Measurement Invariance of the Child Behavior Checklist (CBCL) Across Gender in a Chilean Child Cohort

Measurement Invariance of the Child Behavior Checklist (CBCL) Across Gender in a Chilean Child Cohort | 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 Research Article Measurement Invariance of the Child Behavior Checklist (CBCL) Across Gender in a Chilean Child Cohort Ricardo Mellado This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9431030/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Measurement invariance of the Child Behavior Checklist (CBCL) was examined across gender in a nationally representative sample of Chilean children. Using data from the Chilean Longitudinal Survey of Early Childhood (ELPI; n > 10,000), equivalence of the internalising and externalising domains across boys and girls was evaluated for both the preschool (1.5–5 years) and school-age (6–12 years) versions. Multigroup confirmatory factor analyses with age controlled as a covariate supported full configural, metric, scalar, and strict invariance across gender in both versions. Two sensitivity analyses—an alternative parcelling scheme and item-level invariance testing within each syndrome subscale—confirmed the robustness of these findings. These results indicate that gender comparisons in CBCL scores at each developmental stage are not confounded by differential instrument functioning, providing psychometric support for gender-based comparisons in CBCL research, including longitudinal analyses that rely on valid within-wave gender comparisons, in Latin American settings. Psychology Child Behavior Checklist (CBCL) measurement invariance confirmatory factor analysis gender differences early childhood Chile Figures Figure 1 Introduction Social-emotional development forms a critical foundation for a child's future well-being. Emotional and behavioural symptoms in childhood—such as internalising problems (e.g., anxiety, withdrawal) and externalising behaviours (e.g., aggression, hyperactivity)—are consistently linked to adverse outcomes later in life, including poverty, mental health disorders, and social exclusion (Dekker et al., 2007; Fergusson et al., 2005). Identifying these difficulties early is crucial for guiding timely interventions. This requires tools that can effectively monitor mental health from early childhood and distinguish typical patterns from those that may signal risk. However, efforts to identify and compare patterns of emotional and behavioural problems must also consider how symptoms may be interpreted or reported differently across demographic groups, particularly by gender. Research consistently documents gender differences in symptom prevalence: boys exhibit higher rates of externalising behaviours, while girls show elevated internalising symptoms (Hankin & Abramson, 2001; Rosenfield & Mouzon, 2013; Sterba et al., 2007). Yet observed gender differences may reflect not only true variation in symptom levels, but also systematic differences in how behaviours are interpreted and reported for boys versus girls — a possibility with direct implications for the validity of gender-based comparisons in developmental research. Empirical evidence indicates that parental reports of child behaviour often show systematic gender-based patterns even when children’s self-reports do not: mothers tend to attribute more internalising symptoms to daughters and more externalising symptoms to sons beyond what children report about themselves (Najman et al., 2001). These reporting differences may arise from several mechanisms. First, cultural gender stereotypes may influence interpretation thresholds: identical aggressive behaviours may be normalized in boys (“boys will be boys”) but interpreted as atypical in girls, potentially leading caregivers to underreport externalising behaviours in boys while perceiving similar behaviours in girls as more problematic. Second, symptom manifestations may differ by gender: boys often express externalising problems through physical aggression, whereas girls may demonstrate relational aggression (e.g., social exclusion or rumour-spreading), which may not map equivalently onto CBCL items emphasizing physical behaviours (Crick & Grotpeter, 1995). Third, internalising symptoms in boys may be underrecognized due to masculine socialization discouraging emotional expression, potentially inflating apparent gender differences (Chaplin & Aldao, 2013). In Latin American contexts, where traditional gender role expectations may remain relatively salient, these processes could plausibly be amplified—yet this possibility has received little empirical attention. These concerns raise a critical methodological question: does the CBCL measure internalising and externalising constructs equivalently for boys and girls? Establishing measurement equivalence is essential because developmental researchers routinely use the CBCL to compare symptom levels across gender, identify gender-specific risk profiles, examine whether developmental processes differ for boys and girls, and evaluate whether interventions operate similarly across groups (Achenbach et al., 2016) Such applications assume that the instrument functions equivalently across gender. Measurement invariance testing provides a framework for evaluating this assumption through a sequence of increasingly stringent model constraints. If configural invariance holds, the same factor structure — with internalising and externalising as distinct dimensions defined by the same pattern of syndrome indicators — applies to both boys and girls. If metric invariance holds, each syndrome subscale contributes to its respective factor with equal strength across gender, meaning that a unit change in the latent factor produces the same expected change in observed scores for boys and girls. If scalar invariance holds, boys and girls with the same latent symptom level would obtain the same expected observed score, enabling valid comparisons of group means (Putnick & Bornstein, 2016). Without scalar invariance, observed gender differences may reflect measurement artefacts rather than true psychological variation. This issue is particularly relevant given the CBCL’s widespread use in large-scale research examining gender differences, including national cohort studies and epidemiological surveillance systems. In Chile, the CBCL is integrated into the nationally representative Chilean Longitudinal Survey of Early Childhood (ELPI), where it has been used to examine developmental trajectories of internalising and externalising symptoms (Morales et al., 2024), gender-specific longitudinal processes linking behavioural problems and language development (Mellado, 2025), and early gender differences in socio-emotional outcomes within broader analyses of skill formation and inequality (Behrman et al., 2017). These analyses involve comparing regression coefficients and latent means across gender, which require metric and scalar invariance respectively. Prior Research on Measurement Invariance of the CBCL Measurement invariance of the CBCL has been examined across several types of groups. Cross-national studies have shown that the instrument’s syndrome structure is broadly stable across countries (Ivanova et al., 2007, 2010), and additional work has evaluated invariance across clinical populations and racial or ethnic groups with generally supportive findings (Pandolfi et al., 2009; Yarnell et al., 2013; Stewart et al., 2024). Research on gender-based measurement invariance remains comparatively limited and has focused primarily on the school-age version of the CBCL. Existing studies include analyses in a population-based sample in Mauritius (Yarnell et al., 2013), a recent population-based study in the United States (Stewart et al., 2024), and a clinical sample of children referred for psychiatric evaluation (Sluis et al., 2017). Across these studies, findings generally support configural and metric invariance, with scalar invariance also supported in some cases, suggesting broadly comparable measurement across boys and girls in the school-age CBCL. Despite these contributions, several key gaps remain. First, existing CBCL gender measurement invariance studies have been conducted almost exclusively in high-income countries or European clinical settings, with the exception of Yarnell et al. (2013) in Mauritius. This limits generalizability to contexts where gender socialization norms and parental reporting patterns may differ. No study has examined gender measurement invariance of the CBCL in a large population-based sample from a Latin American context, despite the increasing use of the instrument in such settings. Second, gender-based invariance has not been formally tested in the preschool version of the CBCL, despite early childhood being a developmental stage when gender-typed behavioural expectations are actively forming and parental interpretation of ambiguous behaviours may be especially susceptible to gender bias (Chaplin & Aldao, 2013). Third, measurement invariance studies of child and adolescent mental health instruments frequently test only configural or metric invariance and do not proceed to scalar invariance — the level required for valid comparisons of mean symptom levels across groups (Putnick & Bornstein, 2016). This constrains the conclusions that can be drawn from reported gender differences and underscores the need for systematic evaluation of measurement equivalence in large population-based samples. This study This study tests whether the CBCL measures internalising and externalising problems equivalently across gender in a large, nationally representative Chilean sample. Using data from the Chilean Longitudinal Survey of Early Childhood (n > 10,000), measurement invariance is evaluated across both CBCL versions (preschool 1.5–5 years; school-age 6–18 years). This study is important for three main reasons. First, Chile provides a valuable test case as a middle–upper-income Latin American country where the CBCL is widely used in both research and national policy, yet where cross-group measurement equivalence has not been examined. Patterns of gender socialization and parental reporting in this context may not directly align with those observed in the high-income settings where most existing evidence has been generated. Second, measurement invariance is examined across both CBCL versions using data from the same nationally representative cohort at two developmental stages. Specifically, children in the ELPI cohort were assessed using the preschool version of the CBCL in early childhood (wave 1) and the school-age version in later childhood (wave 3). This design allows for a consistent evaluation of measurement equivalence across the critical transition from early to middle childhood within a single population-based sample. Most existing gender invariance studies focus exclusively on the school-age version (e.g., Stewart et al., 2024; van der Sluis et al., 2017; Yarnell et al., 2013), and formal testing in the preschool version has not been reported. Examining both versions within the same cohort provides more comprehensive evidence on how the CBCL functions across developmental stages. Third, this study provides the necessary psychometric foundation for interpreting gender comparisons in CBCL-based longitudinal research within this cohort. Recent studies using the ELPI have examined gender-specific developmental trajectories of internalising and externalising symptoms (Morales et al., 2024) and tested whether cross-lagged associations between behavioural problems and language development differ for boys and girls (Mellado, 2025). These analyses compare regression coefficients and mean levels across gender — comparisons that require metric and scalar invariance, respectively. In addition, strict invariance was evaluated as a more stringent test of measurement equivalence. By establishing whether these conditions hold, the present study provides empirical grounding for gender comparisons that are routinely conducted in ELPI-based research but have not been formally validated. While cross-sectional measurement invariance does not substitute for longitudinal invariance testing — which is not feasible in this cohort due to the transition from the preschool to the school-age CBCL version across waves (2010, 2012, 2017) — it establishes a necessary condition: that gender comparisons at each measurement occasion are not confounded by differential instrument functioning. Beyond addressing these specific gaps, this study contributes to the broader cross-cultural validation of the CBCL in Latin America. To date, validation work in the region has focused primarily on evaluating the instrument's factor structure and psychometric properties (Lecannelier et al., 2014 in Chile; Viola et al., 2011 in Uruguay). The present study extends this literature by explicitly testing whether constructs derived from an instrument developed in Anglophone contexts retain equivalent meaning across demographic subgroups within a Latin American society. Ethical Approval Ethical approval for the original ELPI data collection was granted by the Ethics Committee of the Microdata Centre of the University of Chile. Written informed consent was obtained from parents or legal guardians. The present study involved secondary analysis of anonymized data and did not require additional ethical approval. Methods Data This study uses data from the Chilean Longitudinal Survey of Early Childhood (ELPI), a nationally representative dataset. ELPI collects information through face-to-face interviews, including a socio-demographic questionnaire administered to mothers and a set of assessments measuring cognitive, socio-emotional, and anthropometric development in children and their mothers. The first wave, conducted in 2010, included a nationally representative sample of approximately 15,000 children aged 6 months to 5 years, drawn from official birth records for children born between 1 January 2006 and 31 August 2009. The second wave, fielded in 2012, followed the original sample and added approximately 3,000 children born between 1 September 2009 and 31 December 2011. The third wave, conducted in 2017, included participants from earlier waves and introduced approximately 5,000 additional children born between 1 January 2012 and 31 December 2016, covering ages 6 months to 12 years. ELPI uses a two-stage cluster-stratified sampling design based on birth records from the Chilean Civil Registry (Behrman et al., 2010). In the first stage, municipalities served as the primary sampling units. Eighty-three municipalities representing the largest urban areas and approximately 74% of the national population were selected with certainty, while the remaining municipalities were grouped into clusters defined by region, socioeconomic composition, and population size, from which one municipality per cluster was randomly selected. In the second stage, children were randomly sampled within selected municipalities in proportion to their population size. In wave 1, 11,231 children aged 18 to 59 months constituted the target sample for assessment with the preschool version of the CBCL, of whom 11,193 completed the measure (99.7%). The wave 1 sample had a balanced distribution by sex assigned at birth, with most children aged 2 or 3 years. Throughout this paper, gender refers to the binary classification (male/female) based on sex assigned at birth as recorded in ELPI administrative records. In wave 3, 11,658 children aged 72 months or older constituted the target sample for the school-age version of the CBCL, and 11,633 completed the assessment (99.8%). This sample was also evenly distributed by gender, with ages ranging from 6 to 12 years and the largest subgroups between 9 and 11 years. Approximately 63% of the wave 3 sample (7,021 children) had also participated in wave 1. Descriptive characteristics of the analytic samples are presented in Table 1 Analyses were restricted to children with valid information on age and CBCL items. Missing data on CBCL items were low: among respondents who completed the CBCL, 34 cases (0.3%) in the preschool sample and 87 cases (0.7%) in the school-age sample had at least one missing item. After excluding cases with missing age and those with missing CBCL items, the final analytic sample consisted of 10,987 children for the preschool CBCL models and 11,626 children for the school-age CBCL models. Given the low proportion of missingness, analyses were conducted using complete cases. Table 1: Sample Characteristics by Wave (ELPI 2010 and 2017; analytic sample) Variables Wave 1 (2010) Wave 3 (2017) Child Characteristics Sample Size 10,987 11,626 Sex Assigned at Birth Male 5,472 (49.8%) 5,917 (50.9%) Female 5,515 (50.2%) 5,709 (49.1%) Age Distribution 1 year 1,923 (17.5%) – 2 years 3,799 (34.6%) – 3 years 3,650 (33.2%) – 4 years 1,615 (14.7%) – 6 years – 801 (6.9%) 7 years – 850 (7.3%) 8 years – 1,486 (12.8%) 9 years – 2,658 (22.9%) 10 years – 2,520 (21.7%) 11 years – 2,584 (22.2%) 12 years – 727 (6.3%) Maternal Characteristics Age of Mother < 25 years 3,177 (29%) 257 (2.3%) 25–34 years 4,872 (44.5%) 4,640 (41.7%) 35+ years 2,905 (26.5%) 6,231 (56.0%) Employment Status Worked last week 4,769 (43.5%) 6,418 (57.7%) Did not work 6,185 (56.4%) 4,710 (42.3%) Education Level Primary 1,932 (17.9%) 1,716 (15.4%) Secondary 6,350 (58.8%) 6,185 (55.6%) Vocational 1,286 (11.9%) 1,621 (14.6%) University 1,237 (11.4%) 1,600 (14.4%) Note. Percentages are based on available observations. Analytic sample sizes reflect children with complete age and CBCL data; maternal characteristics may have smaller denominators due to missing data. Maternal characteristics reflect expected demographic shifts across waves. In wave 1, most mothers were under 35 years old (72.9%), with just over half not in employment (56.4%) and educational attainment concentrated at the secondary level (58.7%). By wave 3, mothers were older (56.8% aged 35+), more frequently employed (57.7%), and slightly more educated. In the preschool sample (wave 1; N = 10,987), boys and girls exhibited similar internalising scores (boys: M = 58.8, SD = 9.48; girls: M = 59.4, SD = 9.45), while boys scored higher on externalising behaviour (boys: M = 60.1, SD = 10.6; girls: M = 58.3, SD = 10.3). In the school-age sample (wave 3; N = 11,626), boys showed slightly higher internalising scores than girls (boys: M = 53.8, SD = 10.7; girls: M = 51.9, SD = 11.0) and similar externalising scores (boys: M = 50.0, SD = 10.1; girls: M = 49.7, SD = 9.6). CBCL Measure Caregivers complete the CBCL by rating items on a 3-point scale: 0 (“Not true”), 1 (“Somewhat or sometimes true”), and 2 (“Very true or often true”). Administration of the CBCL typically takes approximately 20–25 minutes. Items assessing internalising and externalising symptoms are grouped into empirically derived syndrome subscales—six in the preschool version and five in the school-age version. In the preschool form, internalising behaviour is assessed through four subscales: emotional reactivity, anxious/depressed, withdrawn, and somatic complaints. Externalising behaviour is measured using two subscales: attention problems and aggressive behaviour. In the school-age version, internalising behaviour comprises anxious/depressed, withdrawn, and somatic complaints, while externalising behaviour is assessed through rule-breaking and aggressive behaviour. The preschool version includes 99 items, and the school-age version includes 112. Scoring focuses on syndrome-relevant items: 36 internalising and 24 externalising items in the preschool form, and 26 internalising and 25 externalising items in the school-age version. Higher scores indicate greater symptom severity. Analytical Strategy All analyses were conducted using Mplus version 8.11 (Muthén & Muthén, 2017) and the lavaan package in R (Rosseel, 2012). Sampling weights provided by ELPI were incorporated in all models to adjust for unequal probabilities of selection and to produce population-representative estimates. The primary objective of the study was to evaluate measurement invariance of the CBCL internalising and externalising constructs across gender. Before invariance testing could proceed, several measurement modelling decisions were required — specifically, the selection of an appropriate