Confirmatory Factor Analysis and Multidimensional Scaling for Construct Validation of Cognitive Abilities

Aerobic fitness and academic achievement: Disentangling the indirect role of executive functions and intelligence

Research in children points to aerobic fitness as a source of individual differences in academic achievement. By examining the indirect effects of executive functions (EF) and intelligence on the relationship between aerobic fitness and academic achievement, the present study provides novel insight about the cognitive mechanisms underlying this relationship. 218 children (8-10 years) completed the following assessments: (i) a VO2max test to assess aerobic fitness; (ii) four tasks tapping components of EF (i.e., inhibition and cognitive flexibility); (iii) sub-tests of the Kaufman Brief Intelligence Test to assess fluid and crystallized intelligence; and (iv) sub-tests of arithmetic, spelling, and reading achievement (WRAT 3rd edition). Structural equation modeling (SEM) was conducted to examine the indirect role of EF and intelligence on the relationship between aerobic fitness and sub-domains of academic achievement. Covariate analyses included age, pubertal timing, and socio-economic status. Preliminary analysis via linear regression showed a direct effect of aerobic fitness on arithmetic achievement, whereas no effect was observed on spelling and reading achievement. Importantly, multiple mediation SEM revealed the direct effect of aerobic fitness on arithmetic achievement disappeared after accounting for the indirect effects of EF, whereas intelligence did not contribute significantly on this complex mediation process. Moreover, among EF components, cognitive flexibility, was the main driver of the relationship between aerobic fitness and arithmetic achievement. Unpacking which components of EF and intelligence affect the link between aerobic fitness and academic achievement, holds the promise of better understanding the heterogeneity still present in the literature.

Cardiff Online Cognitive Assessment in a National Sample: Cross-Sectional Web-Based Study

BACKGROUND

Psychiatric disorders are associated with cognitive impairment. We have developed a web-based, 9-task cognitive battery to measure the core domains affected in people with psychiatric disorders. To date, this assessment has been used to collect data on a clinical sample of participants with psychiatric disorders.

OBJECTIVE

The aims of this study were (1) to establish a briefer version of the battery (called the Cardiff Online Cognitive Assessment [CONCA]) that can give a valid measure of cognitive ability ("g") and (2) to collect normative data and demonstrate CONCA's application in a health population sample.

METHODS

Based on 6 criteria and data from our previous study, we selected 5 out of the original 9 tasks to include in CONCA. These included 3 core tasks that were sufficient to derive a measure of "g" and 2 optional tasks. Participants from a web-based national cohort study (HealthWise Wales) were invited to complete CONCA. Completion rates, sample characteristics, performance distributions, and associations between cognitive performance and demographic characteristics and mental health measures were examined.

RESULTS

A total of 3679 participants completed at least one CONCA task, of which 3135 completed all 3 core CONCA tasks. Performance on CONCA was associated with age (B=-0.05, SE 0.002; P<.001), device (tablet computer: B=-0.26, SE 0.05; P<.001; smartphone: B=-0.46, SE 0.05; P<.001), education (degree: B=1.68, SE 0.14; P<.001), depression symptoms (B=-0.04, SE 0.01; P<.001), and anxiety symptoms (B=-0.04, SE 0.01; P<.001).

CONCLUSIONS

CONCA provides a valid measure of "g," which can be derived using as few as 3 tasks that take no more than 15 minutes. Performance on CONCA showed associations with demographic characteristics in the expected direction and was associated with current depression and anxiety symptoms. The effect of device on cognitive performance is an important consideration for research using web-based assessments.

Responses to Raven matrices: Governed by visual complexity and centrality

Raven matrices are widely considered a pure test of cognitive abilities. Previous research has examined the extent to which cognitive strategies are predictive of the number of correct responses to Raven items. This study examined whether response times can be explained directly from the centrality and visual complexity of the matrix cells (edge density and perceived complexity). A total of 159 participants completed a 12-item version of the Raven Advanced Progressive Matrices. In addition to item number (an index of item difficulty), the findings demonstrated a positive correlation between the visual complexity of Raven items and both the mean response time and the number of fixations on the matrix (a strong correlate of response time). Moreover, more centrally placed cells as well as more complex cells received more fixations. It is concluded that response times on Raven matrices are impacted by low-level stimulus attributes, namely, visual complexity and eccentricity.

