup\u003e. Therefore, context-specific power analyses are essential\u0026mdash;lower reliability necessitates significantly larger cohorts to avoid Type II errors.\u003c/p\u003e\n\u003cp\u003eOur behavioural effect sizes (|r| \u0026ge; .50) must be contextualized within current brain-wide association study (BWAS) landscapes. Recent large-scale evidence suggests replicable rsfMRI-behaviour associations are often considerably smaller \u003csup\u003e35\u003c/sup\u003e. Given our sample size (N=34), reported correlations should be considered exploratory and potentially subject to inflation. However, our repeated-measures design and reliability-informed filtering help mitigate measurement noise that requires much larger samples in single-session cross-sectional designs. The restricted aggression range in our healthy sample (well below clinical cutoffs) may limit effect sizes and generalisability to clinical populations, though observed correlations within this restricted range are notable, as range restriction typically attenuates effects. High reliability is not only required for longitudinal tracking but also a prerequisite for cross-sectional BWAS validity\u0026mdash;without stable estimates, correlations are attenuated by measurement error, leading to replicability challenges.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLimitations and future research directions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOur sample size (N=34) limits the precision of ICC estimates (confidence intervals approximately \u0026plusmn;0.20), though findings demonstrate substantial reliability patterns are unlikely explained by sampling variability. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSingle-site, male-only design limits the generalisability of specific ICC values\u0026mdash;the exclusive male sample was chosen to reduce variability from sex-related differences in connectivity patterns and hormonal fluctuations \u003csup\u003e36,37\u003c/sup\u003e. Our one-month maximum interval provides guidelines for clinically relevant time frames but may not capture longer-term stability patterns important for trait-based research. The correlational nature of our current brain-behaviour analysis prevents direct causal inferences about directionality, requiring complementary experimental approaches for mechanistic understanding. However, the identified high temporal stability in specific circuits provides the necessary methodological foundation for future longitudinal designs (e.g., latent change score models) that can formally test directionality.\u003c/p\u003e\n\u003cp\u003eFinally, the use of a predefined group-level atlas (CONN networks.32) may not capture individual-specific functional peaks, particularly in networks with high inter-individual spatial variability such as the frontoparietal and attention systems. While this atlas is a widely validated standard, future studies could complement fixed-atlas approaches with individualized parcellations (e.g., Kong et al., 2019) to maximize reliability in high-variability networks.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eStudy Design\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe decision to conduct a dedicated reliability study rather than analysing existing data was motivated by several factors: (1) the need for standardized acquisition parameters optimized for reliability assessment, (2) the requirement for precise timing intervals to capture both short-term and medium-term stability, and (3) the importance of controlling environmental and procedural variables that could affect reliability estimates. This prospective design allowed us to implement optimal protocols specific to reliability assessment rather than adapting analyses from studies designed for other purposes.\u003c/p\u003e\n\u003cp\u003eThe specific timing intervals (1 hour and 1 month) were selected based on methodological considerations critical to our research program: the 1-hour interval allows assessment of short-term measurement stability while controlling for scanner-related variability \u003csup\u003e5\u003c/sup\u003e, while the 1-month interval provides insight into stability over a timeframe commonly used in intervention and clinical studies \u003csup\u003e4\u003c/sup\u003e. These intervals directly address the practical needs of longitudinal connectivity research in our laboratory and similar research environments. The three-time-point design enables comparison of different reliability intervals within the same participants, reducing confounding variables that affect cross-study comparisons. This approach also allows the calculation of reliability estimates that incorporate all time points, potentially providing more stable estimates than pairwise comparisons alone \u003csup\u003e24\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eParticipants\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSample size determination for reliability studies requires different considerations than traditional hypothesis testing studies. Following recommendations for ICC estimation Haga clic o pulse aqu\u0026iacute; para escribir texto.\u003cspan class=\"MsoPlaceholderText\"\u003eHaga clic o pulse aqu\u0026iacute; para escribir texto.\u003c/span\u003e, adequate precision typically requires 30-50 participants for moderate reliability coefficients. For ICC = 0.6 with a desired precision of \u0026plusmn;0.15 (95% CI width of 0.30), approximately 45-50 participants would be optimal. Due to resource constraints and the substantial participant burden of three scanning sessions, we recruited 34 healthy male participants. Recent methodological investigations have successfully demonstrated reliability patterns with N=20 \u003csup\u003e40\u003c/sup\u003e, N=46 for clinical populations \u003csup\u003e10\u003c/sup\u003e, and N=26 for multi-session designs \u003csup\u003e41\u003c/sup\u003e. Even intensive longitudinal studies with 10 repeated measures have used N=30 Haga clic o pulse aqu\u0026iacute; para escribir texto.\u003cspan class=\"MsoPlaceholderText\"\u003eHaga clic o pulse aqu\u0026iacute; para escribir texto.\u003c/span\u003e, demonstrating that our three-session design with N=34 represents a reasonable methodological approach for establishing reliability guidelines. The exclusive focus on male participants was chosen to reduce variability from sex-related differences in connectivity patterns and hormonal fluctuations \u003csup\u003e36,37\u003c/sup\u003e, maximizing our ability to detect reliability patterns within our sample size constraints.\u003c/p\u003e\n\u003cp\u003eInclusion criteria comprised participants aged 18-29 years, free from physical or mental disorders that could affect rsfMRc acquisition or introduce systematic variability. All participants were interviewed by two independent reviewers using a structured protocol validated in previous studies. Thirty-four healthy male young adults participated in this study (see Table 1 for complete demographics). Thirty-six participants were initially recruited from the general university population through institutional announcements at Universidad de Valencia and communication channels at Hospital Universitari i Polit\u0026egrave;cnic La Fe. No selection criteria related to aggression levels or personality traits were applied. Online screening questionnaires assessed eligibility based on standard exclusion criteria before MRI attendance. Two participants were excluded due to incomplete image acquisition, resulting in a final sample of 34 participants.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eParticipants\u0026apos; ages ranged from 19 to 39 years (M = 23.00, SD = 4.44, Mdn = 21). All participants were right-handed, native Spanish speakers with university-level education, and all of them were interviewed by trained researchers (with extensive experience in clinical psychology) to assess their mental health. Cohen\u0026rsquo;s kappa, used to assess the inter-rater agreement between qualitative interviewers, ranged from .78 to .89 in each assessed dimension. Exclusion criteria were: history of neurological disorders, head trauma with loss of consciousness, current psychiatric medication, substance abuse or MRI contraindications. All scanning sessions were conducted between 9:00 AM and 1:00 PM to control for circadian effects on brain function \u003csup\u003e42\u003c/sup\u003e. Participants were instructed to avoid caffeine for at least 4 hours before scanning. Substance use was assessed via self-report questionnaires (AUDIT for alcohol; Severity of Dependence Scale for cannabis); no participants reported problematic use. The study was approved by the Clinical Research Ethics Committee of Hospital Universitari i Polit\u0026egrave;cnic La Fe (2023-MAG-2671408). All participants provided written informed consent and received monetary compensation. All methods were performed in accordance with the relevant guidelines and regulations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eProcedure\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants completed online questionnaires and structured interviews prior to neuroimaging sessions. In the initial session, participants complemented the interview with various health, alcohol, and/or other drugs misuse, personality, and psychopathology aspects. Following the psychological evaluation, participants were scheduled for a resting-state fMRI (rs-fMRI) session within one week. For rsfMRc acquisition, participants were instructed to remain still with eyes open and fixated on a central point, while avoiding deliberate thoughts\u0026mdash;conditions known to optimize signal reliability \u003csup\u003e3,43\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eThe measurement protocol involved three scanning sessions: two sessions separated by one hour on the same day and a third session 30 days later. This design enables assessment of both short-term stability (accounting for factors like positioning differences, scanner drift) and medium-term stability (incorporating biological and state-related variability). Participants were asked to maintain consistent sleep schedules and avoid caffeine on scanning days to minimize state-related variability.