factorial structure and the choice between item-level and parcel-level indicators. Because these decisions depend on empirical considerations such as model fit, convergence, and subscale unidimensionality, preliminary analyses were conducted to inform model specification. Measurement invariance procedures and sensitivity analyses are then described to assess the robustness of the results. Model selection As an initial step, the factorial structure of the CBCL was evaluated separately for the preschool (1.5–5 years) and school-age (6–18 years) versions to identify an appropriate baseline measurement model for subsequent invariance testing. The baseline model must provide adequate fit and converge reliably across groups; if the theoretically preferred specification fails to meet these criteria, an alternative is required. Two item-level confirmatory factor analytic models were compared — a second-order hierarchical model and a first-order two-factor model — both estimated using the robust weighted least squares estimator (WLSMV), appropriate for ordered-categorical data (Flora & Curran, 2004). Items were modelled using their original 3-point response scale (0 = "Not true"; 1 = "Somewhat or sometimes true"; 2 = "Very true or often true"). The first specification was a second-order hierarchical model in which items loaded onto their respective syndrome factors (six in the preschool version and five in the school-age version), which in turn loaded onto higher-order internalising and externalising factors. This model represents the full theoretical structure of the CBCL and is the most faithful representation of the instrument’s design (Achenbach & Rescorla, 2001). The second specification was a first-order two-factor model in which items loaded directly onto internalising and externalising factors, bypassing the syndrome level (Konold & Pianta, 2004). Both models specify internalising and externalising behaviour as the core latent constructs — the dimensions whose measurement equivalence across gender is the focus of this study. Model fit was assessed using the Root Mean Square Error of Approximation (RMSEA), the Comparative Fit Index (CFI), and the Tucker-Lewis Index (TLI). RMSEA values ≤ .05 were considered indicative of good fit and values ≤ .08 of acceptable fit (Browne & Cudeck, 1992; Yu, 2002). For CFI and TLI, values ≥ .90 were considered indicative of good fit, with values between .80 and .90 interpreted as acceptable (Browne & Cudeck, 1992). The more stringent .95 threshold proposed by Hu and Bentler (1999) was not applied, as it has been criticized for over-rejecting correctly specified complex models with large numbers of indicators, as is the case for both CBCL versions (Marsh et al., 2004). Table 2: Fit indices for CFA models across CBCL broadband structures (preschool and school-age versions) Model SB 𝜒 2 (df) CFI TLI RMSEA CBCL 1.5–5 (Preschool version) 2-Factor Model 52843.24 (1709) 0.814 0.807 0.052 Second-Order 2-Factor Model 45461.83 (1703) 0.841 0.834 0.048 CBCL 6–18 (School-age version) 2-Factor Model 28705.68 (1709) 0.811 0.805 0.037 Second-Order 2-Factor Model 26650.01 (1704) 0.826 0.819 0.035 Note. SB𝜒 2 = Satorra–Bentler scaled chi-square; df = degrees of freedom; CFI = Comparative Fit Index; TLI = Tucker–Lewis Index; RMSEA = Root Mean Square Error of Approximation. As shown in Table 2, both specifications produced acceptable RMSEA values and CFI and TLI indices in the .80–.90 range, consistent with prior item-level CBCL factor analyses involving large numbers of ordinal indicators (Ivanova et al., 2010). The second-order model achieved marginally better fit across all indices in both versions (ΔCFI ≈ .02, ΔTLI ≈ .02, ΔRMSEA ≈ .004), however the second-order model failed to converge in both CBCL versions, producing Heywood cases with negative residual variances (see Table S1 in the Supplementary Materials for second-order factor loadings and inter-factor correlations for both model specifications). These issues arose from very high intercorrelations among syndrome factors, causing the second-order factor to over-explain first-order variance. In both versions, this produced standardized second-order loadings exceeding 1.0 (emotional reactivity λ = 1.073 in the preschool version; anxiety/depression λ = 1.024 in the school-age version), indicating Heywood cases. Similar convergence failures have been reported in prior studies applying second-order CFA models to the CBCL in population-based samples (Benninger et al., 2025; Ivanova et al., 2010), likely reflecting the substantial comorbidity across syndrome scales that produces these high inter-factor correlations. Given that the model failed to converge even in single-group analyses, it could not be carried forward into multigroup invariance testing, where estimation demands are substantially greater. The first-order two-factor model, which did not present convergence issues, was therefore retained as the basis for subsequent measurement modelling. However, rather than testing measurement invariance at the item level, established CBCL measurement invariance practice was followed (Konold & Pianta, 2004; Pandolfi et al., 2009), and syndrome subscale scores were used as s factor. Testing multigroup invariance at the item level with 60+ ordinal indicators requires the simultaneous estimation of a large number of parameters (loadings, thresholds, and residual variances for each item across groups), which increases the risk of estimation instability and model non-convergence—particularly with ordinal data, the WLSMV estimator, and the incorporation of sampling weights. Parcelling addresses this by reducing the number of estimated parameters and, through aggregation, attenuating item-specific measurement error (Little et al., 2013), as syndrome subscale scores average across item-level variance, yielding more reliable and approximately continuous indicators. This allowed the use of robust maximum likelihood estimation (MLR) for multigroup invariance testing. The validity of parcelling rests on the assumption that items within each syndrome subscale form approximately unidimensional constructs. If items within a parcel capture multiple underlying dimensions, the resulting parcel indicators may conflate distinct sources of variance, undermining their interpretation as indicators of a single latent factor. To evaluate this assumption, item-level confirmatory factor analyses were conducted separately for each syndrome subscale using the WLSMV estimator on the original ordinal items. Internal consistency was assessed using ordinal Cronbach’s alpha and McDonald's omega. Table 3 reports fit indices and reliability estimate for these models. Table 3: Fit indices and internal consistency estimates for one-factor CFA models of CBCL syndrome subscales (preschool and school-age versions) Version Syndrome SB𝜒 2 df CFI TLI RMSEA SRMR 𝜆̄ 𝜆̃ 𝛼𝑜𝜏𝑑 𝑚 CBCL 1.5–5 Emotional 567.14 27 0.955 0.940 0.051 0.051 0.544 0.524 0.789 0.691 Withdrawn 238.31 20 0.975 0.965 0.038 0.038 0.555 0.629 0.769 0.593 Anxious 989.73 20 0.901 0.862 0.076 0.074 0.557 0.531 0.765 0.638 Somatic 3642.26 44 0.734 0.667 0.091 0.123 0.513 0.543 0.765 0.581 Aggressive 6745.35 152 0.919 0.909 0.075 0.058 0.607 0.622 0.914 0.893 Attention 131.60 5 0.946 0.893 0.054 0.061 0.481 0.473 0.602 0.501 CBCL 6–18 Anxiety 1559.20 65 0.936 0.923 0.053 0.064 0.591 0.587 0.871 0.790 Withdrawn 437.58 20 0.964 0.949 0.053 0.050 0.667 0.678 0.862 0.739 Somatic 3.84 2 0.999 0.996 0.012 0.011 0.603 0.610 0.693 0.500 Rule-breaking 1185.65 119 0.868 0.849 0.032 0.091 0.687 0.672 0.934 0.706 Aggressive 4338.13 135 0.920 0.909 0.061 0.078 0.681 0.710 0.936 0.884 Note. SB𝜒 2 = Satorra–Bentler scaled chi-square; df = degrees of freedom; CFI = Comparative Fit Index; TLI = Tucker–Lewis Index; RMSEA = Root Mean Square Error of Approximation; SRMR = Standardized Root Mean Square Residual; 𝜆̄ = mean standardized factor loading; 𝜆̃ = median standardized factor loading; 𝛼𝑜𝜏𝑑 = ordinal Cronbach’s alpha; 𝜔 = McDonald’s omega. As shown in Table 3, most subscales demonstrated acceptable to good model fit. In the preschool version, CFI values ranged from .901 (Anxious/Depressed) to .975 (Withdrawn), with RMSEA values between .038 and .076. In the school-age version, fit indices were broadly comparable, with CFI ranging from .868 (Rule-Breaking) to .999 (Somatic) and RMSEA from .012 to .061. Internal consistency was adequate to strong across both versions, with ordinal alpha ranging from .60 to .91 and omega from .50 to .89 in the preschool version, and alpha from .69 to .94 and omega from .50 to .88 in the school-age version. Mean standardized factor loadings ranged from .48 to .61 in the preschool version and .59 to .69 in the school-age version. Two subscales showed comparatively poorer fit. The Somatic Complaints subscale in the preschool version exhibited the weakest unidimensional fit (CFI = .734, RMSEA = .091), with modification indices revealing several large residual correlations among items, indicating local dependencies and departures from unidimensionality (see Table S2 in the Supplementary Materials for the largest modification indices corresponding to residual correlations among somatic items). Despite this, the subscale demonstrated acceptable internal consistency (α = .77) and was retained given its theoretical centrality to the internalising domain. Similarly, the Rule-Breaking subscale in the school-age version showed borderline fit (CFI = .868, TLI = .849), though internal consistency was strong (α = .93). In both cases, aggregating items into subscale scores reduces item-specific measurement error and facilitates stable multigroup estimation. The performance of these subscales is revisited in the item-level sensitivity analyses. Distributional properties of the resulting parcels were examined to assess their suitability as continuous indicators. For the preschool version, skewness and kurtosis values were within acceptable ranges (skewness: -0.16 to 1.21; kurtosis: 2.43 to 4.87). For the school-age version, parcels showed moderate to pronounced positive skew, with Rule-Breaking exhibiting the most extreme values (skewness = 4.20, kurtosis = 44.8), consistent with low base rates of these behaviours in population samples (see Table S3 in the Supplementary Materials for parcel-level distributional properties). Robust maximum likelihood estimation (MLR) was used throughout, which provides corrected standard errors and fit statistics under non-normality (Yuan & Bentler, 2000). Measurement invariance testing Measurement invariance was evaluated using multigroup confirmatory factor analysis (CFA) across gender, estimated separately for the preschool and school CBCL versions. Models were estimated using robust maximum likelihood (MLR) on the parcel-level indicators. Following standard practice, a sequence of increasingly constrained models was estimated: configural invariance (equal factor structure), metric invariance (equal factor loadings), scalar invariance (equal intercepts), and strict invariance (equal residual variances). Given the wide age ranges within each version (approximately 1.5 to 5 years in the preschool sample and 6 to 12 years in the school-age sample), age in months was included as a covariate on the latent internalising and externalising factors in all models to ensure that gender comparisons were not confounded by developmental differences within each sample. Model comparisons relied on changes in approximate fit indices rather than chi-square difference testing, which is overly sensitive to sample size in large samples. Following Chen (2007), invariance was supported if decreases in CFI were less than .010 and increases in RMSEA were less than .015. Note that at the parcel level, scalar invariance involves constraining intercepts — not thresholds — as parcels are treated as continuous indicators. Sensitivity Analysis To evaluate the robustness of the primary findings, two complementary sensitivity analyses were conducted. These analyses address a key methodological limitation of parcelling: aggregating items into syndrome-level composites may mask item-level differences in how the CBCL functions across gender. Alternative parcelling scheme First, it was examined whether invariance conclusions depended on how items were aggregated into parcels. The primary analysis used one parcel per syndrome (six parcels in the preschool version and five in the school-age version). An alternative parcelling scheme divided syndromes that were either larger in size or showed marginal unidimensional fit (see Table 3) into two parcels each. For the preschool version, Anxious/Depressed, Somatic Complaints, and Aggressive Behaviour were split, while Emotional Reactivity, Withdrawn, and Attention Problems were retained as single parcels, producing a nine-parcel model. For the school-age version, Anxiety/Depression, Rule-Breaking, and Aggressive Behaviour were split, while Withdrawn and Somatic Complaints were retained, producing an eight-parcel model. The same invariance sequence (configural, metric, scalar, and strict) was applied using MLR with age included as a covariate. Consistent results across parcelling schemes would indicate that invariance conclusions are not dependent on the specific aggregation of items. Item-level invariance within syndromes Second, to assess whether parcelling masked item-level non-invariance, measurement invariance was tested separately within each syndrome subscale. Each syndrome was modelled as a unidimensional factor using the original three-point ordinal items and estimated using WLSMV. Configural, metric, and scalar invariance were evaluated across gender. At this level, scalar invariance involves constraining thresholds rather than intercepts, consistent with ordinal measurement models. This analysis provides a more granular assessment of measurement equivalence and allows identification of potential item-level gender differences that could be obscured when items are aggregated into parcels. Results Confirmatory Factor Analysis As described in the Analytical Strategy section, the first-order two-factor model was retained, and syndrome subscale scores were used as parcel-level indicators of the internalising and externalising factors. Parcel-level CFA models were estimated for both CBCL preschool and school-age versions using MLR. Standardized factor loadings were generally moderate to strong, ranging from .53 to .91 across parcels in both models. The correlation between internalising and externalising factors was high in both versions (r = .72 in the preschool model; r = .77 in the school-age model), indicating substantial shared variance between the constructs. Model fit was acceptable to excellent: preschool χ² (8) = 230.23, CFI = .988, TLI = .977, RMSEA = .050, SRMR = .020; school-age χ² (4) = 107.79, CFI = .991, TLI = .977, RMSEA = .047, SRMR = .015. These results support the adequacy of the parcel-level measurement models for subsequent invariance testing. Standardized factor loadings and residual variances for both models are shown in Figure 1. Measurement Invariance CBCL 1 and CBCL 2 by Gender Measurement invariance of the internalising and externalising factors was tested separately for the preschool and school-age versions of the CBCL across gender. For each version, a sequence of increasingly constrained models was estimated: configural (equal factor structure), metric (equal factor loadings), scalar (equal intercepts), and strict (equal residual variances), with age included as a covariate in all models. Table 4: Measurement invariance across gender (configural, metric, scalar, and strict models): Fit indices for CBCL preschool and school-age versions Model SB 𝜒 2 (df) CFI ( Δ CFI ) TLI ( Δ TLI) RMSEA ( Δ RMSEA) SRMR ( Δ SRMR) CBCL 1.5–5 (Preschool version) Configural 299 (24) 0.986 (–) 0.976 (–) 0.044 (–) 0.018 (–) Metric 308 (28) 0.986 (+0.000) 0.980 (+0.004) 0.041 (-0.003) 0.019 (+0.001) Scalar 410 (32) 0.981 (-0.005) 0.976 (-0.004) 0.045 (+0.004) 0.023 (+0.004) Strict 435 (38) 0.981 (+0.000) 0.979 (+0.003) 0.041 (-0.004) 0.026 (+0.003) CBCL 6–18 (School-age version) Configural 318 (14) 0.984 (–) 0.966 (–) 0.052 (–) 0.019 (–) Metric 376 (17) 0.986 (+0.002) 0.975 (+0.009) 0.045 (-0.007) 0.024 (+0.005) Scalar 428 (20) 0.983 (-0.003) 0.975 (+0.000) 0.045 (+0.000) 0.026 (+0.002) Strict 541 (25) 0.985 (+0.002) 0.982 (+0.007) 0.038 (-0.007) 0.034 (+0.008) Note. ΔCFI, ΔTLI, ΔRMSEA, and ΔSRMR represent changes relative to the preceding (less constrained) model. SB𝜒 2 = Satorra–Bentler scaled chi-square; df = degrees of freedom. Age (in months) was included as a covariate on the latent factors in all models. As shown in Table 4, fit indices for the preschool version remained stable across all four levels of invariance. Changes from configural to metric were negligible (ΔCFI = .000, ΔRMSEA = -.003), and the transition from metric to scalar produced only minor decrements (ΔCFI = -.005, ΔRMSEA = +.004). Strict invariance was also supported, with no change in CFI from scalar to strict (ΔCFI = .000, ΔRMSEA = -.004). All changes were well within recommended thresholds (Chen, 2007), indicating full configural, metric, scalar, and strict invariance of the preschool CBCL across gender. For the school-age version, a similar pattern emerged. Metric invariance was supported with minimal fit change (ΔCFI = +.002, ΔRMSEA = -.007), as was scalar invariance (ΔCFI = -.003, ΔRMSEA = +.001). At the strict level, CFI remained unchanged (ΔCFI = .000) and RMSEA improved (ΔRMSEA = -.007), though SRMR showed a marginal increase (ΔSRMR = .011, slightly above the .010 threshold). Given that CFI and RMSEA were clearly within acceptable limits, strict invariance is considered substantially supported for the school-age version. Together, these results indicate that the CBCL's internalising and externalising constructs are measured equivalently across gender in both versions, supporting valid comparisons of latent means, regression coefficients, and residual variances between boys and girls. Sensitivity Analysis Sensitivity Analysis 1: Alternative parcelling To assess whether invariance conclusions were sensitive to the specific parcelling scheme, the primary analyses were replicated using an alternative specification in which larger or marginally unidimensional syndromes were split into two parcels each. This produced a 9-parcel model for the preschool version and an 8-parcel model for the school-age version. Table 5 presents the results alongside the primary analysis for comparison. Table 5: Sensitivity Analysis: Measurement invariance results for primary and alternative parcelling schemes across gender Version Scheme Model CFI Δ CFI RMSEA Δ RMSEA CBCL 1.5–5 (Preschool) Primary (6 parcels) Configural 0.986 – 0.044 – Metric 0.986 +0.000 0.041 -0.003 Scalar 0.981 -0.005 0.045 +0.004 Strict 0.981 +0.000 0.041 -0.004 Alternative (10 parcels) Configural 0.934 – 0.076 – Metric 0.935 +0.001 0.072 -0.004 Scalar 0.930 -0.005 0.071 -0.001 Strict 0.930 +0.000 0.067 -0.004 CBCL 6–18 (School-age) Primary (5 parcels) Configural 0.984 – 0.052 – Metric 0.986 +0.002 0.045 -0.007 Scalar 0.983 -0.003 0.045 +0.000 Strict 0.985 +0.002 0.038 -0.007 Alternative (8 parcels) Configural 0.959 – 0.060 – Metric 0.966 +0.007 0.051 -0.009 Scalar 0.962 -0.004 0.052 +0.001 Strict 0.971 +0.009 0.043 -0.009 Note. ΔCFI and ΔRMSEA represent changes from the preceding (less constrained) model. Age (in years) was included as a covariate on the latent factors in all models. Primary parcelling used one parcel per syndrome; alternative parcelling split larger or marginally unidimensional syndromes into two parcels each. Despite lower baseline fit in the alternative models — expected given the increased number of indicators — the pattern of fit changes across invariance levels was consistent with the primary analysis. ΔCFI at the scalar level was -.005 for the preschool version and -.004 for the school-age version, both well within the .010 threshold and comparable to the primary parcelling schemes. Strict invariance was also supported in both versions, with ΔCFI values of .000 and +.009 respectively. All changes remained within recommended thresholds (Chen, 2007), indicating that measurement invariance conclusions are robust to the specific aggregation of items into parcels. Sensitivity Analysis 