The Early-Onset Alzheimer's Disease Whole-Genome Sequencing Project: Study design and methodology

INTRODUCTION

Sequencing efforts to identify genetic variants and pathways underlying Alzheimer's disease (AD) have largely focused on late-onset AD although early-onset AD (EOAD), accounting for ∼10% of cases, is largely unexplained by known mutations, resulting in a lack of understanding of its molecular etiology.

METHODS

Whole-genome sequencing and harmonization of clinical, neuropathological, and biomarker data of over 5000 EOAD cases of diverse ancestries.

RESULTS

A publicly available genomics resource for EOAD with extensive harmonized phenotypes. Primary analysis will (1) identify novel EOAD risk loci and druggable targets; (2) assess local-ancestry effects; (3) create EOAD prediction models; and (4) assess genetic overlap with cardiovascular and other traits.

DISCUSSION

This novel resource complements over 50,000 control and late-onset AD samples generated through the Alzheimer's Disease Sequencing Project (ADSP). The harmonized EOAD/ADSP joint call will be available through upcoming ADSP data releases and will allow for additional analyses across the full onset range.

HIGHLIGHTS

Sequencing efforts to identify genetic variants and pathways underlying Alzheimer's disease (AD) have largely focused on late-onset AD although early-onset AD (EOAD), accounting for ∼10% of cases, is largely unexplained by known mutations. This results in a significant lack of understanding of the molecular etiology of this devastating form of the disease. The Early-Onset Alzheimer's Disease Whole-genome Sequencing Project is a collaborative initiative to generate a large-scale genomics resource for early-onset Alzheimer's disease with extensive harmonized phenotype data. Primary analyses are designed to (1) identify novel EOAD risk and protective loci and druggable targets; (2) assess local-ancestry effects; (3) create EOAD prediction models; and (4) assess genetic overlap with cardiovascular and other traits. The harmonized genomic and phenotypic data from this initiative will be available through NIAGADS.

Using dimensionality-reduction techniques to understand the organization of psychotic symptoms in persistent psychotic illness and first episode psychosis

Psychotic disorders are highly heterogeneous. Understanding relationships between symptoms will be relevant to their underlying pathophysiology. We apply dimensionality-reduction methods across two unique samples to characterize the patterns of symptom organization. We analyzed publicly-available data from 153 participants diagnosed with schizophrenia or schizoaffective disorder (fBIRN Data Repository and the Consortium for Neuropsychiatric Phenomics), as well as 636 first-episode psychosis (FEP) participants from the Prevention and Early Intervention Program for Psychosis (PEPP-Montreal). In all participants, the Scale for the Assessment of Positive Symptoms (SAPS) and Scale for the Assessment of Negative Symptoms (SANS) were collected. Multidimensional scaling (MDS) combined with cluster analysis was applied to SAPS and SANS scores across these two groups of participants. MDS revealed relationships between items of SAPS and SANS. Our application of cluster analysis to these results identified: 1 cluster of disorganization symptoms, 2 clusters of hallucinations/delusions, and 2 SANS clusters (asocial and apathy, speech and affect). Those reality distortion items which were furthest from auditory hallucinations had very weak to no relationship with hallucination severity. Despite being at an earlier stage of illness, symptoms in FEP presentations were similarly organized. While hallucinations and delusions commonly co-occur, we found that their specific themes and content sometimes travel together and sometimes do not. This has important implications, not only for treatment, but also for research-particularly efforts to understand the neurocomputational and pathophysiological mechanism underlying delusions and hallucinations.

Multidimensional Scaling of Cognitive Ability and Academic Achievement Scores

Multidimensional scaling (MDS) was used as an alternate multivariate procedure for investigating intelligence and academic achievement test score correlations. Correlation coefficients among Wechsler Intelligence Scale for Children, Fifth Edition (WISC-5) and Wechsler Individual Achievement Test, Third Edition (WIAT-III) validity sample scores and among Kaufman Assessment Battery for Children, Second Edition (KABC-II) and Kaufman Test of Educational Achievement, Second Edition (KTEA-2) co-norming sample scores were analyzed using multidimensional scaling (MDS). Three-dimensional MDS configurations were the best fit for interpretation in both datasets. Subtests were more clearly organized by CHC ability and academic domain instead of complexity. Auditory-linguistic, figural-visual, reading-writing, and quantitative-numeric regions were visible in all models. Results were mostly similar across different grade levels. Additional analysis with WISC-V and WIAT-III tests showed that content (verbal, numeric, figural) and response process facets (verbal, manual, paper-pencil) were also useful in explaining test locations. Two implications from this study are that caution may be needed when interpreting fluency scores across academic areas, and MDS provides more empirically based validity evidence regarding content and response mode processes.