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eAggression Assessment\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants completed the Spanish version of the Reactive-Proactive Aggression Questionnaire (RPQ; Raine et al., 2006; Spanish adaptation: Mu\u0026ntilde;oz-Rivas, 2007). The RPQ comprises 23 items assessing two dimensions: Reactive aggression (11 items; impulsive, emotion-driven responses to provocation, e.g., \u0026quot;Gotten angry or mad when you lost at a game\u0026quot;) and Proactive aggression (12 items; goal-directed, instrumental aggressive behaviour; e.g., \u0026quot;Hurt others to win a game\u0026quot;). Items are rated on a 3-point scale (0=never, 1=sometimes, 2=often). From the more extensive battery, we selected this instrument because its psychometric properties and temporal stability \u003csup\u003e45\u003c/sup\u003e allow us to distinguish levels of propensity towards violence in forensic and non-forensic populations \u003csup\u003e46\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eIn our sample, RPQ scores indicated low-to-moderate aggression levels typical of healthy young adults without clinical psychopathology (Reactive: M = 6.59, SD = 2.96, range = 0-13; Proactive: M = 2.15, SD = 1.89, range = 0-9). These means are substantially below published norms for clinical populations with conduct problems or antisocial traits \u003csup\u003e16\u003c/sup\u003e and reflect natural variability in general population sampling rather than enrichment for high/low aggression profiles. The internal consistency was adequate (Reactive: Cronbach\u0026apos;s \u0026alpha; = .78; Proactive: \u0026alpha; = .72), comparable to published psychometric properties of the Spanish RPQ \u003csup\u003e44\u003c/sup\u003e. The RPQ has shown excellent test-retest reliability (r \u0026gt; .70) in its Spanish validation \u003csup\u003e44\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003efMRI data acquisition and analysis\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe fMRI data were acquired on a 3.0\u0026nbsp;T magnet (Achieva TX, Philips Healthcare Best, The Netherlands) using an 8-channel head coil with parallel acquisition technology (SENSE). Participants were instructed to remain still during the fMRI. The acquisition protocol consisted of a T1-weighted high spatial resolution 3D gradient echo sequence (TE =3\u0026nbsp;ms, TR=6.2\u0026nbsp;s, flip angle=10\u0026ordm;, voxel size=1\u0026times;1\u0026times;1\u0026nbsp;mm\u003csup\u003e3\u003c/sup\u003e, and 6 min duration) and a T2*-weighted 2D EPI BOLD rsfMRc sequence (TE=35 ms, TR=2000 ms, temporal dynamics=265, pixel size=1.8\u0026times;1.8 mm2, slice thickness=5 mm, and 9 min total duration).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFunctional and anatomical data were preprocessed using a modular preprocessing pipeline \u003csup\u003e47\u003c/sup\u003e, based on the CONN toolbox \u003csup\u003e\u003cspan lang=\"EN-GB\"\u003e48\u003c/span\u003e\u003c/sup\u003e and SPM (SPM12-Statistical Parametric Mapping, n.d.), including realignment with correction of susceptibility distortion interactions, slice-timing correction, outlier detection, direct segmentation and MNI-space normalization, and smoothing.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFunctional data were realigned using the SPM realign \u0026amp; unwarp procedure \u003csup\u003e50\u003c/sup\u003e, where all scans were registered to a reference image (first scan of the first session) using a least squares approach and a six-parameter (rigid body) transformation \u003csup\u003e51\u003c/sup\u003e and resampled using b-spline interpolation to correct for motion and magnetic susceptibility interactions. Temporal misalignment between different slices of the functional data (acquired in ascending order) was corrected following SPM slice-timing correction (STC) procedure \u003csup\u003e52\u003c/sup\u003e, using sinc temporal interpolation to resample each slice BOLD time series to a common mid-acquisition time. Potential outlier scans were identified using ART \u003csup\u003e53\u003c/sup\u003e, as acquisitions with framewise displacement above 0.9 mm or global BOLD signal changes above 5 standard deviations \u003csup\u003e54,55\u003c/sup\u003e and a reference BOLD image was computed for each subject by averaging all scans and excluding outliers. Functional and anatomical data were normalized into standard MNI space, segmented into grey matter, white matter, and CSF tissue classes, and resampled to 2 mm isotropic voxels following a direct normalization procedure \u003csup\u003e54,56\u003c/sup\u003e using the SPM unified segmentation and normalization algorithm \u003csup\u003e\u003cspan lang=\"EN-GB\"\u003e57,58\u003c/span\u003e\u003c/sup\u003e. Last, functional data were smoothed using spatial convolution with a Gaussian kernel of 8 mm full width at half maximum (FWHM).\u003c/p\u003e\n\u003cp\u003eFunctional data were denoised using a standard denoising pipeline \u003csup\u003e47\u003c/sup\u003e including the regression of potential confounding effects characterized by white matter time series (5 CompCor noise components), CSF time series (5 CompCor noise components), motion parameters and their first order derivatives (12 factors) \u003csup\u003e59\u003c/sup\u003e, outlier scans (below 28 factors) \u003csup\u003e55\u003c/sup\u003e, session effects and their first order derivatives (2 factors), and linear trends (2 factors) within each functional run, followed by bandpass frequency filtering of the BOLD time series \u003csup\u003e60\u003c/sup\u003e between 0.008 Hz and 0.09 Hz \u003csup\u003e61,62\u003c/sup\u003e. Noise components within white matter and CSF were estimated by computing the average BOLD signal and the largest principal components orthogonal to the BOLD average, motion parameters, and outlier scans within each subject\u0026apos;s eroded segmentation masks.\u003c/p\u003e\n\u003cp\u003eROI-to-ROI connectivity (RRC) matrices were estimated, characterizing the functional connectivity between each pair of ROIs in all three conditions individually. ROIs were selected using the 32-networks atlas in the CONN toolbox, which is derived from a group-ICA decomposition of 497 participants from the Human Connectome Project (HCP). This atlas provides a validated representation of canonical functional networks, offering a balance between spatial specificity, data reduction, and interpretability within established neurobiological frameworks\u0026nbsp;\u003csup\u003e\u003cspan lang=\"EN-GB\"\u003e54\u003c/span\u003e\u003c/sup\u003e. Functional connectivity strength was represented by Fisher-transformed bivariate correlation coefficients from a general linear model (weighted-GLM\u003csup\u003e47\u003c/sup\u003e), estimated separately for each pair of ROIs, characterising the association between their BOLD signal time series.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe analysis focused on seven major functional brain networks and their constituent regions, selected based on established roles in attention, emotion processing, and behavioural regulation, and potential relevance to the clinical variables. The DMN included the lateral parietal regions (LP, left and right), MPFC, and PCC. The Dorsal Attention Network (DAN) comprised the frontal eye fields (FEF) and intraparietal sulcus (IPS), both bilateral. The Frontoparietal Network (FPN) included the lateral prefrontal cortex (LPFC) and posterior parietal cortex (PPC), both bilateral. The Language Network (LN) consisted of the inferior frontal gyrus (IFG) and posterior superior temporal gyrus (pSTG), both bilateral. The Salience Network (SN) encompassed the anterior cingulate cortex (ACC), anterior insula (aInsula), rostral prefrontal cortex (RPFC), and supramarginal gyrus (SMG), all bilateral except for the ACC. The Sensorimotor Network (SMN) included bilateral lateral regions and a superior region, while the Visual Network (VN) comprised bilateral lateral regions, a medial region, and an occipital region. Additionally, cerebellar regions were examined, including both anterior and posterior portions. These networks and their nodes were analysed for their connectivity patterns and stability across time points.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eAge effects on reliability\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo verify that reliability estimates were not confounded by age-related variability, we calculated age-adjusted ICCs by residualizing connectivity values for linear age effects using ordinary least squares regression separately for each connection. For each connection, we regressed connectivity values onto age across subjects, calculated residuals, and added the grand mean to maintain the original scale. ICCs were then computed on these age-adjusted connectivity values.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eReliability assessment\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTest-retest reliability was assessed using multiple complementary approaches. Intraclass correlation coefficients (ICC model 2,k) quantified consistency across the three time points, providing the primary metric for regional reliability assessment. We calculated ICCs for all time points combined and for pairwise comparisons (baseline vs. 1-hour, baseline vs. 1-month, 1-hour vs. 1-month).\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;For each connection, statistical significance of ICC estimates was assessed using the F-test provided by Pingouin\u0026apos;s ICC calculation \u003csup\u003e63\u003c/sup\u003e, which tests the null hypothesis that ICC = 0 (no temporal consistency). The F-test compares between-subject variance to within-subject variance, with degrees of freedom df1 = n_subjects - 1 = 33 and df2 = n_subjects \u0026times; (n_sessions - 1) = 68. Connections showing p \u0026gt; .05 were classified as having non-significant temporal stability. To characterize regional reliability patterns, we calculated the percentage of non-significant connections for each ROI (i.e., the proportion of that ROI\u0026apos;s connections failing to show significant temporal consistency). ROIs for which \u0026gt;55% of connections were non-significant were classified as \u0026apos;highly unstable\u0026apos; based on the median split of this distribution across all ROIs.