2: Item-Level Invariance Within Syndromes To assess whether parcelling masked potential item-level non-invariance, measurement invariance was tested separately within each syndrome using the original ordinal items and WLSMV estimation. Results are summarised in Table 6 Table 6: Item-level measurement invariance across gender for CBCL syndrome subscales (sensitivity analysis). Version Syndrome Items Config. CFI ΔCFIM ΔCFIS ΔCFISt CBCL 1.5–5 (Preschool) Emotional Reactivity 9 .956 .010 -.005 .000 Withdrawn 8 .974 .005 -.006 .000 Anxious/Depressed 8 .901 .014 a -.008 .000 Somatic Complaints 11 .734 .030 a -.009 .000 Aggressive Behaviour 19 .919 .026 a -.021 a .000 Attention Problems 5 .947 .007 -.007 .000 CBCL 6–18 (School-age) Anxiety/Depression 13 .936 .017 a -.007 .000 Withdrawn 8 .964 .008 -.002 .000 Somatic Complaints 4 .998 .001 -.006 .000 Rule-Breaking 17 .886 .034 a -.007 .000 Aggressive Behaviour 18 .923 .027 a -.016 a .000 Note. ΔCFIM, ΔCFIS, and ΔCFISt represent changes in CFI between configural–metric, metric–scalar, and scalar–strict models, respectively. Following Chen (2007), |ΔCFI| > .010 is flagged with a . Invariance tests are sequential; therefore scalar and strict results should be interpreted cautiously when metric invariance is not supported, as later steps assume equality constraints established at earlier stages. Results revealed a broadly consistent pattern across versions. In the preschool version, three of six syndromes (Emotional Reactivity, Withdrawn, and Attention Problems) showed fit changes within recommended thresholds across successive invariance steps. In the school-age version, Withdrawn and Somatic Complaints displayed a similar pattern. Several additional syndromes showed small deviations from the ΔCFI < .010 guideline at the metric stage, though subsequent scalar constraints resulted in minimal additional fit loss (ΔCFI ranging from −.002 to −.009), suggesting that loading differences were relatively small in magnitude. The Aggressive Behaviour subscale showed the greatest item-level complexity in both versions, with ΔCFI at the metric step exceeding .025 and scalar invariance not fully supported. This subscale also contains the largest number of items (19 and 18 respectively) and includes behaviours with closely related content, which may introduce local dependencies among items. Such patterns are consistent with item-level heterogeneity rather than systematic gender-related measurement differences in the broader construct. Importantly, at the parcel level, the Aggressive Behaviour subscale contributed to models achieving strict invariance across gender in both CBCL versions, suggesting that aggregation stabilised the measurement structure without introducing construct-level distortions. Discussion This study evaluated whether the CBCL measures internalising and externalising problems equivalently across gender in a large, nationally representative sample of Chilean children. Using data from the ELPI cohort, we tested measurement invariance of both the preschool (1.5–5 years) and school-age (6–12 years) versions, controlling for age-related developmental variation within each sample. Across both versions, the CBCL demonstrated full configural, metric, scalar, and strict invariance across gender at the parcel level, indicating that the instrument's measurement properties are equivalent for boys and girls. These findings were further supported by two sensitivity analyses — an alternative parcelling scheme and item-level invariance testing within each syndrome — which confirmed that the primary results were robust to analytical choices and not artifacts of item aggregation. These findings have direct practical implications. By establishing measurement invariance across gender, this study provides evidence that observed differences in CBCL scores between boys and girls are likely to reflect true variation in underlying internalising and externalising symptoms rather than artifacts of measurement bias. These results support the use of the CBCL for gender comparisons in Chilean population-based studies, including analyses of symptom prevalence, gender-moderated developmental pathways, and intervention effects. Beyond these immediate implications, the findings strengthen the psychometric foundation for the growing body of ELPI-based research that relies on gender comparisons. Studies using this cohort have examined gender-specific developmental trajectories of internalising and externalising symptoms (Morales et al., 2024) and tested whether cross-lagged associations between behavioural problems and language development differ for boys and girls (e.g., Mellado, 2025). Such analyses rely on metric invariance to validly compare regression coefficients across gender and scalar invariance to compare latent means. The present study therefore provides empirical support for these assumptions, strengthening the psychometric basis for existing and future ELPI-based analyses The findings also extend the existing psychometric literature on the CBCL in several ways. First, nearly all prior gender MI research on the CBCL has been conducted in high-income, English-speaking countries or European clinical settings. By demonstrating full invariance in a nationally representative Chilean sample, this study contributes to the cross-cultural validation of the CBCL in a middle-income, Spanish-speaking context where gender socialization norms and parental reporting patterns may differ from those in previously studied populations. Second, a notable gap is addressed by formally testing gender invariance in the preschool version of the CBCL, which has received limited empirical attention despite being widely used in developmental research. Third, the analytical approach — combining parcel-level multigroup CFA with both alternative parcelling and item-level sensitivity analyses — provides a more comprehensive evaluation of measurement equivalence than has been reported in prior CBCL invariance studies, most of which relied on a single analytical strategy without robustness checks. Some limitations should be acknowledged. First, the primary analysis used item parcelling, which can reduce sensitivity to item-level differences across groups (Meade & Kroustalis, 2006). However, this concern was directly addressed through complementary item-level invariance analyses, which showed that the majority of syndromes achieved invariance at the item level and that the one syndrome showing substantive complexity (Aggressive Behaviour) likely reflects item redundancy rather than systematic gender bias. Nevertheless, researchers interested in item-level differential functioning should consider dedicated DIF analyses as a complementary approach. Several broader limitations and future directions should be noted. First, the study relied exclusively on parent-report data, which may introduce reporting bias — a concern particularly relevant given the theoretical arguments outlined in the introduction regarding gender-differentiated parental interpretation of children's behaviour. Triangulating CBCL assessments with additional informants such as teachers would help disentangle measurement equivalence from informant-specific biases. Second, the present analysis assessed measurement invariance cross-sectionally within each CBCL version. Longitudinal invariance testing was not feasible in this cohort because children transitioned from the preschool to the school-age CBCL as they aged across ELPI waves (2010, 2012, 2017), meaning that different items were administered at different time points — a prerequisite violation for longitudinal MI. Future research using cohorts assessed with the same CBCL version across multiple waves could address this gap. Third, although the sample is nationally representative of Chile, generalizability to other cultural or regional contexts should not be assumed. Cross-national studies across Latin America could evaluate whether the CBCL demonstrates consistent measurement properties in populations with different gender socialization norms and reporting practices. Fourth, although age was controlled as a covariate on the latent factors, the possibility that the CBCL’s measurement properties vary across narrower age bands within each version was not examined. Future research could employ MIMIC models or age-subgroup analyses to evaluate age-related differential item functioning. Finally, the present study tested measurement invariance of the CBCL's established internalising-externalising structure. Alternative structural specifications — such as bifactor models or exploratory structural equation models — may provide additional insights into the dimensionality of the CBCL and could be evaluated in future work Conclusion The parent-reported CBCL demonstrated configural, metric, scalar and strict invariance across gender in a nationally representative Chilean sample, for both the preschool and school-age versions. These findings support the use of the CBCL for valid group comparisons of internalising and externalising problems across key social dimensions in early and middle childhood and strengthen its utility for developmental and policy-relevant research in Latin American settings. Declarations Funding The author acknowledges funding provided by the National Agency for Research and Development (ANID), BECAS/DOCTORADO BECAS CHILE/2021–72220327. References Abufhele, A., Contreras, D., Puentes, E., Telias, A., & Valdebenito, N. (2022). Socioeconomic gradients in child development: Evidence from a Chilean longitudinal study 2010–2017. Advances in Life Course Research, 52 , 100451. https://doi.org/10.1016/j.alcr.2021.100451 Achenbach, T. M. (2019). International findings with the Achenbach system of empirically based assessment (ASEBA): Applications to clinical services, research, and training. Child and Adolescent Psychiatry and Mental Health, 13 (1), 30. https://doi.org/10.1186/s13034-019-0291-2 Achenbach, T. M., Ivanova, M. Y., Rescorla, L. A., Turner, L. V., & Althoff, R. R. (2016). Internalizing/externalising problems: Review and recommendations for clinical and research applications. Journal of the American Academy of Child & Adolescent Psychiatry, 55 (8), 647–656. https://doi.org/10.1016/j.jaac.2016.05.012 Achenbach, T. M., & Rescorla, L. A. (2001). Manual for the ASEBA school-age forms & profiles: An integrated system of multi-informant assessment . University of Vermont, Research Center for Children, Youth, & Families. Behrman, J., Bravo, D., & Urzúa, S. (2010). Encuesta Longitudinal de la Primera Infancia: Aspectos metodológicos y primeros resultados. Retrieved from http://observatorio.ministeriodesarrollosocial.gob.cl/elpi-primera-ronda Behrman, J. R., Contreras, D., Palma, I., & Puentes, E. (2017). Wealth disparities for early childhood anthropometrics and skills: Evidence from Chilean longitudinal data (PIER Working Paper 17-019). Penn Institute for Economic Research, University of Pennsylvania. Benninger, T. L., Norris, M., & Witwer, A. N. (2025). Measurement invariance of the child behavior checklist in autistic toddlers. Research in Autism Spectrum Disorders, 119 , 102500. https://doi.org/10.1016/j.rasd.2024.102500 Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107 (2), 238–246. https://doi.org/10.1037/0033-2909.107.2.238 Browne, M. W., & Cudeck, R. (1992). Alternative ways of assessing model fit. Sociological Methods & Research, 21 (2), 230–258. https://doi.org/10.1177/0049124192021002005 Chaplin, T. M., & Aldao, A. (2013). Gender differences in emotion expression in children: A meta-analytic review. Psychological Bulletin, 139 (4), 735–765. https://doi.org/10.1037/a0030737 Chen, F. F. (2007). Sensitivity of goodness of fit indexes to lack of measurement invariance. Structural Equation Modeling: A Multidisciplinary Journal, 14 (3), 464–504. https://doi.org/10.1080/10705510701301834 Crick, N. R., & Grotpeter, J. K. (1995). Relational aggression, gender, and social-psychological adjustment. Child Development, 66 (3), 710–722. https://doi.org/10.2307/1131945 Dekker, M. C., Ferdinand, R. F., Van Lang, N. D., Bongers, I. L., Van Der Ende, J., & Verhulst, F. C. (2007). Developmental trajectories of depressive symptoms from early childhood to late adolescence: Gender differences and adult outcome. Journal of Child Psychology and Psychiatry, 48 (7), 657–666. https://doi.org/10.1111/j.1469-7610.2007.01742.x Fergusson, D. M., Horwood, L. J., & Ridder, E. M. (2005). Show me the child at seven: The consequences of conduct problems in childhood for psychosocial functioning in adulthood. Journal of Child Psychology and Psychiatry, 46 (8), 837–849. https://doi.org/10.1111/j.1469-7610.2005.01468.x Flora, D. B., & Curran, P. J. (2004). An empirical evaluation of alternative methods of estimation for confirmatory factor analysis with ordinal data. Psychological Methods, 9 (4), 466–491. https://doi.org/10.1037/1082-989X.9.4.466 Gross, D., Fogg, L., Young, M., Ridge, A., Cowell, J. M., Richardson, R., & Sivan, A. (2006). The equivalence of the child behavior checklist/1½–5 across parent race/ethnicity, income level, and language. Psychological Assessment, 18 (3), 313–323. https://doi.org/10.1037/1040-3590.18.3.313 Hankin, B. L., & Abramson, L. Y. (2001). Development of gender differences in depression: An elaborated cognitive vulnerability–transactional stress theory. Psychological Bulletin, 127 (6), 773–796. https://doi.org/10.1037/0033-2909.127.6.773 Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6 (1), 1–55. https://doi.org/10.1080/10705519909540118 Ivanova, M. Y., Achenbach, T. M., Dumenci, L., Rescorla, L. A., Almqvist, F., Weintraub, S., Bilenberg, N., Bird, H., Chen, W. J., Dobrean, A., Döpfner, M., Erol, N., Fombonne, E., Fonseca, A. C., Frigerio, A., Grietens, H., Hannesdóttir, H., Kanbayashi, Y., Lambert, M., ... Verhulst, F. C. (2007). Testing the 8-syndrome structure of the child behavior checklist in 30 societies. Journal of Clinical Child and Adolescent Psychology, 36 (3), 405–417. https://doi.org/10.1080/15374410701444363 Ivanova, M. Y., Achenbach, T. M., Rescorla, L. A., Harder, V. S., Ang, R. P., Bilenberg, N., Bjarnadottir, G., Capron, C., De Pauw, S. S., Dias, P., Dobrean, A., Döpfner, M., Erol, N., Fombonne, E., Frigerio, A., Hannesdottir, H., Kanbayashi, Y., Lambert, M. C., Larsson, B., ... Verhulst, F. C. (2010). Preschool psychopathology reported by parents in 23 societies: Testing the seven-syndrome model of the child behavior checklist for ages 1.5–5. Journal of the American Academy of Child & Adolescent Psychiatry, 49 (12), 1215–1224. https://doi.org/10.1016/j.jaac.2010.08.019 Kline, R. B. (2023). Principles and practice of structural equation modeling (5th ed.). Guilford Publications. Konold, T. R., Walthall, J. C., & Pianta, R. C. (2004). The behavior of child behavior ratings: Measurement structure of the child behavior checklist across time, informants, and child gender. Behavioral Disorders, 29 (4), 372–383. https://doi.org/10.1177/019874290402900404 Lecannelier, F., Pérez Ewert, J. C., Groissman, S., Gallardo, D., Bardet, A. M., Bascuñán, A., & Rodríguez, J. (2014). Validación del Inventario de Conductas Infantiles para niños de entre 1½–5 años (CBCL 1½–5) en la Ciudad de Santiago de Chile. Universitas Psychologica, 13 (2), 491–500. https://doi.org/10.11144/Javeriana.UPSY13-2.vici Little, T. D., Rhemtulla, M., Gibson, K., & Schoemann, A. M. (2013). Why the items versus parcels controversy needn't be one. Psychological Methods, 18 (3), 285–300. https://doi.org/10.1037/a0033266 Little, T. D., Rioux, C., Odejimi, O. A., & Stickley, Z. L. (2022). Parceling in structural equation modeling: A comprehensive introduction for developmental scientists . Cambridge University Press. https://doi.org/10.1017/9781316882054 Marsh, H. W., Hau, K.-T., & Wen, Z. (2004). In search of golden rules: Comment on hypothesis-testing approaches to setting cutoff values for fit indexes and dangers in overgeneralizing Hu and Bentler's (1999) findings. Structural Equation Modeling, 11 (3), 320–341. https://doi.org/10.1207/s15328007sem1103_2 Meade, A. W., & Kroustalis, C. M. (2006). Problems with item parceling for confirmatory factor analytic tests of measurement invariance. Organizational Research Methods, 9 (3), 369–403. https://doi.org/10.1177/1094428105283384 Mellado, R. (2025). Bidirectional relationship between language ability and internalising/externalising behaviour from early to late childhood: Findings from a Chilean cohort. Research on Child and Adolescent Psychopathology, 53 , 597–611. https://doi.org/10.1007/s10802-025-01287-w Millsap, R. E., & Yun-Tein, J. (2004). Assessing factorial invariance in ordered-categorical measures. Multivariate Behavioral Research, 39 (3), 479–515. https://doi.org/10.1207/S15327906MBR3903_4 Morales, M. F., Girard, L.-C., Sawrikar, V., & MacBeth, A. (2023). An examination of behavioural and emotional problems in children exposed prenatally to the 27F Chilean earthquake: Findings from the ELPI cohort. Social Psychiatry and Psychiatric Epidemiology, 58 (7), 1065–1073. https://doi.org/10.1007/s00127-023-02453-5 Morales, M. F., MacBeth, A., Nagin, D., & Girard, L.-C. (2024). Developmental trajectories of aggression, hyperactivity/inattention, and anxious depressed mood: Co-occurring problems within a Chilean context. Current Psychology, 43 (5), 3928–3943. https://doi.org/10.1007/s12144-022-03198-x Muthén, B., & Muthén, L. (2017). Mplus. In Handbook of item response theory (pp. 507–518). Chapman & Hall/CRC. Najman, J. M., Williams, G. M., Nikles, J., Spence, S., Bor, W., O'Callaghan, M., Le Brocque, R., Andersen, M. J., & Shuttlewood, G. (2001). Bias influencing maternal reports of child behaviour and emotional state. Social Psychiatry and Psychiatric Epidemiology, 36 (4), 186–194. https://doi.org/10.1007/s001270170062 Pandolfi, V., Magyar, C. I., & Dill, C. A. (2009). Confirmatory factor analysis of the child behavior checklist 1.5–5 in a sample of children with autism spectrum disorders. Journal of Autism and Developmental Disorders, 39 , 986–995. https://doi.org/10.1007/s10803-009-0709-z Peverill, M., Dirks, M. A., Narvaja, T., Herts, K. L., Comer, J. S., & McLaughlin, K. A. (2021). Socioeconomic status and child psychopathology in the United States: A meta-analysis of population-based studies. Clinical Psychology Review, 83 , 101933. https://doi.org/10.1016/j.cpr.2020.101933 Putnick, D. L., & Bornstein, M. H. (2016). Measurement invariance conventions and reporting: The state of the art and future directions for psychological research. Developmental Review, 41 , 71–90. https://doi.org/10.1016/j.dr.2016.06.004 Rosenfield, S., & Mouzon, D. (2013). Gender and mental health. In C. S. Aneshensel, J. C. Phelan, & A. Bierman (Eds.), Handbook of the sociology of mental health (pp. 277–296). Springer. https://doi.org/10.1007/978-94-007-4276-5_14 Rosseel, Y. (2012). lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48 (2), 1–36. https://doi.org/10.18637/jss.v048.i02 Sterba, S. K., Prinstein, M. J., & Cox, M. J. (2007). Trajectories of internalizing problems across childhood: Heterogeneity, external validity, and gender differences. Development and Psychopathology, 19 (2), 345–366. https://doi.org/10.1017/S0954579407070174 Stewart, L. C., Asadi, S., Rodriguez-Seijas, C., Wilson, S., Michelini, G., Kotov, R., Cicero, D. C., & Olino, T. M. (2024). Measurement invariance of the child behavior checklist (CBCL) across race/ethnicity and sex in the adolescent brain and cognitive development (ABCD) study. Psychological Assessment, 36 (8), 441–452. https://doi.org/10.1037/pas0001319 Tucker, L. R., & Lewis, C. (1973). A reliability coefficient for maximum likelihood factor analysis. Psychometrika, 38 (1), 1–10. https://doi.org/10.1007/BF02291170 van der Sluis, S., Polderman, T. J., Neale, M. C., Verhulst, F. C., Posthuma, D., & Dieleman, G. C. (2017). Sex differences and gender-invariance of mother-reported childhood problem behavior. International Journal of Methods in Psychiatric Research, 26 (3), e1498. https://doi.org/10.1002/mpr.1498 Viola, L., Garrido, G., & Rescorla, L. (2011). Testing multicultural robustness of the child behavior checklist in a national epidemiological sample in Uruguay. Journal of Abnormal Child Psychology, 39 (6), 897–908. https://doi.org/10.1007/s10802-011-9500-z Yarnell, L. M., Sargeant, M. N., Prescott, C. A., Tilley, J. L., Farver, J. A. M., Mednick, S. A., Venables, P. H., Raine, A., & Luczak, S. E. (2013). Measurement invariance of internalizing and externalising behavioral syndrome factors in a non-western sample. Assessment, 20 (5), 642–655. https://doi.org/10.1177/1073191113498114 Yuan, K.