Do method effects generalize across traits (and what if they don't)?

OBJECTIVE: Multitrait-multimethod (MTMM) data can be analyzed with single-indicator confirmatory factor analysis (CFA-MTMM) models. Most single-indicator CFA-MTMM models imply-but do not allow testing-the restrictive assumption that method biases generalize (correlate) perfectly across different traits for a given method. METHOD: To examine the validity of this assumption, we identified and reviewed 20 published applications of multiple-indicator CFA-MTMM models, which allow testing this assumption. Based on simulated data, we demonstrate the consequences of violating the assumption of perfectly general method effects based on the CT-C(M - 1) approach. RESULTS: We extracted 111 heterotrait-monomethod method factor correlation estimates, which varied between |.01| and |1.0| (mean = .52) with most correlations being substantially smaller than |1|. The results of our review and simulations show that violations of the assumption of perfectly general method effects (a) are very common, (b) are difficult to detect based on model fit statistics, and (c) can lead to considerable bias in estimates of convergent validity, method specificity, reliability, and method factor correlations in single-indicator models. CONCLUSIONS: We recommend that researchers abandon the use of single-indicator CFA-MTMM models and that they use multiple-indicator CFA-MTMM models whenever possible.

Are Sluggish Cognitive Tempo, ADHD, and Oppositional Defiant Disorder Trait- or State-Like Constructs from Prekindergarten to Fourth Grade?

The objective was to determine the proportion of trait (consistency across occasions) and occasion-specific variance in sluggish cognitive tempo (SCT), attention deficit/hyperactivity disorder (ADHD)-inattention (IN), ADHD-hyperactivity/impulsivity (HI), and oppositional defiant disorder (ODD) symptom ratings. A single trait factor-multiple state factors model was applied to parent ratings of SCT, ADHD-IN, ADHD-HI, and ODD symptoms for 978 children (50% female) across prekindergarten (M = 4.90 years), kindergarten (M = 6.27 years), 1st-grade (M = 7.42 years), 2nd-grade (M = 8.45 years), and 4th-grade (M = 10.45 years) assessments. For the prekindergarten assessment, SCT, ADHD-IN, ADHD-HI, and ODD contained more occasion-specific than trait variance (54%, 64%, 56%, and 55% occasion-specific variance, respectively). In contrast, SCT, ADHD-IN, ADHD-HI, and ODD contained more trait than occasion-specific variance for the kindergarten through 4th-grade assessments (62%-72%, 65%-68%, 71%-75%, and 60%-69% trait variance, respectively). SCT, ADHD-IN, ADHD-HI, and ODD are slightly to moderately more state-like than trait-like during the prekindergarten developmental period but are more stable traits than fluctuating states from kindergarten to 4th grade. Findings indicate that, particularly after children start formal schooling, these psychopathology dimensions are primarily stable traits; implications for assessment are discussed.

The importance of face-to-face contact and reciprocal relationships and their associations with depressive symptoms and life satisfaction.

OBJECTIVE: The current study aimed to examine how patterns of interpersonal relational contexts (e.g., face-to-face or technology-based) and processes (e.g., initiated or accepted) relate to depressive symptomology and life satisfaction. METHODS: Participants were recruited through Amazon's Mechanical Turk (n = 962 adults [52.1% female; aged 18-78; 16.4% Non-White]). Quota sampling was used to closely match the sample demographics to that of the United States Census data. Latent class analyses (LCA) identified classes of interpersonal relations using the Multidimensional Interpersonal Relations Scale. Next, participants' responses on the Beck Depression Inventory and Satisfaction With Life Scale were examined to evaluate differences in depressive symptoms and life satisfaction across classes. RESULTS: LCA results supported a 4-class model, in which classes were characterized by patterns of relational contexts and processes: Class 1 (50.6%) engagement across all contexts (e.g., face-to-face) and processes (e.g., initiated); Class 2 (12.7%) engagement across all contexts and processes except Facebook; Class 3 (24.0%) engagement in all contexts and only passive processes; and Class 4 (12.7%) engagement in only technology-based contexts and passive processes. Membership in Classes 1 and 2 was associated with lower depressive symptomology and higher life satisfaction as compared to Classes 3 and 4. CONCLUSION: The findings suggest that patterns of relations differentially relate to depressive symptoms and life satisfaction. The findings suggest that multicontextual (e.g., face-to-face and technology-based) and reciprocal relationships with friends (e.g., initiating and accepting connections) may play an important role in the association between interpersonal relations with life satisfaction and depressive symptoms.