Web-Based Cognitive Testing in Psychiatric Research: Validation and Usability Study

Background

Cognitive impairments are features of many psychiatric disorders and affect functioning. A barrier to cognitive research on psychiatric disorders is the lack of large cross-disorder data sets. However, the collection of cognitive data can be logistically challenging and expensive. Web-based collection may be an alternative; however, little is known about who does and does not complete web-based cognitive assessments for psychiatric research.

Objective

The aims of this study are to develop a web-based cognitive battery for use in psychiatric research, validate the battery against the Measurement and Treatment Research to Improve Cognition in Schizophrenia (MATRICS) Consensus Cognitive Battery, and compare the characteristics of the participants who chose to take part with those of the individuals who did not participate.

Methods

Tasks were developed by The Many Brains Project and selected to measure the domains specified by the MATRICS initiative. We undertook a cross-validation study of 65 participants with schizophrenia, bipolar disorder, depression, or no history of psychiatric disorders to compare the web-based tasks with the MATRICS Consensus Cognitive Battery. Following validation, we invited participants from 2 large ongoing genetic studies, which recruited participants with psychiatric disorders to complete the battery and evaluated the demographic and clinical characteristics of those who took part.

Results

Correlations between web-based and MATRICS tasks ranged between 0.26 and 0.73. Of the 961 participants, 887 (92.3%) completed at least one web-based task, and 644 (67%) completed all tasks, indicating adequate completion rates. Predictors of web-based participation included being female (odds ratio [OR] 1.3, 95% CI 1.07-1.58), ethnicity other than White European (OR 0.66, 95% CI 0.46-0.96), higher levels of education (OR 1.19, 95% CI 1.11-1.29), diagnosis of an eating disorder (OR 2.17, 95% CI 1.17-4) or depression and anxiety (OR 5.12, 95% CI 3.38-7.83), and absence of a diagnosis of schizophrenia (OR 0.59, 95% CI 0.35-0.94). Lower performance on the battery was associated with poorer functioning (B=−1.76, SE 0.26; P<.001).

Conclusions

Our findings offer valuable insights into the advantages and disadvantages of testing cognitive function remotely for mental health research.

Higher Education Students' Reflective Journal Writing and Lifelong Learning Skills: Insights From an Exploratory Sequential Study

Reflective journal (RJ) writing has been recognized as an effective pedagogical tool for nurturing students’ lifelong learning skills. With the paucity of empirical work on the dimensionality of reflective writing, this research sought to qualitatively analyze students’ RJ writing and design a generic reflection scheme for identifying dimensions of reflective thinking. Drawing on the theoretical scheme, another aim was to design and validate a questionnaire to measure students’ perceptions of their reflective writing experiences. The last aim was to quantitatively measure the link between perceived reflective writing and students’ tendency to use RJs in their future careers and personal lives. This exploratory sequential research included the following steps: First, experts’ review and analysis of 1312 RJ entries were attained. This step led to the design of a theoretical scheme of reflective writing and a 31-item questionnaire, used to gather data from 171 students (second-year pre-service teachers and third-year health managers). Partial Least Squares analysis corroborated the structure suggested by the theoretical scheme: two timelines–reflections regarding the current course assignments and those related to the student’s future development. Students’ tendency to use reflective skills in their future professional lives was highly connected to their long-term reflections, including learning experiences linked to academic, professional, personal, and multicultural development. The current study’s suggested validated generic scheme can be adapted and integrated into different curricula, thereby possibly increasing the potential of infusing RJ instructional strategies into higher education curricula, improving the quality of reflection in student journals, and promoting lifelong learning skills.