-H., & Bentler, P. M. (2000). Three likelihood-based methods for mean and covariance structure analysis with nonnormal missing data. Sociological Methodology, 30 (1), 165–200. https://doi.org/10.1111/0081-1750.00078 Yu, C. Y. (2002). Evaluation of model fit indices for latent variable models with categorical and continuous outcomes. Paper presented at the Annual Conference of the American Educational Research Association, New Orleans, LA. Additional Declarations The authors declare no competing interests. Supplementary Files SupplementaryMaterials.docx Supplementary Materials Cite Share Download PDF Status: Posted Version 1 posted 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-9431030","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":623876541,"identity":"18324f4d-8b08-4dba-b262-79c377270caf","order_by":0,"name":"Ricardo Mellado","email":"data:image/png;base64,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","orcid":"https://orcid.org/0009-0008-4449-298X","institution":"University College London","correspondingAuthor":true,"prefix":"","firstName":"Ricardo","middleName":"","lastName":"Mellado","suffix":""}],"badges":[],"createdAt":"2026-04-15 21:02:58","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-9431030/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9431030/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107484843,"identity":"19476b6f-b68d-4e19-9dc7-0e64e0857050","added_by":"auto","created_at":"2026-04-22 02:33:08","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":70724,"visible":true,"origin":"","legend":"\u003cp\u003eMeasurement models for the CBCL preschool (top) and school-age (bottom) versions. Rectangles represent syndrome subscale parcels: ER = emotional reactivity; WD = withdrawn; AD = anxious/depressed; SC = somatic complaints; AB = aggressive behaviour; AT = attention problems; DR = rule-breaking behaviour. Ovals represent latent internalising and externalising factors. Single-headed arrows denote standardized factor loadings. Values above the rectangles indicate standardized residual variances. Correlations between latent factors were estimated but are omitted for clarity.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-9431030/v1/80e58f18e5d828da4fc51444.png"},{"id":107487197,"identity":"177d4fc9-ea9c-409c-bc52-a38022bd824e","added_by":"auto","created_at":"2026-04-22 02:40:01","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":805131,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9431030/v1/d1b90a9d-1b4a-44dc-aadc-f00ac740e998.pdf"},{"id":107259147,"identity":"57b50f7c-7ac9-4b66-aeea-dfe5f73a82d7","added_by":"auto","created_at":"2026-04-19 12:46:35","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":24407,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Materials\u003c/p\u003e","description":"","filename":"SupplementaryMaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-9431030/v1/464e710dcc77aa8b343262f1.docx"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eMeasurement Invariance of the Child Behavior Checklist (CBCL) Across Gender in a Chilean Child Cohort\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eSocial-emotional development forms a critical foundation for a child\u0026apos;s future well-being. Emotional and behavioural symptoms in childhood\u0026mdash;such as internalising problems (e.g., anxiety, withdrawal) and externalising behaviours (e.g., aggression, hyperactivity)\u0026mdash;are consistently linked to adverse outcomes later in life, including poverty, mental health disorders, and social exclusion (Dekker et al., 2007; Fergusson et al., 2005). Identifying these difficulties early is crucial for guiding timely interventions. This requires tools that can effectively monitor mental health from early childhood and distinguish typical patterns from those that may signal risk.\u003c/p\u003e\n\u003cp\u003eHowever, efforts to identify and compare patterns of emotional and behavioural problems must also consider how symptoms may be interpreted or reported differently across demographic groups, particularly by gender. Research consistently documents gender differences in symptom prevalence: boys exhibit higher rates of externalising behaviours, while girls show elevated internalising symptoms (Hankin \u0026amp; Abramson, 2001; Rosenfield \u0026amp; Mouzon, 2013; Sterba et al., 2007). Yet observed gender differences may reflect not only true variation in symptom levels, but also systematic differences in how behaviours are interpreted and reported for boys versus girls \u0026mdash; a possibility with direct implications for the validity of gender-based comparisons in developmental research.\u003c/p\u003e\n\u003cp\u003eEmpirical evidence indicates that parental reports of child behaviour often show systematic gender-based patterns even when children\u0026rsquo;s self-reports do not: mothers tend to attribute more internalising symptoms to daughters and more externalising symptoms to sons beyond what children report about themselves (Najman et al., 2001). These reporting differences may arise from several mechanisms. First, cultural gender stereotypes may influence interpretation thresholds: identical aggressive behaviours may be normalized in boys (\u0026ldquo;boys will be boys\u0026rdquo;) but interpreted as atypical in girls, potentially leading caregivers to underreport externalising behaviours in boys while perceiving similar behaviours in girls as more problematic. Second, symptom manifestations may differ by gender: boys often express externalising problems through physical aggression, whereas girls may demonstrate relational aggression (e.g., social exclusion or rumour-spreading), which may not map equivalently onto CBCL items emphasizing physical behaviours (Crick \u0026amp; Grotpeter, 1995). Third, internalising symptoms in boys may be underrecognized due to masculine socialization discouraging emotional expression, potentially inflating apparent gender differences (Chaplin \u0026amp; Aldao, 2013). In Latin American contexts, where traditional gender role expectations may remain relatively salient, these processes could plausibly be amplified\u0026mdash;yet this possibility has received little empirical attention.\u003c/p\u003e\n\u003cp\u003eThese concerns raise a critical methodological question: does the CBCL measure internalising and externalising constructs equivalently for boys and girls? Establishing measurement equivalence is essential because developmental researchers routinely use the CBCL to compare symptom levels across gender, identify gender-specific risk profiles, examine whether developmental processes differ for boys and girls, and evaluate whether interventions operate similarly across groups (Achenbach et al., 2016) Such applications assume that the instrument functions equivalently across gender. Measurement invariance testing provides a framework for evaluating this assumption through a sequence of increasingly stringent model constraints. If configural invariance holds, the same factor structure \u0026mdash; with internalising and externalising as distinct dimensions defined by the same pattern of syndrome indicators \u0026mdash; applies to both boys and girls. If metric invariance holds, each syndrome subscale contributes to its respective factor with equal strength across gender, meaning that a unit change in the latent factor produces the same expected change in observed scores for boys and girls. If scalar invariance holds, boys and girls with the same latent symptom level would obtain the same expected observed score, enabling valid comparisons of group means (Putnick \u0026amp; Bornstein, 2016). Without scalar invariance, observed gender differences may reflect measurement artefacts rather than true psychological variation.\u003c/p\u003e\n\u003cp\u003eThis issue is particularly relevant given the CBCL\u0026rsquo;s widespread use in large-scale research examining gender differences, including national cohort studies and epidemiological surveillance systems. In Chile, the CBCL is integrated into the nationally representative Chilean Longitudinal Survey of Early Childhood (ELPI), where it has been used to examine developmental trajectories of internalising and externalising symptoms (Morales et al., 2024), gender-specific longitudinal processes linking behavioural problems and language development (Mellado, 2025), and early gender differences in socio-emotional outcomes within broader analyses of skill formation and inequality (Behrman et al., 2017). These analyses involve comparing regression coefficients and latent means across gender, which require metric and scalar invariance respectively.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePrior Research on Measurement Invariance of the CBCL\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMeasurement invariance of the CBCL has been examined across several types of groups. Cross-national studies have shown that the instrument\u0026rsquo;s syndrome structure is broadly stable across countries (Ivanova et al., 2007, 2010), and additional work has evaluated invariance across clinical populations and racial or ethnic groups with generally supportive findings (Pandolfi et al., 2009; Yarnell et al., 2013; Stewart et al., 2024). Research on gender-based measurement invariance remains comparatively limited and has focused primarily on the school-age version of the CBCL. Existing studies include analyses in a population-based sample in Mauritius (Yarnell et al., 2013), a recent population-based study in the United States (Stewart et al., 2024), and a clinical sample of children referred for psychiatric evaluation (Sluis et al., 2017). Across these studies, findings generally support configural and metric invariance, with scalar invariance also supported in some cases, suggesting broadly comparable measurement across boys and girls in the school-age CBCL.\u003c/p\u003e\n\u003cp\u003eDespite these contributions, several key gaps remain. First, existing CBCL gender measurement invariance studies have been conducted almost exclusively in high-income countries or European clinical settings, with the exception of Yarnell et al. (2013) in Mauritius. This limits generalizability to contexts where gender socialization norms and parental reporting patterns may differ. No study has examined gender measurement invariance of the CBCL in a large population-based sample from a Latin American context, despite the increasing use of the instrument in such settings.\u003c/p\u003e\n\u003cp\u003eSecond, gender-based invariance has not been formally tested in the preschool version of the CBCL, despite early childhood being a developmental stage when gender-typed behavioural expectations are actively forming and parental interpretation of ambiguous behaviours may be especially susceptible to gender bias (Chaplin \u0026amp; Aldao, 2013). Third, measurement invariance studies of child and adolescent mental health instruments frequently test only configural or metric invariance and do not proceed to scalar invariance \u0026mdash; the level required for valid comparisons of mean symptom levels across groups (Putnick \u0026amp; Bornstein, 2016). This constrains the conclusions that can be drawn from reported gender differences and underscores the need for systematic evaluation of measurement equivalence in large population-based samples.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThis study\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study tests whether the CBCL measures internalising and externalising problems equivalently across gender in a large, nationally representative Chilean sample. Using data from the Chilean Longitudinal Survey of Early Childhood (n \u0026gt; 10,000), measurement invariance is evaluated across both CBCL versions (preschool 1.5\u0026ndash;5 years; school-age 6\u0026ndash;18 years).\u003c/p\u003e\n\u003cp\u003eThis study is important for three main reasons. First, Chile provides a valuable test case as a middle\u0026ndash;upper-income Latin American country where the CBCL is widely used in both research and national policy, yet where cross-group measurement equivalence has not been examined. Patterns of gender socialization and parental reporting in this context may not directly align with those observed in the high-income settings where most existing evidence has been generated.\u003c/p\u003e\n\u003cp\u003eSecond, measurement invariance is examined across both CBCL versions using data from the same nationally representative cohort at two developmental stages. Specifically, children in the ELPI cohort were assessed using the preschool version of the CBCL in early childhood (wave 1) and the school-age version in later childhood (wave 3). This design allows for a consistent evaluation of measurement equivalence across the critical transition from early to middle childhood within a single population-based sample. Most existing gender invariance studies focus exclusively on the school-age version (e.g., Stewart et al., 2024; van der Sluis et al., 2017; Yarnell et al., 2013), and formal testing in the preschool version has not been reported. Examining both versions within the same cohort provides more comprehensive evidence on how the CBCL functions across developmental stages.\u003c/p\u003e\n\u003cp\u003eThird, this study provides the necessary psychometric foundation for interpreting gender comparisons in CBCL-based longitudinal research within this cohort. Recent studies using the ELPI have examined gender-specific developmental trajectories of internalising and externalising symptoms (Morales et al., 2024) and tested whether cross-lagged associations between behavioural problems and language development differ for boys and girls (Mellado, 2025). These analyses compare regression coefficients and mean levels across gender \u0026mdash; comparisons that require metric and scalar invariance, respectively. \u0026nbsp;In addition, strict invariance was evaluated as a more stringent test of measurement equivalence. \u0026nbsp;By establishing whether these conditions hold, the present study provides empirical grounding for gender comparisons that are routinely conducted in ELPI-based research but have not been formally validated. While cross-sectional measurement invariance does not substitute for longitudinal invariance testing \u0026mdash; which is not feasible in this cohort due to the transition from the preschool to the school-age CBCL version across waves (2010, 2012, 2017) \u0026mdash; it establishes a necessary condition: that gender comparisons at each measurement occasion are not confounded by differential instrument functioning.\u003c/p\u003e\n\u003cp\u003eBeyond addressing these specific gaps, this study contributes to the broader cross-cultural validation of the CBCL in Latin America. To date, validation work in the region has focused primarily on evaluating the instrument\u0026apos;s factor structure and psychometric properties (Lecannelier et al., 2014 in Chile; Viola et al., 2011 in Uruguay). The present study extends this literature by explicitly testing whether constructs derived from an instrument developed in Anglophone contexts retain equivalent meaning across demographic subgroups within a Latin American society.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthical Approval\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEthical approval for the original ELPI data collection was granted by the Ethics Committee of the Microdata Centre of the University of Chile. Written informed consent was obtained from parents or legal guardians. The present study involved secondary analysis of anonymized data and did not require additional ethical approval.\u003c/p\u003e"},{"header":"Methods","content":"\u003ch2\u003eData\u0026nbsp;\u003c/h2\u003e\n\u003cp\u003eThis study uses data from the Chilean Longitudinal Survey of Early Childhood (ELPI), a nationally representative dataset. ELPI collects information through face-to-face interviews, including a socio-demographic questionnaire administered to mothers and a set of assessments measuring cognitive, socio-emotional, and anthropometric development in children and their mothers.\u003c/p\u003e\n\u003cp\u003eThe first wave, conducted in 2010, included a nationally representative sample of approximately 15,000 children aged 6 months to 5 years, drawn from official birth records for children born between 1 January 2006 and 31 August 2009. The second wave, fielded in 2012, followed the original sample and added approximately 3,000 children born between 1 September 2009 and 31 December 2011. The third wave, conducted in 2017, included participants from earlier waves and introduced approximately 5,000 additional children born between 1 January 2012 and 31 December 2016, covering ages 6 months to 12 years.\u003c/p\u003e\n\u003cp\u003eELPI uses a two-stage cluster-stratified sampling design based on birth records from the Chilean Civil Registry (Behrman et al., 2010). In the first stage, municipalities served as the primary sampling units. Eighty-three municipalities representing the largest urban areas and approximately 74% of the national population were selected with certainty, while the remaining municipalities were grouped into clusters defined by region, socioeconomic composition, and population size, from which one municipality per cluster was randomly selected. In the second stage, children were randomly sampled within selected municipalities in proportion to their population size.\u003c/p\u003e\n\u003cp\u003eIn wave 1, 11,231 children aged 18 to 59 months constituted the target sample for assessment with the preschool version of the CBCL, of whom 11,193 completed the measure (99.7%). The wave 1 sample had a balanced distribution by sex assigned at birth, with most children aged 2 or 3 years. Throughout this paper, gender refers to the binary classification (male/female) based on sex assigned at birth as recorded in ELPI administrative records. In wave 3, 11,658 children aged 72 months or older constituted the target sample for the school-age version of the CBCL, and 11,633 completed the assessment (99.8%). This sample was also evenly distributed by gender, with ages ranging from 6 to 12 years and the largest subgroups between 9 and 11 years. Approximately 63% of the wave 3 sample (7,021 children) had also participated in wave 1. Descriptive characteristics of the analytic samples are presented in Table 1\u003c/p\u003e\n\u003cp\u003eAnalyses were restricted to children with valid information on age and CBCL items. Missing data on CBCL items were low: among respondents who completed the CBCL, 34 cases (0.3%) in the preschool sample and 87 cases (0.7%) in the school-age sample had at least one missing item. After excluding cases with missing age and those with missing CBCL items, the final analytic sample consisted of 10,987 children for the preschool CBCL models and 11,626 children for the school-age CBCL models. Given the low proportion of missingness, analyses were conducted using complete cases.\u003c/p\u003e\n\u003cp\u003eTable 1: Sample Characteristics by Wave (ELPI 2010 and 2017; analytic sample)\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eVariables\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eWave\u0026nbsp;1 (2010)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eWave\u0026nbsp;3 (2017)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eChild Characteristics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\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: 41.5385%;\"\u003e\n \u003cp\u003eSample Size\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e10,987\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e11,626\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003eSex\u0026nbsp;Assigned\u0026nbsp;at Birth\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\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: 41.5385%;\"\u003e\n \u003cp\u003eMale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e5,472 (49.