\u003c/p\u003e\n\u003cp\u003eTo complement regional ICC analyses, we calculated several matrix-wide metrics providing global perspectives on connectivity stability. Pearson correlation was used to assess the overall similarity of connectivity matrices between sessions, with higher coefficients indicating greater consistency in global network organization \u003csup\u003e6\u003c/sup\u003e. Kendall\u0026rsquo;s W, a non-parametric measure, was applied to evaluate the consistency of rank-order connectivity strengths across sessions, offering robust insights into global reproducibility \u003csup\u003e64\u003c/sup\u003e. The Frobenius distance quantified the overall differences between connectivity matrices by computing the root mean square of element-wise differences, with smaller distances indicating greater similarity in global network structure \u003csup\u003e65\u003c/sup\u003e. The Procrustes disparity assessed the geometric alignment of connectivity matrices, considering both scale and orientation, in which values close to zero indicated high structural similarity and preservation of global connectivity patterns over time \u003csup\u003e\u003cspan lang=\"EN-GB\"\u003e66\u003c/span\u003e\u003c/sup\u003e. These metrics provide information about global network organization stability that may not be captured by regional ICC values alone.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eVariance decomposition and quality assessment\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo ensure that reliability differences were not driven by regional signal properties, we quantified signal quality for all 32 ROIs. Signal amplitude was calculated as the temporal standard deviation of pre-processed low-frequency BOLD fluctuations (0.01-0.1 Hz), reflecting neuronal signal variability after removal of motion artifacts, physiological noise, and scanner drift (equivalent to Amplitude of Low-Frequency Fluctuations [ALFF]; \u003csup\u003e15,67\u003c/sup\u003e. All data passed CONN toolbox\u0026apos;s automated quality control procedures prior to analysis.\u003c/p\u003e\n\u003cp\u003eTo determine whether high-reliability connections reflect stable individual differences or merely low measurement noise, we decomposed the total variance into between-subject and within-subject components for all 496 connections \u003csup\u003e63\u003c/sup\u003e. This analysis directly quantifies the relative magnitude of stable individual differences versus temporal fluctuations, contextualizing which networks are most suitable for individual-differences research \u003csup\u003e68\u003c/sup\u003e. For each connection, we calculated: (1) between-subject variance (\u0026sigma;\u0026sup2;_B) as the variance of subject-specific means across the three sessions, representing stable individual differences, and (2) within-subject variance (\u0026sigma;\u0026sup2;_W) as the average variance within subjects across sessions, representing temporal fluctuations and measurement error.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe between/within (B/W) ratio (\u0026sigma;\u0026sup2;_B / \u0026sigma;\u0026sup2;_W) quantifies whether individual differences exceed temporal variability: ratios \u0026gt; 1.0 indicate that stable differences between individuals exceed fluctuations within individuals, with larger ratios reflecting more pronounced individual differentiation.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eBrain-behaviour reliability analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo examine relationships between stable connectivity patterns and aggression, we employed a reliability-informed analytical strategy, using the 23 most reliable ROIs (ICC \u0026gt; .50 and \u0026lt;55% of connections showing non-significant temporal consistency; n = 253 possible connections) from the previous analysis for the proof-of-concept biomarker analysis. In addition, we ensured that retained connections reflected stable inter-individual differences rather than measurement noise or session-specific fluctuations.\u003c/p\u003e\n\u003cp\u003eFor each connection among these high-reliability ROIs, we calculated functional connectivity values for each subject at each of the three scanning sessions. We then used the repeated-measures correlation (rm_corr; Bakdash \u0026amp; Marusich, 2017) to examine associations between session-specific connectivity values and trait aggression scores (RPQ reactive and proactive subscales).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe repeated-measures correlation accounts for within-subject dependencies across sessions by computing the common within-individual association between two variables measured repeatedly. A key prerequisite for the valid application of the repeated-measures correlation (\u003cem\u003erm_corr\u003c/em\u003e) is the assumption of stationarity across repeated measurements. Bakdash \u0026amp; Marusich (2017) emphasize that the method may be inappropriate if systematic changes or dominant within-subject fluctuations exist across time points. To ensure the integrity of our behavioural inferences, we restricted our analysis to high-reliability connections, where measured signal should primarily reflect stable inter-individual trait differences rather than session-specific noise or fluctuations. Connections showing |r| \u0026ge; .50 (p-FDR \u0026lt; .05) were considered meaningful associations, following conventional guidelines for effect size interpretation in correlational research (Cohen, 1988). False discovery rate (FDR) correction via the Benjamini-Hochberg procedure controlled for multiple comparisons across all tested connections.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis investigation establishes that rsfMRc reliability varies substantially across networks and must be considered in all phases of longitudinal study design. Network-specific reliability patterns provide empirical guidance for methodological decisions, while reliability-informed analysis strategies offer frameworks for robust biomarker development. Networks with excellent reliability offer immediate opportunities for clinical applications, while those with poor reliability require innovative methodological solutions. The identification of stable neural signatures associated with aggression traits enhances our understanding of their neurobiological basis and offers promising opportunities for developing connectivity-based biomarkers and targeted interventions. The temporal stability in brain-behaviour relationships provides a methodological framework for future research seeking to establish reliable neural markers of psychological traits, representing a significant step toward precision approaches in mental health, where stable neural signatures could inform personalized intervention strategies based on individual connectivity profiles.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research was partially supported by grant number PID2022-142287OA-I00; funded by MCIN/AEI/10.13039/501100011033 and by ERDF, EU\u0026rdquo; and The Prometeo Program for research groups of excellence of the Ministry of Innovation, Universities, Science and Digital Society of the Generalitat Valenciana, grant number CIPROM/2021/46.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003estatement\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe raw MRI data that support the findings of this study are not publicly available due to legal and ethical restrictions imposed by the approving Clinical Research Ethics Committee. Processed functional connectivity matrices, clinical and behavioural variables, and all analysis code are publicly available in the GitHub repository at [https://github.com/beserma/test_retest]. Requests for further information regarding data access may be directed to the corresponding author.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe author(s) declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eM.B.-R.: Conceptualization, Data curation, Formal analysis, Methodology, Software, Visualization, Writing \u0026ndash; original draft, Writing \u0026ndash; review \u0026amp; editing. \u0026Aacute;.R.-M.: Data curation, Funding acquisition, Project administration, Supervision, Writing \u0026ndash; review \u0026amp; editing. C.S.-C.: Data curation. L.C.-A.: Writing \u0026ndash; review \u0026amp; editing. F.A.: Writing \u0026ndash; review \u0026amp; editing. L.M.-A.: Writing \u0026ndash; review \u0026amp; editing. L.M.-B.: Writing \u0026ndash; review \u0026amp; editing.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eCosta, P. T. \u0026amp; McCrae, R. R. The Five-Factor Model of Personality and Its Relevance to Personality Disorders. \u003cem\u003eJ. Pers. Disord\u003c/em\u003e. \u003cb\u003e6\u003c/b\u003e, 343\u0026ndash;359 (1992).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWatson, D. Stability versus change, dependability versus error: Issues in the assessment of personality over time. \u003cem\u003eJ. Res. Pers.\u003c/em\u003e \u003cb\u003e38\u003c/b\u003e, 319\u0026ndash;350 (2004).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNoble, S. et al. Influences on the Test\u0026ndash;Retest Reliability of Functional Connectivity MRI and its Relationship with Behavioral Utility. \u003cem\u003eCereb. Cortex\u003c/em\u003e. \u003cb\u003e27\u003c/b\u003e, 5415\u0026ndash;5429 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eElliott, M. L. et al. What Is the Test-Retest Reliability of Common Task-Functional MRI Measures? New Empirical Evidence and a Meta-Analysis. \u003cem\u003ePsychol. Sci.\u003c/em\u003e \u003cb\u003e31\u003c/b\u003e, 792\u0026ndash;806 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShah, L. M., Cramer, J. A., Ferguson, M. A., Birn, R. M. \u0026amp; Anderson, J. S. Reliability and reproducibility of individual differences in functional connectivity acquired during task and resting state. \u003cem\u003eBrain Behav\u003c/em\u003e \u003cb\u003e6\u003c/b\u003e, (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShehzad, Z. et al. The Resting Brain: Unconstrained yet Reliable. \u003cem\u003eCereb. Cortex\u003c/em\u003e. \u003cb\u003e19\u003c/b\u003e, 2209\u0026ndash;2229 (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBraun, U. et al. Test\u0026ndash;retest reliability of resting-state connectivity network characteristics using fMRI and graph theoretical measures. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e59\u003c/b\u003e, 1404\u0026ndash;1412 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNoble, S., Scheinost, D. \u0026amp; Constable, R. T. A decade of test-retest reliability of functional connectivity: A systematic review and meta-analysis. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e203\u003c/b\u003e, 116157 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZuo, X. N. \u0026amp; Xing, X. X. Test-retest reliabilities of resting-state FMRI measurements in human brain functional connectomics: A systems neuroscience perspective. \u003cem\u003eNeurosci. Biobehav Rev.\u003c/em\u003e \u003cb\u003e45\u003c/b\u003e, 100\u0026ndash;118 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnderson, J. S., Ferguson, M. A. \u0026amp; Lopez-Larson, M. Yurgelun-Todd, D. Reproducibility of Single-Subject Functional Connectivity Measurements. \u003cem\u003eAm. J. Neuroradiol.\u003c/em\u003e \u003cb\u003e32\u003c/b\u003e, 548\u0026ndash;555 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBirn, R. M. et al. The effect of scan length on the reliability of resting-state fMRI connectivity estimates. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e83\u003c/b\u003e, 550\u0026ndash;558 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, J. H. et al. Graph Theoretical Analysis of Functional Brain Networks: Test-Retest Evaluation on Short- and Long-Term Resting-State Functional MRI Data. \u003cem\u003ePLoS One\u003c/em\u003e. \u003cb\u003e6\u003c/b\u003e, e21976 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBlautzik, J. et al. Classifying fMRI-derived resting-state connectivity patterns according to their daily rhythmicity. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e71\u003c/b\u003e, 298\u0026ndash;306 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChou, Y., -h., Panych, L. P., Dickey, C. C., Petrella, J. R. \u0026amp; Chen, N. -k. Investigation of Long-Term Reproducibility of Intrinsic Connectivity Network Mapping: A Resting-State fMRI Study. \u003cem\u003eAm. J. Neuroradiol.\u003c/em\u003e \u003cb\u003e33\u003c/b\u003e, 833\u0026ndash;838 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZuo, X. N. et al. Reliable intrinsic connectivity networks: Test\u0026ndash;retest evaluation using ICA and dual regression approach. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e49\u003c/b\u003e, 2163\u0026ndash;2177 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRaine, A. et al. The reactive\u0026ndash;proactive aggression questionnaire: differential correlates of reactive and proactive aggression in adolescent boys. \u003cem\u003eAggress. Behav.\u003c/em\u003e \u003cb\u003e32\u003c/b\u003e, 159\u0026ndash;171 (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGiromini, L. et al. Validity and Reliability of the Interpersonal Competence Questionnaire: Empirical Evidence from an Italian Study. \u003cem\u003eJ. Psychopathol. Behav. Assess.\u003c/em\u003e \u003cb\u003e38\u003c/b\u003e, 113\u0026ndash;123 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRomero-Mart\u0026iacute;nez, \u0026Aacute;., Sarrate-Costa, C. \u0026amp; Moya-Albiol, L. Reactive vs proactive aggression: A differential psychobiological profile? Conclusions derived from a systematic review. \u003cem\u003eNeurosci. Biobehav Rev.\u003c/em\u003e \u003cb\u003e136\u003c/b\u003e, 104626 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRomero-Mart\u0026iacute;nez, \u0026Aacute;. et al. The Brain Resting-State Functional Connectivity Underlying Violence Proneness: Is It a Reliable Marker for Neurocriminology? A Systematic Review. \u003cem\u003eBehav. Sci.\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e, 11 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDavidson, R. J., Putnam, K. M. \u0026amp; Larson, C. L. Dysfunction in the Neural Circuitry of Emotion Regulation\u0026ndash;A Possible Prelude to Violence. \u003cem\u003eScience (\u003c/em\u003e). 289, 591\u0026ndash;594 (2000).). 289, 591\u0026ndash;594 (2000). (1979).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFulwiler, C. E., King, J. A. \u0026amp; Zhang, N. Amygdala\u0026ndash;orbitofrontal resting-state functional connectivity is associated with trait anger. \u003cem\u003eNeuroreport\u003c/em\u003e \u003cb\u003e23\u003c/b\u003e, 606\u0026ndash;610 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBlair, R. J. R. The Neurobiology of Impulsive Aggression. \u003cem\u003eJ. Child. Adolesc. Psychopharmacol.\u003c/em\u003e \u003cb\u003e26\u003c/b\u003e, 4\u0026ndash;9 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRosell, D. R. \u0026amp; Siever, L. J. The neurobiology of aggression and violence. \u003cem\u003eCNS Spectr.\u003c/em\u003e \u003cb\u003e20\u003c/b\u003e, 254\u0026ndash;279 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePannunzi, M. et al. Resting-state fMRI correlations: From link-wise unreliability to whole brain stability. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e157\u003c/b\u003e, 250\u0026ndash;262 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaxali, A., Angstadt, M., Rutherford, S. \u0026amp; Sripada, C. Boost in Test\u0026ndash;Retest Reliability in Resting State fMRI with Predictive Modeling. \u003cem\u003eCereb. Cortex\u003c/em\u003e. \u003cb\u003e31\u003c/b\u003e, 2822\u0026ndash;2833 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCamp, C. C., Noble, S., Scheinost, D., Stringaris, A. \u0026amp; Nielson, D. M. Test-Retest Reliability of Functional Connectivity in Adolescents With Depression. \u003cem\u003eBiol. Psychiatry Cogn. Neurosci. Neuroimaging\u003c/em\u003e. \u003cb\u003e9\u003c/b\u003e, 21\u0026ndash;29 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBressler, S. L. \u0026amp; Menon, V. Large-scale brain networks in cognition: emerging methods and principles. \u003cem\u003eTrends Cogn. Sci.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 277\u0026ndash;290 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNoble, S., Curtiss, J., Pessoa, L. \u0026amp; Scheinost, D. The tip of the iceberg: A call to embrace anti-localizationism in human neuroscience research. \u003cem\u003eImaging Neuroscience\u003c/em\u003e \u003cb\u003e2\u003c/b\u003e, (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRodriguez, R. X., Noble, S., Camp, C. C. \u0026amp; Scheinost, D. Connectome caricatures remove large-amplitude coactivation patterns in resting-state fMRI to emphasize individual differences. \u003cem\u003eNat. Neurosci.\u003c/em\u003e \u003cb\u003e28\u003c/b\u003e, 2601\u0026ndash;2611 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZuo, X. N. et al. An open science resource for establishing reliability and reproducibility in functional connectomics. \u003cem\u003eSci. Data\u003c/em\u003e. \u003cb\u003e1\u003c/b\u003e, 140049 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eParkes, L., Fulcher, B., Y\u0026uuml;cel, M. \u0026amp; Fornito, A. An evaluation of the efficacy, reliability, and sensitivity of motion correction strategies for resting-state functional MRI. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e171\u003c/b\u003e, 415\u0026ndash;436 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePatriat, R. et al. The effect of resting condition on resting-state fMRI reliability and consistency: A comparison between resting with eyes open, closed, and fixated. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e78\u003c/b\u003e, 463\u0026ndash;473 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMenon, V. Large-scale brain networks and psychopathology: a unifying triple network model. \u003cem\u003eTrends Cogn. Sci.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 483\u0026ndash;506 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYankouskaya, A. \u0026amp; Sui, J. Self-prioritization is supported by interactions between large‐scale brain networks. \u003cem\u003eEur. J. Neurosci.\u003c/em\u003e \u003cb\u003e55\u003c/b\u003e, 1244\u0026ndash;1261 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarek, S. et al. Reproducible brain-wide association studies require thousands of individuals. \u003cem\u003eNature\u003c/em\u003e \u003cb\u003e603\u003c/b\u003e, 654\u0026ndash;660 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEngman, J., Linnman, C., Van Dijk, K. R. A. \u0026amp; Milad, M. R. Amygdala subnuclei resting-state functional connectivity sex and estrogen differences. \u003cem\u003ePsychoneuroendocrinology\u003c/em\u003e \u003cb\u003e63\u003c/b\u003e, 34\u0026ndash;42 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRitchie, S. J. et al. Sex Differences in the Adult Human Brain: Evidence from 5216 UK Biobank Participants. \u003cem\u003eCereb. Cortex\u003c/em\u003e. \u003cb\u003e28\u003c/b\u003e, 2959\u0026ndash;2975 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKong, R. et al. Spatial Topography of Individual-Specific Cortical Networks Predicts Human Cognition, Personality, and Emotion. \u003cem\u003eCereb. Cortex\u003c/em\u003e. \u003cb\u003e29\u003c/b\u003e, 2533\u0026ndash;2551 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKoo, T. K. \u0026amp; Li, M. Y. A Guideline of Selecting and Reporting Intraclass Correlation Coefficients for Reliability Research. \u003cem\u003eJ. Chiropr. Med.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 155\u0026ndash;163 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuo, C. C. et al. One-year test\u0026ndash;retest reliability of intrinsic connectivity network fMRI in older adults. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e61\u003c/b\u003e, 1471\u0026ndash;1483 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSheu, L. K., Jennings, J. R. \u0026amp; Gianaros, P. J. Test\u0026ndash;retest reliability of an \u0026lt;\u0026thinsp;scp\u0026gt;fMRI paradigm for studies of cardiovascular reactivity. \u003cem\u003ePsychophysiology\u003c/em\u003e \u003cb\u003e49\u003c/b\u003e, 873\u0026ndash;884 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang, C. et al. Diurnal Variations in Neural Activity of Healthy Human Brain Decoded with Resting-State Blood Oxygen Level Dependent fMRI. \u003cem\u003eFront. Hum. Neurosci.