Students’ Perceived Constructivist Learning Environment

Abstract. Theoretical classifications suggest three key tenets of the constructivist learning environment: constructivist task, teacher-student interaction, and social activity. This study assessed two scales, measuring the perceptions of high school and college students of their constructivist learning environment. Facet theory (FT) approach with smallest space analysis (SSA) as well as confirmatory factor analysis (CFA) were used to confirm the structure of the constructivist learning theory and to test the structural validity of scores on two scales. In the first study, data were gathered by using the Constructivist Learning Environment Survey (CLES). In this study, compared with the CFA result, the SSA gave additional information showing the existence of the three theoretical key tenets which were absent from the factor analysis. This facet was derived from the constructivist learning theory, framed in the mapping sentence, purposely designed as part of the FT research strategy. In the second study, students’ perceptions of the occurrence of constructivist practices in higher education learning environments were assessed. The comparison between SSA and CFA showed that the CFA was limited to capture merely one facet at a time and failed to illustrate the correspondence between the content facets of the mapping sentence.

The unity and diversity of executive functions: A systematic review and re-analysis of latent variable studies

Confirmatory factor analysis (CFA) has been frequently applied to executive function measurement since first used to identify a three-factor model of inhibition, updating, and shifting; however, subsequent CFAs have supported inconsistent models across the life span, ranging from unidimensional to nested-factor models (i.e., bifactor without inhibition). This systematic review summarized CFAs on performance-based tests of executive functions and reanalyzed summary data to identify best-fitting models. Eligible CFAs involved 46 samples (N = 9,756). The most frequently accepted models varied by age (i.e., preschool = one/two-factor; school-age = three-factor; adolescent/adult = three/nested-factor; older adult = two/three-factor), and most often included updating/working memory, inhibition, and shifting factors. A bootstrap reanalysis simulated 5,000 samples from 21 correlation matrices (11 child/adolescent; 10 adult) from studies including the three most common factors, fitting seven competing models. Model results were summarized as the mean percent accepted (i.e., average rate at which models converged and met fit thresholds: CFI ≥ .90/RMSEA ≤ .08) and mean percent selected (i.e., average rate at which a model showed superior fit to other models: ΔCFI ≥ .005/.010/ΔRMSEA ≤ -.010/-.015). No model consistently converged and met fit criteria in all samples. Among adult samples, the nested-factor was accepted (41-42%) and selected (8-30%) most often. Among child/adolescent samples, the unidimensional model was accepted (32-36%) and selected (21-53%) most often, with some support for two-factor models without a differentiated shifting factor. Results show some evidence for greater unidimensionality of executive function among child/adolescent samples and both unity and diversity among adult samples. However, low rates of model acceptance/selection suggest possible bias toward the publication of well-fitting but potentially nonreplicable models with underpowered samples. (PsycINFO Database Record (c) 2018 APA, all rights reserved).

Psychometric properties of the Resilience Scale for Adults (RSA) and its relationship with life-stress, anxiety and depression in a Hispanic Latin-American community sample

Resilience is a multi-dimensional construct associated with health and well-being. At present, we do not yet have a valid, scientific instrument that is designed to evaluate adult resilience in Spanish-speaking countries and that accounts for family, social and individual components. This study aimed at investigating the construct and cross-cultural validity of the Resilience Scale for Adults (RSA) by combining Confirmatory Factor Analysis (CFA), Multidimensional Scaling (MDS) and Hierarchical Regression models in a Hispanic Latin-American group. A community sample of 805 adults answered the RSA, Spanish Language Stressful Life-Events checklist (SL-SLE), and the Hopkins Symptom Checklist-25 (HSCL-25). First-order CFA verified the six factors structure for the RSA (RMSEA = .037, SRMR = .047, CFI = .91, TLI = .90). Five RSA scales and total score have good internal consistency (scales α > .70; total score α = .90). Two second-order CFA verified the intrapersonal and interpersonal dimensions of the protector factors of resilience, as well as their commonality and uniqueness with affective symptoms (anxiety and depression). An exploratory MDS reproduced the relations of RSA items and factors at first and second-order levels against random simulated data, thereby providing initial evidence of its cross-cultural validity in a Spanish-speaking group. The Four-steps hierarchical model showed that the RSA scales are the strongest predictors of anxiety and depression–greater than gender, age, education and stressful life-events. Three RSA scales are significant unique predictors of affective symptoms. In addition, similar to findings in diverse cultural settings, resilience is positively associated with age but not with education. Women report higher scores of Social Resources and Social Competence and lower scores of Perception of the Self. In conclusion, this study demonstrates the construct and criterion-related validity of the RSA in broad, diverse and Spanish speaking sample.