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e5,917 (50.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003eFemale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e5,515 (50.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e5,709 (49.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003eAge\u0026nbsp;Distribution 1 year\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1,923 (17.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003e2\u0026nbsp;years\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e3,799 (34.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003e3\u0026nbsp;years\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e3,650 (33.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003e4\u0026nbsp;years\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e1,615 (14.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003e6\u0026nbsp;years\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e801 (6.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003e7\u0026nbsp;years\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e850 (7.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003e8\u0026nbsp;years\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e1,486 (12.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003e9\u0026nbsp;years\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e2,658 (22.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003e10 years\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e2,520 (21.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003e11 years\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e2,584 (22.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003e12 years\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e727 (6.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"3\" valign=\"top\" style=\"width: 100%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMaternal Characteristics\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eAge\u0026nbsp;of Mother\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003e\u0026lt;\u0026nbsp;25\u0026nbsp;years\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e3,177 (29%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e257 (2.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003e25\u0026ndash;34 years\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e4,872 (44.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e4,640 (41.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003e35+ years\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e2,905 (26.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e6,231 (56.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003eEmployment Status\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\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: 41.5385%;\"\u003e\n \u003cp\u003eWorked\u0026nbsp;last\u0026nbsp;week\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e4,769 (43.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e6,418 (57.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003eDid\u0026nbsp;not\u0026nbsp;work\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e6,185 (56.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e4,710 (42.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003eEducation Level\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\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: 41.5385%;\"\u003e\n \u003cp\u003ePrimary\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e1,932 (17.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e1,716 (15.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003eSecondary\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e6,350 (58.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e6,185 (55.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003eVocational\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e1,286 (11.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e1,621 (14.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 41.5385%;\"\u003e\n \u003cp\u003eUniversity\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 28.9231%;\"\u003e\n \u003cp\u003e1,237 (11.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 29.5385%;\"\u003e\n \u003cp\u003e1,600 (14.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eNote.\u0026nbsp;\u003c/em\u003ePercentages are based on available observations. Analytic sample sizes reflect children with complete age and CBCL data; maternal characteristics may have smaller denominators due to missing data.\u003c/p\u003e\n\u003cp\u003eMaternal characteristics reflect expected demographic shifts across waves. In wave 1, most mothers were under 35 years old (72.9%), with just over half not in employment (56.4%) and educational attainment concentrated at the secondary level (58.7%). By wave 3, mothers were older (56.8% aged 35+), more frequently employed (57.7%), and slightly more educated.\u003c/p\u003e\n\u003cp\u003eIn the preschool sample (wave 1; N = 10,987), boys and girls exhibited similar internalising scores (boys: M = 58.8, SD = 9.48; girls: M = 59.4, SD = 9.45), while boys scored higher on externalising behaviour (boys: M = 60.1, SD = 10.6; girls: M = 58.3, SD = 10.3). In the school-age sample (wave 3; N = 11,626), boys showed slightly higher internalising scores than girls (boys: M = 53.8, SD = 10.7; girls: M = 51.9, SD = 11.0) and similar externalising scores (boys: M = 50.0, SD = 10.1; girls: M = 49.7, SD = 9.6).\u003c/p\u003e\n\u003ch2\u003eCBCL Measure\u003c/h2\u003e\n\u003cp\u003eCaregivers complete the CBCL by rating items on a 3-point scale: 0 (\u0026ldquo;Not true\u0026rdquo;), 1 (\u0026ldquo;Somewhat or sometimes true\u0026rdquo;), and 2 (\u0026ldquo;Very true or often true\u0026rdquo;). Administration of the CBCL typically takes approximately 20\u0026ndash;25 minutes. Items assessing internalising and externalising symptoms are grouped into empirically derived syndrome subscales\u0026mdash;six in the preschool version and five in the school-age version.\u003c/p\u003e\n\u003cp\u003eIn the preschool form, internalising behaviour is assessed through four subscales: emotional reactivity, anxious/depressed, withdrawn, and somatic complaints. Externalising behaviour is measured using two subscales: attention problems and aggressive behaviour. In the school-age version, internalising behaviour comprises anxious/depressed, withdrawn, and somatic complaints, while externalising behaviour is assessed through rule-breaking and aggressive behaviour.\u003c/p\u003e\n\u003cp\u003eThe preschool version includes 99 items, and the school-age version includes 112. Scoring focuses on syndrome-relevant items: 36 internalising and 24 externalising items in the preschool form, and 26 internalising and 25 externalising items in the school-age version. Higher scores indicate greater symptom severity.\u003c/p\u003e\n\u003ch2\u003eAnalytical Strategy\u003c/h2\u003e\n\u003cp\u003eAll analyses were conducted using Mplus version 8.11 (Muth\u0026eacute;n \u0026amp; Muth\u0026eacute;n, 2017) and the lavaan package in R (Rosseel, 2012). Sampling weights provided by ELPI were incorporated in all models to adjust for unequal probabilities of selection and to produce population-representative estimates. The primary objective of the study was to evaluate measurement invariance of the CBCL internalising and externalising constructs across gender.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBefore invariance testing could proceed, several measurement modelling decisions were required \u0026mdash; specifically, the selection of an appropriate factorial structure and the choice between item-level and parcel-level indicators. Because these decisions depend on empirical considerations such as model fit, convergence, and subscale unidimensionality, preliminary analyses were conducted to inform model specification. Measurement invariance procedures and sensitivity analyses are then described to assess the robustness of the results.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModel selection\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAs an initial step, the factorial structure of the CBCL was evaluated separately for the preschool (1.5\u0026ndash;5 years) and school-age (6\u0026ndash;18 years) versions to identify an appropriate baseline measurement model for subsequent invariance testing. The baseline model must provide adequate fit and converge reliably across groups; if the theoretically preferred specification fails to meet these criteria, an alternative is required. Two item-level confirmatory factor analytic models were compared \u0026mdash; a second-order hierarchical model and a first-order two-factor model \u0026mdash; both estimated using the robust weighted least squares estimator (WLSMV), appropriate for ordered-categorical data (Flora \u0026amp; Curran, 2004). Items were modelled using their original 3-point response scale (0 = \u0026quot;Not true\u0026quot;; 1 = \u0026quot;Somewhat or sometimes true\u0026quot;; 2 = \u0026quot;Very true or often true\u0026quot;).\u003c/p\u003e\n\u003cp\u003eThe first specification was a second-order hierarchical model in which items loaded onto their respective syndrome factors (six in the preschool version and five in the school-age version), which in turn loaded onto higher-order internalising and externalising factors. This model represents the full theoretical structure of the CBCL and is the most faithful representation of the instrument\u0026rsquo;s design (Achenbach \u0026amp; Rescorla, 2001). The second specification was a first-order two-factor model in which items loaded directly onto internalising and externalising factors, bypassing the syndrome level (Konold \u0026amp; Pianta, 2004). Both models specify internalising and externalising behaviour as the core latent constructs \u0026mdash; the dimensions whose measurement equivalence across gender is the focus of this study.\u003c/p\u003e\n\u003cp\u003eModel fit was assessed using the Root Mean Square Error of Approximation (RMSEA), the Comparative Fit Index (CFI), and the Tucker-Lewis Index (TLI). RMSEA values \u0026le; .05 were considered indicative of good fit and values \u0026le; .08 of acceptable fit (Browne \u0026amp; Cudeck, 1992; Yu, 2002). For CFI and TLI, values \u0026ge; .90 were considered indicative of good fit, with values between .80 and .90 interpreted as acceptable (Browne \u0026amp; Cudeck, 1992). The more stringent .95 threshold proposed by Hu and Bentler (1999) was not applied, as it has been criticized for over-rejecting correctly specified complex models with large numbers of indicators, as is the case for both CBCL versions (Marsh et al., 2004).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2: Fit indices for CFA models across CBCL broadband structures (preschool and school-age versions)\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 184px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eModel\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSB\u003c/strong\u003e𝜒\u003csup\u003e2\u003c/sup\u003e \u003cstrong\u003e(df)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCFI\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTLI\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRMSEA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 184px;\"\u003e\n \u003cp\u003e\u003cem\u003eCBCL\u0026nbsp;1.5\u0026ndash;5\u0026nbsp;(Preschool version)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\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: 184px;\"\u003e\n \u003cp\u003e2-Factor Model\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 98px;\"\u003e\n \u003cp\u003e52843.24 (1709)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.814\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.807\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.052\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 281px;\"\u003e\n \u003cp\u003eSecond-Order\u0026nbsp;2-Factor Model 45461.83 (1703)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.841\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.834\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.048\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 281px;\"\u003e\n \u003cp\u003e\u003cem\u003eCBCL\u0026nbsp;6\u0026ndash;18\u0026nbsp;(School-age version)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 281px;\"\u003e\n \u003cp\u003e2-Factor Model 28705.68 (1709)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.811\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.805\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.037\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 281px;\"\u003e\n \u003cp\u003eSecond-Order\u0026nbsp;2-Factor Model 26650.01 (1704)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.826\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.819\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.035\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eNote.\u0026nbsp;\u003c/em\u003eSB𝜒\u003csup\u003e2\u003c/sup\u003e = Satorra\u0026ndash;Bentler scaled chi-square; df = degrees of freedom; CFI = Comparative Fit Index; TLI = Tucker\u0026ndash;Lewis Index; RMSEA = Root Mean Square Error of Approximation.\u003c/p\u003e\n\u003cp\u003eAs shown in Table 2, both specifications produced acceptable RMSEA values and CFI and TLI indices in the .80\u0026ndash;.90 range, consistent with prior item-level CBCL factor analyses involving large numbers of ordinal indicators (Ivanova et al., 2010). The second-order model achieved marginally better fit across all indices in both versions (\u0026Delta;CFI \u0026asymp; .02, \u0026Delta;TLI \u0026asymp; .02, \u0026Delta;RMSEA \u0026asymp; .004), \u0026nbsp;however the second-order model failed to converge in both CBCL versions, producing Heywood cases with negative residual variances (see Table S1 in the Supplementary Materials for second-order factor loadings and inter-factor correlations for both model specifications). These issues arose from very high intercorrelations among syndrome factors, causing the second-order factor to over-explain first-order variance. In both versions, this produced standardized second-order loadings exceeding 1.0 (emotional reactivity \u0026lambda; = 1.073 in the preschool version; anxiety/depression \u0026lambda; = 1.024 in the school-age version), indicating Heywood cases. Similar convergence failures have been reported in prior studies applying second-order CFA models to the CBCL in population-based samples (Benninger et al., 2025; Ivanova et al., 2010), likely reflecting the substantial comorbidity across syndrome scales that produces these high inter-factor correlations. Given that the model failed to converge even in single-group analyses, it could not be carried forward into multigroup invariance testing, where estimation demands are substantially greater.\u003c/p\u003e\n\u003cp\u003eThe first-order two-factor model, which did not present convergence issues, was therefore retained as the basis for subsequent measurement modelling. However, rather than testing measurement invariance at the item level, established CBCL measurement invariance practice was followed (Konold \u0026amp; Pianta, 2004; Pandolfi et al., 2009), and syndrome subscale scores were used as s factor. Testing multigroup invariance at the item level with 60+ ordinal indicators requires the simultaneous estimation of a large number of parameters (loadings, thresholds, and residual variances for each item across groups), which increases the risk of estimation instability and model non-convergence\u0026mdash;particularly with ordinal data, the WLSMV estimator, and the incorporation of sampling weights. Parcelling addresses this by reducing the number of estimated parameters and, through aggregation, attenuating item-specific measurement error (Little et al., 2013), as syndrome subscale scores average across item-level variance, yielding more reliable and approximately continuous indicators.\u0026nbsp;This allowed the use of robust maximum likelihood estimation (MLR) for multigroup invariance testing.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe validity of parcelling rests on the assumption that items within each syndrome subscale form approximately unidimensional constructs. If items within a parcel capture multiple underlying dimensions, the resulting parcel indicators may conflate distinct sources of variance, undermining their interpretation as indicators of a single latent factor. To evaluate this assumption, item-level confirmatory factor analyses were conducted separately for each syndrome subscale using the WLSMV estimator on the original ordinal items. Internal consistency was assessed using ordinal Cronbach\u0026rsquo;s alpha and McDonald\u0026apos;s omega. Table 3 reports fit indices and reliability estimate for these models.\u003c/p\u003e\n\u003cp\u003eTable\u0026nbsp;3:\u0026nbsp;Fit\u0026nbsp;indices\u0026nbsp;and\u0026nbsp;internal\u0026nbsp;consistency\u0026nbsp;estimates\u0026nbsp;for\u0026nbsp;one-factor\u0026nbsp;CFA\u0026nbsp;models\u0026nbsp;of\u0026nbsp;CBCL\u0026nbsp;syndrome\u0026nbsp;subscales\u0026nbsp;(preschool and school-age versions)\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003eVersion Syndrome\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003eSB𝜒\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 36px;\"\u003e\n \u003cp\u003edf\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003eCFI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003eTLI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 62px;\"\u003e\n \u003cp\u003eRMSEA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003eSRMR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e𝜆̄\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e𝜆̃\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e𝛼𝑜𝜏𝑑\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e𝑚\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e\u003cem\u003eCBCL 1.5\u0026ndash;5\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 36px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 62px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\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: 148px;\"\u003e\n \u003cp\u003eEmotional\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003e567.14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 36px;\"\u003e\n \u003cp\u003e27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.955\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.940\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 62px;\"\u003e\n \u003cp\u003e0.051\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e0.051\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.544\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.524\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.789\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.691\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003eWithdrawn\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003e238.31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 36px;\"\u003e\n \u003cp\u003e20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.975\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.965\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 62px;\"\u003e\n \u003cp\u003e0.038\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e0.038\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.555\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.629\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.769\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.593\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003eAnxious\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003e989.73\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 36px;\"\u003e\n \u003cp\u003e20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.901\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.862\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 62px;\"\u003e\n \u003cp\u003e0.076\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e0.074\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.557\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.531\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.765\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.638\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003eSomatic\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003e3642.26\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 36px;\"\u003e\n \u003cp\u003e44\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.734\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.667\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 62px;\"\u003e\n \u003cp\u003e0.091\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e0.123\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.513\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.543\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.765\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.581\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003eAggressive\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003e6745.35\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 36px;\"\u003e\n \u003cp\u003e152\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.919\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.909\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 62px;\"\u003e\n \u003cp\u003e0.075\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e0.058\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.607\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.622\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.914\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.893\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003eAttention\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003e131.60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 36px;\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.946\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.893\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 62px;\"\u003e\n \u003cp\u003e0.054\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e0.061\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.481\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.473\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.602\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.501\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003e\u003cem\u003eCBCL 6\u0026ndash;18\u003c/em\u003e\u003c/p\u003e\n \u003cp\u003eAnxiety\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1559.