\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e, 634 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZou, Q. et al. Reliability comparison of spontaneous brain activities between BOLD and CBF contrasts in eyes-open and eyes-closed resting states. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e121\u003c/b\u003e, 91\u0026ndash;105 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMu\u0026ntilde;oz-Rivas, M. J. M. D. P. Validaci\u0026oacute;n de la versi\u0026oacute;n modificada de la Conflicts Tactics Scale (M-CTS) en poblaci\u0026oacute;n juvenil espa\u0026ntilde;ola. \u003cem\u003ePsicothema\u003c/em\u003e \u003cb\u003e19\u003c/b\u003e, 693\u0026ndash;698 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcAuliffe, M. D., Hubbard, J. A., Rubin, R. M., Morrow, M. T. \u0026amp; Dearing, K. F. Reactive and Proactive Aggression: Stability of Constructs and Relations to Correlates. \u003cem\u003eJ. Genet. Psychol.\u003c/em\u003e \u003cb\u003e167\u003c/b\u003e, 365\u0026ndash;382 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrugman, S. et al. Examining the reactive proactive questionnaire in adults in forensic and non-forensic settings: A variable‐ and person‐based approach. \u003cem\u003eAggress. Behav.\u003c/em\u003e \u003cb\u003e43\u003c/b\u003e, 155\u0026ndash;162 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNieto-Castanon, A. \u003cem\u003eHandbook of Functional Connectivity Magnetic Resonance Imaging Methods in CONN\u003c/em\u003e (Hilbert, 2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.56441/hilbertpress.2207.6598\u003c/span\u003e\u003cspan address=\"10.56441/hilbertpress.2207.6598\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWhitfield-Gabrieli, S., Nieto-Castanon, A. \u0026amp; Conn A Functional Connectivity Toolbox for Correlated and Anticorrelated Brain Networks. \u003cem\u003eBrain Connect.\u003c/em\u003e \u003cb\u003e2\u003c/b\u003e, 125\u0026ndash;141 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSPM12 -Statistical Parametric Mapping. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.fil.ion.ucl.ac.uk/spm/software/spm12/\u003c/span\u003e\u003cspan address=\"https://www.fil.ion.ucl.ac.uk/spm/software/spm12/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndersson, J. L. R., Hutton, C., Ashburner, J., Turner, R. \u0026amp; Friston, K. Modeling Geometric Deformations in EPI Time Series. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e, 903\u0026ndash;919 (2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFriston, K. J. et al. Spatial registration and normalization of images. \u003cem\u003eHum. Brain Mapp.\u003c/em\u003e \u003cb\u003e3\u003c/b\u003e, 165\u0026ndash;189 (1995).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSladky, R. et al. Slice-timing effects and their correction in functional MRI. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e58\u003c/b\u003e, 588\u0026ndash;594 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWhitfield-Gabrieli, S., Nieto-Castanon, A. \u0026amp; Ghosh, S. Artifact detection tools (ART). \u003cem\u003eCambridge, MA\u003c/em\u003e: (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNieto-Castanon, A. Preparing fMRI Data for Statistical Analysis. (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePower, J. D. et al. Methods to detect, characterize, and remove motion artifact in resting state fMRI. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e84\u003c/b\u003e, 320\u0026ndash;341 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCalhoun, V. D. et al. The impact of T1 versus EPI spatial normalization templates for fMRI data analyses. \u003cem\u003eHum. Brain Mapp.\u003c/em\u003e \u003cb\u003e38\u003c/b\u003e, 5331\u0026ndash;5342 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAshburner, J. A fast diffeomorphic image registration algorithm. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e38\u003c/b\u003e, 95\u0026ndash;113 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAshburner, J. \u0026amp; Friston, K. J. Unified segmentation. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e26\u003c/b\u003e, 839\u0026ndash;851 (2005).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFriston, K. J., Williams, S., Howard, R., Frackowiak, R. S. J. \u0026amp; Turner, R. Movement-Related effects in fMRI time‐series. \u003cem\u003eMagn. Reson. Med.\u003c/em\u003e \u003cb\u003e35\u003c/b\u003e, 346\u0026ndash;355 (1996).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHallquist, M. N., Hwang, K. \u0026amp; Luna, B. The nuisance of nuisance regression: Spectral misspecification in a common approach to resting-state fMRI preprocessing reintroduces noise and obscures functional connectivity. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e82\u003c/b\u003e, 208\u0026ndash;225 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChai, X. J., Casta\u0026ntilde;\u0026oacute;n, A. N., \u0026Ouml;ng\u0026uuml;r, D. \u0026amp; Whitfield-Gabrieli, S. Anticorrelations in resting state networks without global signal regression. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e59\u003c/b\u003e, 1420\u0026ndash;1428 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBehzadi, Y., Restom, K., Liau, J. \u0026amp; Liu, T. T. A component based noise correction method (CompCor) for BOLD and perfusion based fMRI. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e37\u003c/b\u003e, 90\u0026ndash;101 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShrout, P. E. \u0026amp; Fleiss, J. L. Intraclass correlations: Uses in assessing rater reliability. \u003cem\u003ePsychol. Bull.\u003c/em\u003e \u003cb\u003e86\u003c/b\u003e, 420\u0026ndash;428 (1979).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLegendre, P. Species associations: the Kendall coefficient of concordance revisited. \u003cem\u003eJ. Agric. Biol. Environ. Stat.\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e, 226\u0026ndash;245 (2005).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLoan, C. F. V. The ubiquitous Kronecker product. \u003cem\u003eJ. Comput. Appl. Math.\u003c/em\u003e \u003cb\u003e123\u003c/b\u003e, 85\u0026ndash;100 (2000).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGower, J. C. \u0026amp; Dijksterhuis, G. B. \u003cem\u003eProcrustes Problems\u003c/em\u003e (Oxford University Press, 2004). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/acprof:oso/9780198510581.001.0001\u003c/span\u003e\u003cspan address=\"10.1093/acprof:oso/9780198510581.001.0001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang, H. et al. Amplitude of low frequency fluctuation within visual areas revealed by resting-state functional MRI. \u003cem\u003eNeuroimage\u003c/em\u003e \u003cb\u003e36\u003c/b\u003e, 144\u0026ndash;152 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGratton, C. et al. Functional Brain Networks Are Dominated by Stable Group and Individual Factors, Not Cognitive or Daily Variation. \u003cem\u003eNeuron\u003c/em\u003e \u003cb\u003e98\u003c/b\u003e, 439\u0026ndash;452e5 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBakdash, J. Z. \u0026amp; Marusich, L. R. Repeated Measures Correlation. \u003cem\u003eFront Psychol\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e, (2017).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003e\u003cstrong\u003eTable 1. Participant Demographic and Clinical Characteristics\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eVariable\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e\u003cstrong\u003en = 34\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAge (years)\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003eMean (SD)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e23.00 (4.44)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003eMedian\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e21\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003eRange\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e19-39\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSex\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003eMale\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e34 (100%)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMarital Status\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003eSingle\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e34 (100%)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eEducation Level\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003eUniversity\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e34 (100%)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRPQ Reactive\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003eMean (SD)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e6.59 (2.96)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003eMedian\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e7.0\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003eRange\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e0-13\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRPQ Proactive\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003eMean (SD)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e2.15 (1.89)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003eMedian\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e1.5\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003eRange\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 237px;\"\u003e\n \u003cp\u003e0-9\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eNote: All participants were male university students recruited from the general population. RPQ = Reactive-Proactive Aggression Questionnaire. Age outliers defined as \u0026gt;1.5 IQR above Q3 (\u0026gt;30 years). Main analyses reported for n=34; sensitivity analyses excluding age outliers (n=32) showed highly similar reliability patterns (see Supplementary Results).\u003c/em\u003e \u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable\u0026nbsp;2.\u0026nbsp;Test-retest\u0026nbsp;reliability of functional connectivity across brain networks.