Comparable Consistency, Coherence, and Commonality of Measures of Cognitive Functioning Across Adulthood

Increased age is associated with lower scores in many cognitive tests, but interpretation of those results is based on the assumption that the measurement at different ages is equivalent, such that the differences reflect quantitative rather than qualitative changes. The assumption of measurement equivalence was investigated by comparing adult age differences in the relations among alternative versions of the same tests, among different tests of the same ability, and among different cognitive abilities. Results from three independent data sets revealed that only modest age differences were apparent at each level, which implies that cognitive abilities have similar measurement properties at different ages in adulthood.

Scores in Space: Multidimensional Scaling of the WISC-V

The purpose of this study was to use multidimensional scaling (MDS) to investigate relations among scores from the standardization sample of the Wechsler Intelligence Scale for Children–Fifth edition (WISC-V; Wechsler, 2014). Nonmetric two-dimensional MDS maps were selected for interpretation. The most cognitively complex subtests and indexes were located near the center of the maps. Subtests were also grouped together in the way that they are organized in the primary and complementary indexes, and not necessarily on content surface features. Naming Speed and Symbol Translation scores may be better off kept as separate indexes, and Digit Span Sequencing appeared to add more complexity to the Digit Span subtest. One implication related to score interpretation is that General Ability Index (GAI) and Cognitive Proficiency Index (CPI) may be distinguished by the level of complexity of the subtests included in each index (i.e., more complex subtests included in GAI and less complex subtests included in the CPI).

Strong genetic overlap between executive functions and intelligence

Executive functions (EFs) are cognitive processes that control, monitor, and coordinate more basic cognitive processes. EFs play instrumental roles in models of complex reasoning, learning, and decision making, and individual differences in EFs have been consistently linked with individual differences in intelligence. By middle childhood, genetic factors account for a moderate proportion of the variance in intelligence, and these effects increase in magnitude through adolescence. Genetic influences on EFs are very high, even in middle childhood, but the extent to which these genetic influences overlap with those on intelligence is unclear. We examined genetic and environmental overlap between EFs and intelligence in a racially and socioeconomically diverse sample of 811 twins ages 7 to 15 years (M = 10.91, SD = 1.74) from the Texas Twin Project. A general EF factor representing variance common to inhibition, switching, working memory, and updating domains accounted for substantial proportions of variance in intelligence, primarily via a genetic pathway. General EF continued to have a strong, genetically mediated association with intelligence even after controlling for processing speed. Residual variation in general intelligence was influenced only by shared and nonshared environmental factors, and there remained no genetic variance in general intelligence that was unique of EF. Genetic variance independent of EF did remain, however, in a more specific perceptual reasoning ability. These results provide evidence that genetic influences on general intelligence are highly overlapping with those on EF. (PsycINFO Database Record

Self-Reported Personality Pathology Has Complex Structure and Imposing Simple Structure Degrades Test Information

Although simple structure is often used as a theoretical guideline for rotation of exploratory factor structures, its empirical foundation is sometimes not investigated. We examined the degree of simple structure in 2 self-report inventories advocated for the description of personality pathology. The multivariate complexity of the inventories was compared against simulated factorial solutions with increasing degrees of simple structure. Results revealed a high degree of complexity in both inventories, regardless of whether focusing on a few broad or many narrow factors and whether items or scales were analyzed. It is concluded that Thurstone's (1934) simple structure criteria are rarely fulfilled in multivariate personality pathology solutions, and, therefore, that alternate rotations represent equally justifiable ways to operationalize the space. Furthermore, using iterative test development to impose simple structure where none originally existed increased standard errors around the factor scores in the interstitial areas between imposed simple structure dimensions. In both simulated and real data, multidimensional item response theory test information functions based on complex solutions demonstrated that test information was uniformly distributed across the space. In contrast, in simple structure solutions, information was weak in areas of the space where nonconforming items had been removed. Implications are discussed.

Behind the wheel and on the map: Genetic and environmental associations between drunk driving and other externalizing behaviors

Drunk driving, a major contributor to alcohol-related mortality, has been linked to a variety of other alcohol-related (e.g., Alcohol Dependence, early age at first drink) and non-alcohol-related externalizing behaviors. In a sample of 517 same-sex twin pairs from the National Longitudinal Study of Adolescent Health, we examined 3 conceptualizations of the etiology of drunk driving in relation to other externalizing behaviors. A series of behavioral-genetic models found consistent evidence for drunk driving as a manifestation of genetic vulnerabilities toward a spectrum of alcohol-related and non-alcohol-related externalizing behaviors. Most notably, multidimensional scaling analyses produced a genetic "map" with drunk driving located near its center, supporting the strength of drunk driving's genetic relations with a broad range of externalizing behaviors. In contrast, nonshared environmental associations with drunk driving were weaker and more diffuse. Drunk driving may be a manifestation of genetic vulnerabilities toward a broad externalizing spectrum.