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 36px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.936\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.923\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 62px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.053\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.064\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.591\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.587\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.871\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.790\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003eWithdrawn\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003e437.58\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 36px;\"\u003e\n \u003cp\u003e20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.964\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.949\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 62px;\"\u003e\n \u003cp\u003e0.053\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e0.050\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.667\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.678\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.862\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.739\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003eSomatic\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003e3.84\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 36px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.999\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.996\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 62px;\"\u003e\n \u003cp\u003e0.012\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e0.011\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.603\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.610\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.693\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.500\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003eRule-breaking\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003e1185.65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 36px;\"\u003e\n \u003cp\u003e119\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.868\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.849\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 62px;\"\u003e\n \u003cp\u003e0.032\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e0.091\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.687\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.672\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.934\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.706\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 148px;\"\u003e\n \u003cp\u003eAggressive\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003e4338.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 36px;\"\u003e\n \u003cp\u003e135\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.920\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.909\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 62px;\"\u003e\n \u003cp\u003e0.061\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e0.078\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.681\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.710\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.936\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.884\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eNote.\u0026nbsp;\u003c/em\u003eSB𝜒\u003csup\u003e2\u003c/sup\u003e = Satorra\u0026ndash;Bentler scaled chi-square; df = degrees of freedom; CFI = Comparative Fit Index; TLI = Tucker\u0026ndash;Lewis Index; RMSEA = Root Mean Square Error of Approximation; SRMR = Standardized Root Mean Square Residual; 𝜆̄ = mean standardized factor loading; 𝜆̃ = median standardized factor loading; 𝛼𝑜𝜏𝑑 = ordinal Cronbach\u0026rsquo;s alpha; 𝜔 = McDonald\u0026rsquo;s omega.\u003c/p\u003e\n\u003cp\u003eAs shown in Table 3, most subscales demonstrated acceptable to good model fit. In the preschool version, CFI values ranged from .901 (Anxious/Depressed) to .975 (Withdrawn), with RMSEA values between .038 and .076. In the school-age version, fit indices were broadly comparable, with CFI ranging from .868 (Rule-Breaking) to .999 (Somatic) and RMSEA from .012 to .061. Internal consistency was adequate to strong across both versions, with ordinal alpha ranging from .60 to .91 and omega from .50 to .89 in the preschool version, and alpha from .69 to .94 and omega from .50 to .88 in the school-age version. Mean standardized factor loadings ranged from .48 to .61 in the preschool version and .59 to .69 in the school-age version.\u003c/p\u003e\n\u003cp\u003eTwo subscales showed comparatively poorer fit. The Somatic Complaints subscale in the preschool version exhibited the weakest unidimensional fit (CFI = .734, RMSEA = .091), with modification indices revealing several large residual correlations among items, indicating local dependencies and departures from unidimensionality (see Table S2 in the Supplementary Materials for the largest modification indices corresponding to residual correlations among somatic items). Despite this, the subscale demonstrated acceptable internal consistency (\u0026alpha; = .77) and was retained given its theoretical centrality to the internalising domain. Similarly, the Rule-Breaking subscale in the school-age version showed borderline fit (CFI = .868, TLI = .849), though internal consistency was strong (\u0026alpha; = .93). In both cases, aggregating items into subscale scores reduces item-specific measurement error and facilitates stable multigroup estimation. The performance of these subscales is revisited in the item-level sensitivity analyses.\u003c/p\u003e\n\u003cp\u003eDistributional properties of the resulting parcels were examined to assess their suitability as continuous indicators. For the preschool version, skewness and kurtosis values were within acceptable ranges (skewness: -0.16 to 1.21; kurtosis: 2.43 to 4.87). For the school-age version, parcels showed moderate to pronounced positive skew, with Rule-Breaking exhibiting the most extreme values (skewness = 4.20, kurtosis = 44.8), consistent with low base rates of these behaviours in population samples (see Table S3 in the Supplementary Materials for parcel-level distributional properties). Robust maximum likelihood estimation (MLR) was used throughout, which provides corrected standard errors and fit statistics under non-normality (Yuan \u0026amp; Bentler, 2000).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMeasurement invariance testing\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMeasurement invariance was evaluated using multigroup confirmatory factor analysis (CFA) across gender, estimated separately for the preschool and school CBCL versions. Models were estimated using robust maximum likelihood (MLR) on the parcel-level indicators. Following standard practice, a sequence of increasingly constrained models was estimated: configural invariance (equal factor structure), metric invariance (equal factor loadings), scalar invariance (equal intercepts), and strict invariance (equal residual variances). Given the wide age ranges within each version (approximately 1.5 to 5 years in the preschool sample and 6 to 12 years in the school-age sample), age in months was included as a covariate on the latent internalising and externalising factors in all models to ensure that gender comparisons were not confounded by developmental differences within each sample. Model comparisons relied on changes in approximate fit indices rather than chi-square difference testing, which is overly sensitive to sample size in large samples. Following Chen (2007), invariance was supported if decreases in CFI were less than .010 and increases in RMSEA were less than .015. Note that at the parcel level, scalar invariance involves constraining intercepts \u0026mdash; not thresholds \u0026mdash; as parcels are treated as continuous indicators.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSensitivity Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo evaluate the robustness of the primary findings, two complementary sensitivity analyses were conducted. These analyses address a key methodological limitation of parcelling: aggregating items into syndrome-level composites may mask item-level differences in how the CBCL functions across gender.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eAlternative parcelling scheme\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eFirst, it was examined whether invariance conclusions depended on how items were aggregated into parcels. The primary analysis used one parcel per syndrome (six parcels in the preschool version and five in the school-age version).\u0026nbsp;An alternative parcelling scheme divided syndromes that were either larger in size or showed marginal unidimensional fit (see Table 3) into two parcels each. For the preschool version, Anxious/Depressed, Somatic Complaints, and Aggressive Behaviour were split, while Emotional Reactivity, Withdrawn, and Attention Problems were retained as single parcels, producing a nine-parcel model. For the school-age version, Anxiety/Depression, Rule-Breaking, and Aggressive Behaviour were split, while Withdrawn and Somatic Complaints were retained, producing an eight-parcel model. The same invariance sequence (configural, metric, scalar, and strict) was applied using MLR with age included as a covariate. Consistent results across parcelling schemes would indicate that invariance conclusions are not dependent on the specific aggregation of items.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eItem-level invariance within syndromes\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eSecond, to assess whether parcelling masked item-level non-invariance, measurement invariance was tested separately within each syndrome subscale. Each syndrome was modelled as a unidimensional factor using the original three-point ordinal items and estimated using WLSMV. Configural, metric, and scalar invariance were evaluated across gender. At this level, scalar invariance involves constraining thresholds rather than intercepts, consistent with ordinal measurement models. This analysis provides a more granular assessment of measurement equivalence and allows identification of potential item-level gender differences that could be obscured when items are aggregated into parcels.\u003c/p\u003e"},{"header":"Results","content":"\u003ch2\u003e\u003cstrong\u003eConfirmatory Factor Analysis\u003c/strong\u003e\u0026nbsp;\u003c/h2\u003e\n\u003cp\u003eAs described in the Analytical Strategy section, the first-order two-factor model was retained, and syndrome subscale scores were used as parcel-level indicators of the internalising and externalising factors. Parcel-level CFA models were estimated for both CBCL preschool and school-age versions using MLR. Standardized factor loadings were generally moderate to strong, ranging from .53 to .91 across parcels in both models. The correlation between internalising and externalising factors was high in both versions (r = .72 in the preschool model; r = .77 in the school-age model), indicating substantial shared variance between the constructs. Model fit was acceptable to excellent: preschool \u0026chi;\u0026sup2; (8) = 230.23, CFI = .988, TLI = .977, RMSEA = .050, SRMR = .020; school-age \u0026chi;\u0026sup2; (4) = 107.79, CFI = .991, TLI = .977, RMSEA = .047, SRMR = .015. These results support the adequacy of the parcel-level measurement models for subsequent invariance testing. Standardized factor loadings and residual variances for both models are shown in Figure 1.\u003c/p\u003e\n\u003ch2\u003eMeasurement Invariance CBCL 1 and CBCL 2 by Gender\u003c/h2\u003e\n\u003cp\u003eMeasurement invariance of the internalising and externalising factors was tested separately for the preschool and school-age versions of the CBCL across gender. For each version, a sequence of increasingly constrained models was estimated: configural (equal factor structure), metric (equal factor loadings), scalar (equal intercepts), and strict (equal residual variances), with age included as a covariate in all models.\u0026nbsp;\u003c/p\u003e\n\u003ch3\u003eTable 4: Measurement invariance across gender (configural, metric, scalar, and strict models): Fit indices for CBCL preschool and school-age versions\u003c/h3\u003e\n\u003ctable border=\"0\" width=\"100%\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 78px;\"\u003e\u003cstrong\u003eModel\u003c/strong\u003e\u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\u003cstrong\u003eSB\u003c/strong\u003e𝜒\u003csup\u003e2\u003c/sup\u003e\u003cstrong\u003e(df)\u003c/strong\u003e\u003c/td\u003e\n \u003ctd style=\"width: 100px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCFI (\u003c/strong\u003e\u0026Delta;\u003cstrong\u003eCFI\u003c/strong\u003e\u003cstrong\u003e)\u003c/strong\u003e\u003c/p\u003e\u003cbr\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\u003cstrong\u003eTLI (\u003c/strong\u003e\u0026Delta;\u003cstrong\u003eTLI)\u003c/strong\u003e\u003c/td\u003e\n \u003ctd style=\"width: 115px;\"\u003e\u003cstrong\u003eRMSEA (\u003c/strong\u003e\u0026Delta;\u003cstrong\u003eRMSEA)\u003c/strong\u003e\u003c/td\u003e\n \u003ctd style=\"width: 138px;\"\u003e\u003cstrong\u003eSRMR (\u003c/strong\u003e\u0026Delta;\u003cstrong\u003eSRMR)\u003c/strong\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 100%;\" colspan=\"6\"\u003e\u003cem\u003eCBCL 1.5\u0026ndash;5 (Preschool version)\u003c/em\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 78px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eConfigural\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e299\u0026nbsp;(24)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e0.986 (\u0026ndash;)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.976 (\u0026ndash;)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e0.044 (\u0026ndash;)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.018 (\u0026ndash;)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 78px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMetric\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e308\u0026nbsp;(28)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e0.986 (+0.000)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.980 (+0.004)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e0.041\u0026nbsp;(-0.003)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.019 (+0.001)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 78px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eScalar\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e410\u0026nbsp;(32)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e0.981\u0026nbsp;(-0.005)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.976\u0026nbsp;(-0.004)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e0.045 (+0.004)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.023 (+0.004)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 78px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eStrict\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e435\u0026nbsp;(38)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e0.981 (+0.000)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.979 (+0.003)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e0.041\u0026nbsp;(-0.004)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.026 (+0.003)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"6\" valign=\"top\" style=\"width: 593px;\"\u003e\n \u003cp\u003e\u003cem\u003eCBCL\u0026nbsp;6\u0026ndash;18\u0026nbsp;(School-age version)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 78px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eConfigural\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e318\u0026nbsp;(14)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e0.984 (\u0026ndash;)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.966 (\u0026ndash;)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e0.052 (\u0026ndash;)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.019 (\u0026ndash;)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 78px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMetric\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e376\u0026nbsp;(17)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e0.986 (+0.002)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.975 (+0.009)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e0.045\u0026nbsp;(-0.007)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.024 (+0.005)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 78px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eScalar\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e428\u0026nbsp;(20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e0.983\u0026nbsp;(-0.003)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.975 (+0.000)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e0.045 (+0.000)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.026 (+0.002)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 78px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eStrict\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 66px;\"\u003e\n \u003cp\u003e541\u0026nbsp;(25)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 100px;\"\u003e\n \u003cp\u003e0.985 (+0.002)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e0.982 (+0.007)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e0.038\u0026nbsp;(-0.007)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 138px;\"\u003e\n \u003cp\u003e0.034 (+0.008)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eNote.\u0026nbsp;\u003c/em\u003e\u0026Delta;CFI, \u0026Delta;TLI, \u0026Delta;RMSEA, and \u0026Delta;SRMR represent changes relative to the preceding (less constrained) model. SB𝜒\u003csup\u003e2\u003c/sup\u003e = Satorra\u0026ndash;Bentler scaled chi-square; df = degrees of freedom. Age (in months) was included as a covariate on the latent factors in all models.