\u003c/strong\u003e \u0026nbsp;\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" style=\"width: 169px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eROIs\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 151px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAll time\u0026nbsp;points\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eBaseline vs. 1 Hour\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eBaseline vs. 1 Month\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 151px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e1 Hour vs. 1 Month\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 84px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNetwork\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRegion\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMean ICC\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e% Non-significant\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 59px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMean ICC\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 72px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e% Non-significant\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMean ICC\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e% Non-significant\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMean ICC\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e% Non-significant\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" style=\"width: 84px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCerebellar\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eAnterior\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.480\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e10%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.348\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e71%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.320\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e84%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.501\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e26%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003ePosterior\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.527\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e13%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.444\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e39%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.404\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e55%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.488\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e48%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"4\" style=\"width: 84px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDMN\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eLP (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.717\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e6%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.635\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e19%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.617\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e16%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.649\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e10%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eLP (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.654\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e10%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.601\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e19%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.557\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e29%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.609\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e13%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eMPFC\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.466\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e29%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.335\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e68%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.395\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e61%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.455\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e55%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003ePCC\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.531\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e12%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.512\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e32%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.466\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e52%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.440\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e55%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"4\" style=\"width: 84px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDAN\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eFEF (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.357\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e48%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.346\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e81%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.373\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e61%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.305\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e74%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eFEF (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.301\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e61%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.296\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e77%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.311\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e77%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.317\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e77%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eIPS (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.601\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e6%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.547\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e26%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.447\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e35%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.517\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e26%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eIPS (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.615\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e3%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.570\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e19%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.504\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e32%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.504\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e29%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"4\" style=\"width: 84px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eFPN\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eLPFC (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.510\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e23%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.435\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e48%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.427\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e42%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.443\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e45%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eLPFC (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.569\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e13%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.464\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e42%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.470\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e42%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.513\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e32%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003ePPC (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.599\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e13%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.507\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e35%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.516\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e42%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.518\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e35%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003ePPC (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.589\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e16%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.487\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e42%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.515\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e26%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.498\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e26%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"4\" style=\"width: 84px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eLN\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eIFG (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.578\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e16%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.477\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e45%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.566\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e26%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.525\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e39%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eIFG (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.628\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e16%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.528\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e29%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.567\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e23%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.542\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e23%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003epSTG (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.624\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e14%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.553\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e26%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.569\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e23%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.512\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e35%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003epSTG (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.573\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e10%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.506\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e35%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.526\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e32%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.535\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e32%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"7\" style=\"width: 84px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSN\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eACC\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.605\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e10%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.511\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e29%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.498\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e35%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.531\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e35%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eaInsula (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.601\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e23%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.479\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e39%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.467\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e42%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.592\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e26%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eaInsula (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.601\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e20%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.486\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e35%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.504\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e39%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.565\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e29%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eRPFC (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.590\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e10%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.564\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e23%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.476\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e35%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.490\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e32%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eRPFC (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.636\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e10%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.592\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e19%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.558\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e32%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.549\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e29%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eSMG (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.699\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e13%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.616\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e19%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.643\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e13%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.657\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e10%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eSMG (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.685\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e10%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.637\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e16%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.568\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e23%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.615\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e19%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"3\" style=\"width: 84px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSMN\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eLateral (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.499\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e29%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.403\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e55%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.440\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e45%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.398\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e52%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eLateral (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.561\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e10%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.441\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e55%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.459\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e45%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.473\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e42%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eSuperior\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.537\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e19%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.383\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e55%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.497\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e35%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.475\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e26%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"4\" style=\"width: 84px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eVN\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eLateral (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.526\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e19%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.491\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e42%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.398\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e65%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.448\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e52%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eLateral (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.471\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e29%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.462\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e42%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.389\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e58%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.382\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e58%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eMedial\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.505\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e19%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.431\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e45%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.400\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e61%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.431\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e52%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 85px;\"\u003e\n \u003cp\u003eOccipital\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.382\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e42%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 59px;\"\u003e\n \u003cp\u003e.412*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 72px;\"\u003e\n \u003cp\u003e55%\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e.364\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 71px;\"\u003e\n \u003cp\u003e61%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e.312\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 75px;\"\u003e\n \u003cp\u003e84%*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" style=\"width: 169px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTotal ROIs\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 77px;\"\u003e\n \u003cp\u003e.557\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 73px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 59px;\"\u003e\n \u003cp\u003e.484\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 72px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 60px;\"\u003e\n \u003cp\u003e.475\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 75px;\"\u003e\n \u003cp\u003e.493\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 75px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eNote. Mean intraclass correlation coefficients (ICC) and percentage of non-significant connections for each functional network region across different temporal comparisons. Asterisks (*) show regions with more than 55% non-significant connections or\u0026nbsp;\u003c/em\u003e|\u003cem\u003er| \u0026lt; 0.5,\u003c/em\u003e\u003cem\u003e\u0026nbsp;classified as highly unstable. Abbreviations: DMN = Default Mode Network; DAN = Dorsal Attention Network; FPN = Frontoparietal Network; LN = Language Network; SN = Salience Network; SMN = Sensorimotor Network; VN = Visual Network; LP = Lateral Parietal; MPFC = Medial Prefrontal Cortex; PCC = Posterior Cingulate Cortex; FEF = Frontal Eye Fields; IPS = Intraparietal Sulcus; LPFC = Lateral Prefrontal Cortex; PPC = Posterior Parietal Cortex; IFG = Inferior Frontal Gyrus; pSTG = posterior Superior Temporal Gyrus; ACC = Anterior Cingulate Cortex; RPFC = Rostral Prefrontal Cortex; SMG = Supramarginal Gyrus; L = Left; R = Right.