Socioeconomic Status Modifies Interest-Knowledge Associations among Adolescents

Researchers have recently taken a renewed interest in examining the patterns by which noncognitive traits and cognitive traits relate to one another. Few researchers, however, have examined the possibility that such patterns might differ according to environmental context. Using data from a nationally representative sample of approximately 375,000 students from 1,300 high schools in the United States, we examined the relations between socioeconomic status (SES), interests, and knowledge in eleven academic, vocational/professional, and recreational domains. We found little support for the hypothesis that SES-related differences in levels of interest mediate SES-related differences in levels of knowledge. In contrast, we found robust and consistent support for the hypothesis that SES moderates interest-knowledge associations. For 10 out of 11 of the knowledge domains examined, the interest-knowledge association was stronger for individuals living in higher SES contexts. Moderation persisted after controlling for an index of general intelligence. These findings are consistent with the hypothesis that low SES inhibits individuals from selectively investing their time and attention in learning experiences that are consistent with their interests.

Measuring Working Memory Capacity With Automated Complex Span Tasks

Individual differences in working memory capacity are related to a variety of behaviors both within and outside of the lab. Recently developed automated complex span tasks have contributed to increasing our knowledge concerning working memory capacity by making valid and reliable assessments freely available for use by researchers. Combining the samples from three testing locations yielded data from over 6,000 young adult participants who performed at least one of three such tasks (Operation, Symmetry, and Reading Span). Normative data are presented here for researchers interested in applying cutoffs for their own applications, and information on the validity and reliability of the tasks is also reported. In addition, the data were analyzed as a function of sex and college status. While automated complex span tasks are just one way to measure working memory capacity, the use of a standardized procedure for administration and scoring greatly facilitates comparison across studies.

Global and domain-specific changes in cognition throughout adulthood

Normative adult age-related decrements are well documented for many diverse forms of effortful cognitive processing. However, it is currently unclear whether each of these decrements reflects a distinct and independent developmental phenomenon, or, in part, a more global phenomenon. A number of studies have recently been published that show moderate to large magnitudes of positive relations among individual differences in rates of changes in different cognitive variables during adulthood. This suggests that a small number of common dimensions or even a single common dimension may underlie substantial proportions of individual differences in aging-related cognitive declines. This possibility was directly examined using data from 1,281 adults 18-95 years of age who were followed longitudinally over up to 7 years on 12 different measures of effortful processing. Multivariate growth curve models were applied to examine the dimensionality of individual differences in longitudinal changes. Results supported a hierarchical structure of aging-related changes, with an average of 39% of individual differences in change in a given variable attributable to global (domain-general) developmental processes, 33% attributable to domain-specific developmental processes (abstract reasoning, spatial visualization, episodic memory, and processing speed), and 28% attributable to test-specific developmental processes. Although it is often assumed that systematic and pervasive sources of cognitive decline only emerge in later adulthood, domain-general influences on change were apparent for younger (18-49 years), middle aged (50-69 years), and older (70-95 years) adults.

Combining nonlinear biometric and psychometric models of cognitive abilities

It is well-established that genetic factors account for large proportions of individual differences in multiple cognitive abilities. It is also well-established that individual differences in performance on many different cognitive ability measures are strongly correlated. Recent empirical investigations, however, have suggested two interesting qualifications to these well-established findings: Genetic variance in cognitive abilities is higher in richer home environments (gene-by-environment interaction), and common variance in different cognitive abilities is lower at higher levels of overall ability (nonlinear factor structure). Although they have been investigated independently, these two phenomena may interact, because richer environments are routinely associated with higher ability levels. Using simulation we demonstrate how un-modeled nonlinear factor structure can obscure interpretation of gene-by-environment interaction. We then reanalyze data from the National Collaborative Perinatal Project, previously used by Turkheimer et al. (2003; Psychol Science), with a two-step method to model both phenomena.