\u003c/p\u003e\n\u003cp\u003eAs shown in Table 4, fit indices for the preschool version remained stable across all four levels of invariance. Changes from configural to metric were negligible (\u0026Delta;CFI = .000, \u0026Delta;RMSEA = -.003), and the transition from metric to scalar produced only minor decrements (\u0026Delta;CFI = -.005, \u0026Delta;RMSEA = +.004). Strict invariance was also supported, with no change in CFI from scalar to strict (\u0026Delta;CFI = .000, \u0026Delta;RMSEA = -.004). All changes were well within recommended thresholds (Chen, 2007), indicating full configural, metric, scalar, and strict invariance of the preschool CBCL across gender.\u003c/p\u003e\n\u003cp\u003eFor the school-age version, a similar pattern emerged. Metric invariance was supported with minimal fit change (\u0026Delta;CFI = +.002, \u0026Delta;RMSEA = -.007), as was scalar invariance (\u0026Delta;CFI = -.003, \u0026Delta;RMSEA = +.001). At the strict level, CFI remained unchanged (\u0026Delta;CFI = .000) and RMSEA improved (\u0026Delta;RMSEA = -.007), though SRMR showed a marginal increase (\u0026Delta;SRMR = .011, slightly above the .010 threshold). Given that CFI and RMSEA were clearly within acceptable limits, strict invariance is considered substantially supported for the school-age version. Together, these results indicate that the CBCL\u0026apos;s internalising and externalising constructs are measured equivalently across gender in both versions, supporting valid comparisons of latent means, regression coefficients, and residual variances between boys and girls.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSensitivity Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eSensitivity Analysis 1: Alternative parcelling\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eTo assess whether invariance conclusions were sensitive to the specific parcelling scheme, the primary analyses were replicated using an alternative specification in which larger or marginally unidimensional syndromes were split into two parcels each. This produced a 9-parcel model for the preschool version and an 8-parcel model for the school-age version. Table 5 presents the results alongside the primary analysis for comparison.\u003c/p\u003e\n\u003ch3\u003eTable 5: Sensitivity Analysis: Measurement invariance results for primary and alternative parcelling schemes across gender\u003c/h3\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eVersion\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;Scheme\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eModel\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCFI\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e\u0026Delta;\u003cstrong\u003eCFI\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRMSEA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u0026Delta;\u003cstrong\u003eRMSEA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u003cem\u003eCBCL\u0026nbsp;1.5\u0026ndash;5 (Preschool)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\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: 201px;\"\u003e\n \u003cp\u003ePrimary\u0026nbsp;(6 parcels)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eConfigural\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.986\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.044\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eMetric\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.986\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e+0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.041\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.003\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eScalar\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.981\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e-0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.045\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e+0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eStrict\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.981\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e+0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.041\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eAlternative\u0026nbsp;(10 parcels)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eConfigural\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.934\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.076\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eMetric\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.935\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e+0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.072\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eScalar\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.930\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e-0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.071\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eStrict\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.930\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e+0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.067\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u003cem\u003eCBCL\u0026nbsp;6\u0026ndash;18\u0026nbsp;(School-age)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\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: 201px;\"\u003e\n \u003cp\u003ePrimary\u0026nbsp;(5 parcels)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eConfigural\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.984\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.052\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eMetric\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.986\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e+0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.045\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.007\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eScalar\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.983\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e-0.003\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.045\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e+0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eStrict\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.985\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e+0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.038\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.007\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003eAlternative\u0026nbsp;(8 parcels)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eConfigural\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.959\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.060\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e\u0026ndash;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eMetric\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.966\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e+0.007\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.051\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.009\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eScalar\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.962\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e-0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.052\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e+0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 201px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eStrict\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 46px;\"\u003e\n \u003cp\u003e0.971\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e+0.009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 64px;\"\u003e\n \u003cp\u003e0.043\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 75px;\"\u003e\n \u003cp\u003e-0.009\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eNote.\u0026nbsp;\u003c/em\u003e\u0026Delta;CFI and \u0026Delta;RMSEA represent changes from the preceding (less constrained) model. Age (in years) was included as a covariate on the latent factors in all models. Primary parcelling used one parcel per syndrome; alternative parcelling split larger or marginally unidimensional syndromes into two parcels each.\u003c/p\u003e\n\u003cp\u003eDespite lower baseline fit in the alternative models \u0026mdash; expected given the increased number of indicators \u0026mdash; the pattern of fit changes across invariance levels was consistent with the primary analysis. \u0026Delta;CFI at the scalar level was -.005 for the preschool version and -.004 for the school-age version, both well within the .010 threshold and comparable to the primary parcelling schemes. Strict invariance was also supported in both versions, with \u0026Delta;CFI values of .000 and +.009 respectively. All changes remained within recommended thresholds (Chen, 2007), indicating that measurement invariance conclusions are robust to the specific aggregation of items into parcels. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eSensitivity Analysis 2: Item-Level Invariance Within Syndromes\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eTo assess whether parcelling masked potential item-level non-invariance, measurement invariance was tested separately within each syndrome using the original ordinal items and WLSMV estimation. Results are summarised in Table 6\u003c/p\u003e\n\u003ch3\u003eTable 6: Item-level measurement invariance across gender for CBCL syndrome subscales (sensitivity analysis).\u003c/h3\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 188px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eVersion\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;Syndrome\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eItems\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eConfig. CFI\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e\u0026Delta;CFIM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003e\u0026Delta;CFIS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e\u0026Delta;CFISt\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 188px;\"\u003e\n \u003cp\u003e\u003cem\u003eCBCL\u0026nbsp;1.5\u0026ndash;5 (Preschool)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\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: 188px;\"\u003e\n \u003cp\u003eEmotional Reactivity\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e.956\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.010\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003e-.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 188px;\"\u003e\n \u003cp\u003eWithdrawn\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e.974\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003e-.006\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 188px;\"\u003e\n \u003cp\u003eAnxious/Depressed\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e.901\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.014\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003e-.008\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 188px;\"\u003e\n \u003cp\u003eSomatic Complaints\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e.734\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.030\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003e-.009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 188px;\"\u003e\n \u003cp\u003eAggressive Behaviour\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e.919\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.026\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003e-.021\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 188px;\"\u003e\n \u003cp\u003eAttention Problems\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e.947\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.007\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003e-.007\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 188px;\"\u003e\n \u003cp\u003e\u003cem\u003eCBCL\u0026nbsp;6\u0026ndash;18\u0026nbsp;(School-age)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\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: 188px;\"\u003e\n \u003cp\u003eAnxiety/Depression\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e.936\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.017\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003e-.007\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 188px;\"\u003e\n \u003cp\u003eWithdrawn\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e.964\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.008\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003e-.002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 188px;\"\u003e\n \u003cp\u003eSomatic Complaints\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e.998\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003e-.006\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 188px;\"\u003e\n \u003cp\u003eRule-Breaking\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e.886\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.034\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003e-.007\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 188px;\"\u003e\n \u003cp\u003eAggressive Behaviour\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 48px;\"\u003e\n \u003cp\u003e18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 85px;\"\u003e\n \u003cp\u003e.923\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.027\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003e-.016\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 57px;\"\u003e\n \u003cp\u003e.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eNote.\u0026nbsp;\u003c/em\u003e\u0026Delta;CFIM, \u0026Delta;CFIS, and \u0026Delta;CFISt represent changes in CFI between configural\u0026ndash;metric, metric\u0026ndash;scalar, and scalar\u0026ndash;strict models, respectively. Following Chen (2007), |\u0026Delta;CFI| \u0026gt; .010 is flagged with \u003csup\u003ea\u003c/sup\u003e. Invariance tests are sequential; therefore scalar and strict results should be interpreted cautiously when metric invariance is not supported, as later steps assume equality constraints established at earlier stages.\u003c/p\u003e\n\u003cp\u003eResults revealed a broadly consistent pattern across versions. In the preschool version, three of six syndromes (Emotional Reactivity, Withdrawn, and Attention Problems) showed fit changes within recommended thresholds across successive invariance steps. In the school-age version, Withdrawn and Somatic Complaints displayed a similar pattern. Several additional syndromes showed small deviations from the \u0026Delta;CFI \u0026lt; .010 guideline at the metric stage, though subsequent scalar constraints resulted in minimal additional fit loss (\u0026Delta;CFI ranging from \u0026minus;.002 to \u0026minus;.009), suggesting that loading differences were relatively small in magnitude.\u003c/p\u003e\n\u003cp\u003eThe Aggressive Behaviour subscale showed the greatest item-level complexity in both versions, with \u0026Delta;CFI at the metric step exceeding .025 and scalar invariance not fully supported. This subscale also contains the largest number of items (19 and 18 respectively) and includes behaviours with closely related content, which may introduce local dependencies among items. Such patterns are consistent with item-level heterogeneity rather than systematic gender-related measurement differences in the broader construct. Importantly, at the parcel level, the Aggressive Behaviour subscale contributed to models achieving strict invariance across gender in both CBCL versions, suggesting that aggregation stabilised the measurement structure without introducing construct-level distortions.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis study evaluated whether the CBCL measures internalising and externalising problems equivalently across gender in a large, nationally representative sample of Chilean children. Using data from the ELPI cohort, we tested measurement invariance of both the preschool (1.5\u0026ndash;5 years) and school-age (6\u0026ndash;12 years) versions, controlling for age-related developmental variation within each sample. Across both versions, the CBCL demonstrated full configural, metric, scalar, and strict invariance across gender at the parcel level, indicating that the instrument\u0026apos;s measurement properties are equivalent for boys and girls. These findings were further supported by two sensitivity analyses \u0026mdash; an alternative parcelling scheme and item-level invariance testing within each syndrome \u0026mdash; which confirmed that the primary results were robust to analytical choices and not artifacts of item aggregation.\u003c/p\u003e\n\u003cp\u003eThese findings have direct practical implications. By establishing measurement invariance across gender, this study provides evidence that observed differences in CBCL scores between boys and girls are likely to reflect true variation in underlying internalising and externalising symptoms rather than artifacts of measurement bias. These results support the use of the CBCL for gender comparisons in Chilean population-based studies, including analyses of symptom prevalence, gender-moderated developmental pathways, and intervention effects.\u003c/p\u003e\n\u003cp\u003eBeyond these immediate implications, the findings strengthen the psychometric foundation for the growing body of ELPI-based research that relies on gender comparisons. Studies using this cohort have examined gender-specific developmental trajectories of internalising and externalising symptoms (Morales et al., 2024) and tested whether cross-lagged associations between behavioural problems and language development differ for boys and girls (e.g., Mellado, 2025). Such analyses rely on metric invariance to validly compare regression coefficients across gender and scalar invariance to compare latent means. The present study therefore provides empirical support for these assumptions, strengthening the psychometric basis for existing and future ELPI-based analyses\u003c/p\u003e\n\u003cp\u003eThe findings also extend the existing psychometric literature on the CBCL in several ways. First, nearly all prior gender MI research on the CBCL has been conducted in high-income, English-speaking countries or European clinical settings. By demonstrating full invariance in a nationally representative Chilean sample, this study contributes to the cross-cultural validation of the CBCL in a middle-income, Spanish-speaking context where gender socialization norms and parental reporting patterns may differ from those in previously studied populations. Second, a notable gap is addressed by formally testing gender invariance in the preschool version of the CBCL, which has received limited empirical attention despite being widely used in developmental research. Third, the analytical approach \u0026mdash; combining parcel-level multigroup CFA with both alternative parcelling and item-level sensitivity analyses \u0026mdash; provides a more comprehensive evaluation of measurement equivalence than has been reported in prior CBCL invariance studies, most of which relied on a single analytical strategy without robustness checks.\u003c/p\u003e\n\u003cp\u003eSome limitations should be acknowledged. First, the primary analysis used item parcelling, which can reduce sensitivity to item-level differences across groups (Meade \u0026amp; Kroustalis, 2006). However, this concern was directly addressed through complementary item-level invariance analyses, which showed that the majority of syndromes achieved invariance at the item level and that the one syndrome showing substantive complexity (Aggressive Behaviour) likely reflects item redundancy rather than systematic gender bias. Nevertheless, researchers interested in item-level differential functioning should consider dedicated DIF analyses as a complementary approach.