\u003c/em\u003e \u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable\u0026nbsp;3. Stable connectivity patterns associated with aggression dimensions\u0026nbsp;(RPQ\u0026nbsp;proactive\u0026nbsp;and\u0026nbsp;reactive).\u003c/strong\u003e \u0026nbsp;\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRegion 1\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRegion 2\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePROACTIVE\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eREACTIVE\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eDMN LP (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eSN RPFC (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e-0.441*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e-0.467*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" style=\"width: 148px;\"\u003e\n \u003cp\u003eDMN LP (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eFPN LPFC (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e-0.467*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e-0.212\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eSMN Superior\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e-0.616*\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e-0.554*\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"3\" style=\"width: 148px;\"\u003e\n \u003cp\u003eDMN PCC\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eSN RPFC (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e-0.518*\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e-0.072\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eSN RPFC (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e0.496*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e0.454*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eSN SMG (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e-0.512*\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e-0.396\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"6\" style=\"width: 148px;\"\u003e\n \u003cp\u003eDAN IPS (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eFPN LPFC (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e-0.127*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e-0.053\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eLN IFG (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e-0.467*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e-0.231*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eLN pSTG (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e0.186*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e0.345\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eFPN LPFC (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e0.391*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e0.040\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eLN IFG (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e0.227*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e0.493*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eLN pSTG (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e-0.435*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e-0.344\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eFPN LPFC (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eLN pSTG (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e-0.505*\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e-0.502*\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eFPN PPC (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eLN pSTG (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e-0.525*\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e-0.545*\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eSN SMG (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eDAN IPS (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e0.498*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e0.471*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" style=\"width: 148px;\"\u003e\n \u003cp\u003eSMN Lateral (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eSN ACC\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e0.383*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e0.459*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eSN RPFC (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e-0.292*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e-0.440*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eSMN Lateral (R)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eDAN IPS (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e-0.410*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e-0.445*\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" style=\"width: 148px;\"\u003e\n \u003cp\u003eSMN Superior\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eFPN LPFC (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.538*\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.522*\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003eLN pSTG (L)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.501*\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 148px;\"\u003e\n \u003cp\u003e0.407\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eNote: Values represent Pearson correlation coefficients. * Indicates statistically significant correlations (p-FDR \u0026lt; 0.005) after false discovery rate correction for multiple comparisons (253 comparisons). Bold values indicate both statistical significance (p-FDR \u0026lt; 0.005) and strong correlations (|r| \u0026gt; 0.5). Region 1 labels are shown only at the beginning of each grouping. DMN = Default Mode Network; DAN = Dorsal Attention Network; FPN = Frontoparietal Network; LN = Language Network; SN = Salience Network; SMN = Sensorimotor Network; LP = Lateral Parietal; PCC = Posterior Cingulate Cortex; IPS = Intraparietal Sulcus; LPFC = Lateral Prefrontal Cortex; PPC = Posterior Parietal Cortex; IFG = Inferior Frontal Gyrus; pSTG = Posterior Superior Temporal Gyrus; ACC = Anterior Cingulate Cortex; AIns = Anterior Insula; RPFC = Rostral Prefrontal Cortex; SMG = Supramarginal Gyrus; L = Left; R = Right. \u0026nbsp;\u003c/em\u003e\u0026nbsp;\u003c/p\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":"aggression, brain networks, resting-state functional connectivity, test-retest reliability","lastPublishedDoi":"10.21203/rs.3.rs-9290839/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9290839/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe temporal stability of resting-state functional MR connectivity (rsfMRc) is a fundamental methodological consideration for longitudinal neuroimaging research, yet systematic evaluations across clinically relevant time frames remain limited. This study examined test-retest reliability of rsfMRc over short-term (1 hour) and medium-term (1 month) intervals in 34 healthy young adult males across three sessions. Intraclass correlation coefficients (ICC) and matrix-wide reliability metrics were calculated to evaluate temporal stability across major brain networks, and associations between stable connectivity patterns and aggression measures were examined. Overall reliability showed moderate values (mean ICC = .557) with significant network-specific variation. Salience Network regions, particularly the supramarginal gyrus, and lateral parietal regions of the Default Mode Network demonstrated the highest stability (ICC \u0026gt; .600). Matrix-wide analyses indicated high consistency in connectivity rankings (Kendall's W = .798). Among the 23 most temporally stable regions, several showed systematic associations with both reactive and proactive aggression. These findings reveal substantial heterogeneity in rsfMRc reliability across brain networks and measurement intervals, providing empirical guidance for researchers planning longitudinal connectivity studies. Network-specific considerations are critical in study design, and identifying stable connectivity patterns offers a foundation for biomarker development, while regions with poor reliability may require alternative acquisition strategies.\u003c/p\u003e","manuscriptTitle":"Test-retest reliability of resting-state functional connectivity: A methodological investigation with implications for aggression research","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-20 15:49:07","doi":"10.21203/rs.3.rs-9290839/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2026-04-28T10:02:37+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-22T18:56:17+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-19T13:32:05+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"3602785171282142893737026863569466548","date":"2026-04-15T07:32:40+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"162191710796141063943345343373302876444","date":"2026-04-13T01:41:06+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"115175300672257839096749806639614911477","date":"2026-04-10T09:47:15+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"65152237532001284288459609798606543716","date":"2026-04-10T09:45:55+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-10T09:31:10+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-10T09:27:35+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-04-08T11:51:18+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-07T23:00:09+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-04-07T22:53:52+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":"30a8193c-6f9f-454a-b3ef-fdf20b520446","owner":[],"postedDate":"April 20th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":66598162,"name":"Health sciences/Biomarkers"},{"id":66598163,"name":"Health sciences/Medical research"},{"id":66598164,"name":"Health sciences/Neurology"},{"id":66598165,"name":"Biological sciences/Neuroscience"}],"tags":[],"updatedAt":"2026-04-20T15:49:07+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-20 15:49:07","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9290839","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9290839","identity":"rs-9290839","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}