\u003c/p\u003e\n\u003cp\u003eSeveral broader limitations and future directions should be noted. First, the study relied exclusively on parent-report data, which may introduce reporting bias \u0026mdash; a concern particularly relevant given the theoretical arguments outlined in the introduction regarding gender-differentiated parental interpretation of children\u0026apos;s behaviour. Triangulating CBCL assessments with additional informants such as teachers would help disentangle measurement equivalence from informant-specific biases. Second, the present analysis assessed measurement invariance cross-sectionally within each CBCL version. Longitudinal invariance testing was not feasible in this cohort because children transitioned from the preschool to the school-age CBCL as they aged across ELPI waves (2010, 2012, 2017), meaning that different items were administered at different time points \u0026mdash; a prerequisite violation for longitudinal MI. Future research using cohorts assessed with the same CBCL version across multiple waves could address this gap. Third, although the sample is nationally representative of Chile, generalizability to other cultural or regional contexts should not be assumed. Cross-national studies across Latin America could evaluate whether the CBCL demonstrates consistent measurement properties in populations with different gender socialization norms and reporting practices. Fourth, although age was controlled as a covariate on the latent factors, the possibility that the CBCL\u0026rsquo;s measurement properties vary across narrower age bands within each version was not examined.\u003c/p\u003e\n\u003cp\u003eFuture research could employ MIMIC models or age-subgroup analyses to evaluate age-related differential item functioning. Finally, the present study tested measurement invariance of the CBCL\u0026apos;s established internalising-externalising structure. Alternative structural specifications \u0026mdash; such as bifactor models or exploratory structural equation models \u0026mdash; may provide additional insights into the dimensionality of the CBCL and could be evaluated in future work\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThe parent-reported CBCL demonstrated configural, metric, scalar and strict invariance across gender in a nationally representative Chilean sample, for both the preschool and school-age versions. These findings support the use of the CBCL for valid group comparisons of internalising and externalising problems across key social dimensions in early and middle childhood and strengthen its utility for developmental and policy-relevant research in Latin American settings.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eFunding\u0026nbsp;\u003c/h2\u003e\n\u003cp\u003eThe author acknowledges funding provided by the National Agency for Research and Development (ANID), BECAS/DOCTORADO BECAS CHILE/2021\u0026ndash;72220327.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eAbufhele, A., Contreras, D., Puentes, E., Telias, A., \u0026amp; Valdebenito, N. (2022). Socioeconomic gradients in child development: Evidence from a Chilean longitudinal study 2010\u0026ndash;2017. \u003cem\u003eAdvances in Life Course Research, 52\u003c/em\u003e, 100451. https://doi.org/10.1016/j.alcr.2021.100451\u003c/li\u003e\n \u003cli\u003eAchenbach, T. M. (2019). International findings with the Achenbach system of empirically based assessment (ASEBA): Applications to clinical services, research, and training. \u003cem\u003eChild and Adolescent Psychiatry and Mental Health, 13\u003c/em\u003e(1), 30. https://doi.org/10.1186/s13034-019-0291-2\u003c/li\u003e\n \u003cli\u003eAchenbach, T. M., Ivanova, M. Y., Rescorla, L. A., Turner, L. V., \u0026amp; Althoff, R. R. (2016). Internalizing/externalising problems: Review and recommendations for clinical and research applications. \u003cem\u003eJournal of the American Academy of Child \u0026amp; Adolescent Psychiatry, 55\u003c/em\u003e(8), 647\u0026ndash;656. https://doi.org/10.1016/j.jaac.2016.05.012\u003c/li\u003e\n \u003cli\u003eAchenbach, T. M., \u0026amp; Rescorla, L. A. (2001). \u003cem\u003eManual for the ASEBA school-age forms \u0026amp; profiles: An integrated system of multi-informant assessment\u003c/em\u003e. University of Vermont, Research Center for Children, Youth, \u0026amp; Families.\u003c/li\u003e\n \u003cli\u003eBehrman, J., Bravo, D., \u0026amp; Urz\u0026uacute;a, S. (2010). Encuesta Longitudinal de la Primera Infancia: Aspectos metodol\u0026oacute;gicos y primeros resultados. Retrieved from http://observatorio.ministeriodesarrollosocial.gob.cl/elpi-primera-ronda\u003c/li\u003e\n \u003cli\u003eBehrman, J. R., Contreras, D., Palma, I., \u0026amp; Puentes, E. (2017). Wealth disparities for early childhood anthropometrics and skills: Evidence from Chilean longitudinal data (PIER Working Paper 17-019). Penn Institute for Economic Research, University of Pennsylvania.\u003c/li\u003e\n \u003cli\u003eBenninger, T. L., Norris, M., \u0026amp; Witwer, A. N. (2025). Measurement invariance of the child behavior checklist in autistic toddlers. \u003cem\u003eResearch in Autism Spectrum Disorders, 119\u003c/em\u003e, 102500. https://doi.org/10.1016/j.rasd.2024.102500\u003c/li\u003e\n \u003cli\u003eBentler, P. M. (1990). Comparative fit indexes in structural models. \u003cem\u003ePsychological Bulletin, 107\u003c/em\u003e(2), 238\u0026ndash;246. https://doi.org/10.1037/0033-2909.107.2.238\u003c/li\u003e\n \u003cli\u003eBrowne, M. W., \u0026amp; Cudeck, R. (1992). Alternative ways of assessing model fit. \u003cem\u003eSociological Methods \u0026amp; Research, 21\u003c/em\u003e(2), 230\u0026ndash;258. https://doi.org/10.1177/0049124192021002005\u003c/li\u003e\n \u003cli\u003eChaplin, T. M., \u0026amp; Aldao, A. (2013). Gender differences in emotion expression in children: A meta-analytic review. \u003cem\u003ePsychological Bulletin, 139\u003c/em\u003e(4), 735\u0026ndash;765. https://doi.org/10.1037/a0030737\u003c/li\u003e\n \u003cli\u003eChen, F. F. (2007). Sensitivity of goodness of fit indexes to lack of measurement invariance. \u003cem\u003eStructural Equation Modeling: A Multidisciplinary Journal, 14\u003c/em\u003e(3), 464\u0026ndash;504. https://doi.org/10.1080/10705510701301834\u003c/li\u003e\n \u003cli\u003eCrick, N. R., \u0026amp; Grotpeter, J. K. (1995). Relational aggression, gender, and social-psychological adjustment. \u003cem\u003eChild Development, 66\u003c/em\u003e(3), 710\u0026ndash;722. https://doi.org/10.2307/1131945\u003c/li\u003e\n \u003cli\u003eDekker, M. C., Ferdinand, R. F., Van Lang, N. D., Bongers, I. L., Van Der Ende, J., \u0026amp; Verhulst, F. C. (2007). Developmental trajectories of depressive symptoms from early childhood to late adolescence: Gender differences and adult outcome. \u003cem\u003eJournal of Child Psychology and Psychiatry, 48\u003c/em\u003e(7), 657\u0026ndash;666. https://doi.org/10.1111/j.1469-7610.2007.01742.x\u003c/li\u003e\n \u003cli\u003eFergusson, D. M., Horwood, L. J., \u0026amp; Ridder, E. M. (2005). Show me the child at seven: The consequences of conduct problems in childhood for psychosocial functioning in adulthood. \u003cem\u003eJournal of Child Psychology and Psychiatry, 46\u003c/em\u003e(8), 837\u0026ndash;849. https://doi.org/10.1111/j.1469-7610.2005.01468.x\u003c/li\u003e\n \u003cli\u003eFlora, D. B., \u0026amp; Curran, P. J. (2004). An empirical evaluation of alternative methods of estimation for confirmatory factor analysis with ordinal data. \u003cem\u003ePsychological Methods, 9\u003c/em\u003e(4), 466\u0026ndash;491. https://doi.org/10.1037/1082-989X.9.4.466\u003c/li\u003e\n \u003cli\u003eGross, D., Fogg, L., Young, M., Ridge, A., Cowell, J. M., Richardson, R., \u0026amp; Sivan, A. (2006). The equivalence of the child behavior checklist/1\u0026frac12;\u0026ndash;5 across parent race/ethnicity, income level, and language. \u003cem\u003ePsychological Assessment, 18\u003c/em\u003e(3), 313\u0026ndash;323. https://doi.org/10.1037/1040-3590.18.3.313\u003c/li\u003e\n \u003cli\u003eHankin, B. L., \u0026amp; Abramson, L. Y. (2001). Development of gender differences in depression: An elaborated cognitive vulnerability\u0026ndash;transactional stress theory. \u003cem\u003ePsychological Bulletin, 127\u003c/em\u003e(6), 773\u0026ndash;796. https://doi.org/10.1037/0033-2909.127.6.773\u003c/li\u003e\n \u003cli\u003eHu, L., \u0026amp; Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. \u003cem\u003eStructural Equation Modeling: A Multidisciplinary Journal, 6\u003c/em\u003e(1), 1\u0026ndash;55. https://doi.org/10.1080/10705519909540118\u003c/li\u003e\n \u003cli\u003eIvanova, M. Y., Achenbach, T. M., Dumenci, L., Rescorla, L. A., Almqvist, F., Weintraub, S., Bilenberg, N., Bird, H., Chen, W. J., Dobrean, A., D\u0026ouml;pfner, M., Erol, N., Fombonne, E., Fonseca, A. C., Frigerio, A., Grietens, H., Hannesd\u0026oacute;ttir, H., Kanbayashi, Y., Lambert, M., ... Verhulst, F. C. (2007). Testing the 8-syndrome structure of the child behavior checklist in 30 societies. \u003cem\u003eJournal of Clinical Child and Adolescent Psychology, 36\u003c/em\u003e(3), 405\u0026ndash;417. https://doi.org/10.1080/15374410701444363\u003c/li\u003e\n \u003cli\u003eIvanova, M. Y., Achenbach, T. M., Rescorla, L. A., Harder, V. S., Ang, R. P., Bilenberg, N., Bjarnadottir, G., Capron, C., De Pauw, S. S., Dias, P., Dobrean, A., D\u0026ouml;pfner, M., Erol, N., Fombonne, E., Frigerio, A., Hannesdottir, H., Kanbayashi, Y., Lambert, M. C., Larsson, B., ... Verhulst, F. C. (2010). Preschool psychopathology reported by parents in 23 societies: Testing the seven-syndrome model of the child behavior checklist for ages 1.5\u0026ndash;5. \u003cem\u003eJournal of the American Academy of Child \u0026amp; Adolescent Psychiatry, 49\u003c/em\u003e(12), 1215\u0026ndash;1224. https://doi.org/10.1016/j.jaac.2010.08.019\u003c/li\u003e\n \u003cli\u003eKline, R. B. (2023). \u003cem\u003ePrinciples and practice of structural equation modeling\u003c/em\u003e (5th ed.). Guilford Publications.\u003c/li\u003e\n \u003cli\u003eKonold, T. R., Walthall, J. C., \u0026amp; Pianta, R. C. (2004). The behavior of child behavior ratings: Measurement structure of the child behavior checklist across time, informants, and child gender. \u003cem\u003eBehavioral Disorders, 29\u003c/em\u003e(4), 372\u0026ndash;383. https://doi.org/10.1177/019874290402900404\u003c/li\u003e\n \u003cli\u003eLecannelier, F., P\u0026eacute;rez Ewert, J. C., Groissman, S., Gallardo, D., Bardet, A. M., Bascu\u0026ntilde;\u0026aacute;n, A., \u0026amp; Rodr\u0026iacute;guez, J. (2014). Validaci\u0026oacute;n del Inventario de Conductas Infantiles para ni\u0026ntilde;os de entre 1\u0026frac12;\u0026ndash;5 a\u0026ntilde;os (CBCL 1\u0026frac12;\u0026ndash;5) en la Ciudad de Santiago de Chile. \u003cem\u003eUniversitas Psychologica, 13\u003c/em\u003e(2), 491\u0026ndash;500. https://doi.org/10.11144/Javeriana.UPSY13-2.vici\u003c/li\u003e\n \u003cli\u003eLittle, T. D., Rhemtulla, M., Gibson, K., \u0026amp; Schoemann, A. M. (2013). Why the items versus parcels controversy needn\u0026apos;t be one. \u003cem\u003ePsychological Methods, 18\u003c/em\u003e(3), 285\u0026ndash;300. https://doi.org/10.1037/a0033266\u003c/li\u003e\n \u003cli\u003eLittle, T. D., Rioux, C., Odejimi, O. A., \u0026amp; Stickley, Z. L. (2022). \u003cem\u003eParceling in structural equation modeling: A comprehensive introduction for developmental scientists\u003c/em\u003e. Cambridge University Press. https://doi.org/10.1017/9781316882054\u003c/li\u003e\n \u003cli\u003eMarsh, H. W., Hau, K.-T., \u0026amp; Wen, Z. (2004). In search of golden rules: Comment on hypothesis-testing approaches to setting cutoff values for fit indexes and dangers in overgeneralizing Hu and Bentler\u0026apos;s (1999) findings. \u003cem\u003eStructural Equation Modeling, 11\u003c/em\u003e(3), 320\u0026ndash;341. https://doi.org/10.1207/s15328007sem1103_2\u003c/li\u003e\n \u003cli\u003eMeade, A. W., \u0026amp; Kroustalis, C. M. (2006). Problems with item parceling for confirmatory factor analytic tests of measurement invariance. \u003cem\u003eOrganizational Research Methods, 9\u003c/em\u003e(3), 369\u0026ndash;403. https://doi.org/10.1177/1094428105283384\u003c/li\u003e\n \u003cli\u003eMellado, R. (2025). Bidirectional relationship between language ability and internalising/externalising behaviour from early to late childhood: Findings from a Chilean cohort. \u003cem\u003eResearch on Child and Adolescent Psychopathology, 53\u003c/em\u003e, 597\u0026ndash;611. https://doi.org/10.1007/s10802-025-01287-w\u003c/li\u003e\n \u003cli\u003eMillsap, R. E., \u0026amp; Yun-Tein, J. (2004). Assessing factorial invariance in ordered-categorical measures. \u003cem\u003eMultivariate Behavioral Research, 39\u003c/em\u003e(3), 479\u0026ndash;515. https://doi.org/10.1207/S15327906MBR3903_4\u003c/li\u003e\n \u003cli\u003eMorales, M. F., Girard, L.-C., Sawrikar, V., \u0026amp; MacBeth, A. (2023). An examination of behavioural and emotional problems in children exposed prenatally to the 27F Chilean earthquake: Findings from the ELPI cohort. \u003cem\u003eSocial Psychiatry and Psychiatric Epidemiology, 58\u003c/em\u003e(7), 1065\u0026ndash;1073. https://doi.org/10.1007/s00127-023-02453-5\u003c/li\u003e\n \u003cli\u003eMorales, M. F., MacBeth, A., Nagin, D., \u0026amp; Girard, L.-C. (2024). Developmental trajectories of aggression, hyperactivity/inattention, and anxious depressed mood: Co-occurring problems within a Chilean context. \u003cem\u003eCurrent Psychology, 43\u003c/em\u003e(5), 3928\u0026ndash;3943. https://doi.org/10.1007/s12144-022-03198-x\u003c/li\u003e\n \u003cli\u003eMuth\u0026eacute;n, B., \u0026amp; Muth\u0026eacute;n, L. (2017). Mplus. In \u003cem\u003eHandbook of item response theory\u003c/em\u003e (pp. 507\u0026ndash;518). Chapman \u0026amp; Hall/CRC.\u003c/li\u003e\n \u003cli\u003eNajman, J. M., Williams, G. M., Nikles, J., Spence, S., Bor, W., O\u0026apos;Callaghan, M., Le Brocque, R., Andersen, M. J., \u0026amp; Shuttlewood, G. (2001). Bias influencing maternal reports of child behaviour and emotional state. \u003cem\u003eSocial Psychiatry and Psychiatric Epidemiology, 36\u003c/em\u003e(4), 186\u0026ndash;194. https://doi.org/10.1007/s001270170062\u003c/li\u003e\n \u003cli\u003ePandolfi, V., Magyar, C. I., \u0026amp; Dill, C. A. (2009). Confirmatory factor analysis of the child behavior checklist 1.5\u0026ndash;5 in a sample of children with autism spectrum disorders. \u003cem\u003eJournal of Autism and Developmental Disorders, 39\u003c/em\u003e, 986\u0026ndash;995. https://doi.org/10.1007/s10803-009-0709-z\u003c/li\u003e\n \u003cli\u003ePeverill, M., Dirks, M. A., Narvaja, T., Herts, K. L., Comer, J. S., \u0026amp; McLaughlin, K. A. (2021). Socioeconomic status and child psychopathology in the United States: A meta-analysis of population-based studies. \u003cem\u003eClinical Psychology Review, 83\u003c/em\u003e, 101933. https://doi.org/10.1016/j.cpr.2020.101933\u003c/li\u003e\n \u003cli\u003ePutnick, D. L., \u0026amp; Bornstein, M. H. (2016). Measurement invariance conventions and reporting: The state of the art and future directions for psychological research. \u003cem\u003eDevelopmental Review, 41\u003c/em\u003e, 71\u0026ndash;90. https://doi.org/10.1016/j.dr.2016.06.004\u003c/li\u003e\n \u003cli\u003eRosenfield, S., \u0026amp; Mouzon, D. (2013). Gender and mental health. In C. S. Aneshensel, J. C. Phelan, \u0026amp; A. Bierman (Eds.), \u003cem\u003eHandbook of the sociology of mental health\u003c/em\u003e (pp. 277\u0026ndash;296). Springer. https://doi.org/10.1007/978-94-007-4276-5_14\u003c/li\u003e\n \u003cli\u003eRosseel, Y. (2012). lavaan: An R package for structural equation modeling. \u003cem\u003eJournal of Statistical Software, 48\u003c/em\u003e(2), 1\u0026ndash;36. https://doi.org/10.18637/jss.v048.i02\u003c/li\u003e\n \u003cli\u003eSterba, S. K., Prinstein, M. J., \u0026amp; Cox, M. J. (2007). Trajectories of internalizing problems across childhood: Heterogeneity, external validity, and gender differences. \u003cem\u003eDevelopment and Psychopathology, 19\u003c/em\u003e(2), 345\u0026ndash;366. https://doi.org/10.1017/S0954579407070174\u003c/li\u003e\n \u003cli\u003eStewart, L. C., Asadi, S., Rodriguez-Seijas, C., Wilson, S., Michelini, G., Kotov, R., Cicero, D. C., \u0026amp; Olino, T. M. (2024). Measurement invariance of the child behavior checklist (CBCL) across race/ethnicity and sex in the adolescent brain and cognitive development (ABCD) study. \u003cem\u003ePsychological Assessment, 36\u003c/em\u003e(8), 441\u0026ndash;452. https://doi.org/10.1037/pas0001319\u003c/li\u003e\n \u003cli\u003eTucker, L. R., \u0026amp; Lewis, C. (1973). A reliability coefficient for maximum likelihood factor analysis. \u003cem\u003ePsychometrika, 38\u003c/em\u003e(1), 1\u0026ndash;10. https://doi.org/10.1007/BF02291170\u003c/li\u003e\n \u003cli\u003evan der Sluis, S., Polderman, T. J., Neale, M. C., Verhulst, F. C., Posthuma, D., \u0026amp; Dieleman, G. C. (2017). Sex differences and gender-invariance of mother-reported childhood problem behavior. \u003cem\u003eInternational Journal of Methods in Psychiatric Research, 26\u003c/em\u003e(3), e1498. https://doi.org/10.1002/mpr.1498\u003c/li\u003e\n \u003cli\u003eViola, L., Garrido, G., \u0026amp; Rescorla, L. (2011). Testing multicultural robustness of the child behavior checklist in a national epidemiological sample in Uruguay. \u003cem\u003eJournal of Abnormal Child Psychology, 39\u003c/em\u003e(6), 897\u0026ndash;908. https://doi.org/10.1007/s10802-011-9500-z\u003c/li\u003e\n \u003cli\u003eYarnell, L. M., Sargeant, M. N., Prescott, C. A., Tilley, J. L., Farver, J. A. M., Mednick, S. A., Venables, P. H., Raine, A., \u0026amp; Luczak, S. E. (2013). Measurement invariance of internalizing and externalising behavioral syndrome factors in a non-western sample. \u003cem\u003eAssessment, 20\u003c/em\u003e(5), 642\u0026ndash;655. https://doi.org/10.1177/1073191113498114\u003c/li\u003e\n \u003cli\u003eYuan, K.-H., \u0026amp; Bentler, P. M. (2000). Three likelihood-based methods for mean and covariance structure analysis with nonnormal missing data. \u003cem\u003eSociological Methodology, 30\u003c/em\u003e(1), 165\u0026ndash;200. https://doi.org/10.1111/0081-1750.00078\u003c/li\u003e\n \u003cli\u003eYu, C. Y. (2002). Evaluation of model fit indices for latent variable models with categorical and continuous outcomes. Paper presented at the Annual Conference of the American Educational Research Association, New Orleans, LA.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"University College London","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Child Behavior Checklist (CBCL), measurement invariance, confirmatory factor analysis, gender differences, early childhood, Chile","lastPublishedDoi":"10.21203/rs.3.rs-9431030/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9431030/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eMeasurement invariance of the Child Behavior Checklist (CBCL) was examined across gender in a nationally representative sample of Chilean children. Using data from the Chilean Longitudinal Survey of Early Childhood (ELPI; n\u0026thinsp;\u0026gt;\u0026thinsp;10,000), equivalence of the internalising and externalising domains across boys and girls was evaluated for both the preschool (1.5\u0026ndash;5 years) and school-age (6\u0026ndash;12 years) versions. Multigroup confirmatory factor analyses with age controlled as a covariate supported full configural, metric, scalar, and strict invariance across gender in both versions. Two sensitivity analyses\u0026mdash;an alternative parcelling scheme and item-level invariance testing within each syndrome subscale\u0026mdash;confirmed the robustness of these findings. These results indicate that gender comparisons in CBCL scores at each developmental stage are not confounded by differential instrument functioning, providing psychometric support for gender-based comparisons in CBCL research, including longitudinal analyses that rely on valid within-wave gender comparisons, in Latin American settings.\u003c/p\u003e","manuscriptTitle":"Measurement Invariance of the Child Behavior Checklist (CBCL) Across Gender in a Chilean Child Cohort","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-19 12:46:31","doi":"10.21203/rs.3.rs-9431030/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"82a589be-af0e-4d8a-8c97-5eab31d0cf69","owner":[],"postedDate":"April 19th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":66496326,"name":"Psychology"}],"tags":[],"updatedAt":"2026-04-19T12:46:31+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-19 12:46:31","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9431030","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9431030","identity":"rs-9431030","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}