A latent state-trait model for analyzing states, traits, situations, method effects, and their interactions.

OBJECTIVE: Studies on situation and Person × Situation interaction effects often use only one method of measurement, so that the extent to which the effects may be method-specific cannot be determined. We introduce a new multimethod latent state-trait model for random and fixed situations (MM-LST-RF), which allows examining person, situation, and Person × Situation interaction effects in the context of multimethod measurement designs (e.g., studies with multiple reporters), and discuss potential areas of application for the new model in personality research. METHOD: The new model allows analyzing novel features of multimethod, multi-situation data, including (a) the convergent validity and method specificity of trait (person) effects, situation effects, and Person × Situation interaction effects; (b) the degree of situation specificity of method effects; and (c) potential Method × Situation interactions. An application to smoker's affect (N = 235; 57% female; 93% Caucasian) before and after quitting smoking is presented with positively and negatively worded items as methods. RESULTS: The MM-LST-RF model fit the smoking data well. Method specificity of many effects was high. CONCLUSIONS: The MM-LST-RF model provides researchers with a new framework for testing method specificity of person, situation, and interaction effects.

Trait and State Variance in Multi-Informant Assessments of ADHD and Academic Impairment in Spanish First-Grade Children.

OBJECTIVE: We examined the stable trait and variable state components of ADHD-inattention (IN), ADHD-hyperactivity/impulsivity (HI), and academic impairment (AI) dimensions using mothers', fathers', primary and secondary teachers' ratings of children's behavior at home and school. We also examined between-informant agreement with regard to trait and state components. METHOD: Mothers, fathers, primary and secondary teachers rated HI, IN, and AI in N = 758 Spanish first grade children (55% boys) over three measurement occasions across 12 months. RESULTS: Latent state-trait analyses revealed that mothers', fathers', and primary teachers' (but not secondary teachers') ratings reflected more trait variance for ADHD-HI (M = 73%), ADHD-IN (M = 74%), and AI (M = 76%) than occasion-specific variance (M = 27%, M = 26%, and M = 24%, respectively). Fathers' ratings shared a meaningful level of trait variance with mothers' ratings of ADHD-HI and ADHD-IN (range 78% to 82%), whereas primary and secondary teachers' ratings shared lower levels of trait variance with mothers' ratings (range 41% to 63%). The trait components of fathers', primary teachers', and secondary teachers' ratings of AI showed high levels of convergence with mothers' ratings (88%, 70%, and 59% respectively). CONCLUSIONS: ADHD symptom reports reflect both trait (48 to 86%) and state (14 to 53%) variance components. The lower amount of shared variability between home and school suggests the setting-specificity of trait and state components of ADHD symptoms. Our findings indicate that ADHD symptom reports may reflect context-specific traits, suggesting the importance of differentiating and targeting ADHD behaviors across different settings.

Mechanisms in the relation between GABRA2 and adolescent externalizing problems.

Conduct problems, alcohol problems and hyperactive-inattentive symptoms co-occur at a high rate and are heritable in adolescence. The γ-aminobutyric acid A receptor, α2 gene (GABRA2) is associated with a broad spectrum of externalizing problems and disinhibitory-related traits. The current study tested whether two important forms of disinhibition in adolescence, impulsivity and sensation seeking, mediated the effects of GABRA2 on hyperactive-inattentive symptoms, conduct problems, and alcohol problems. Participants were assessed at two waves (11-17 and 12-18 years old; N = 292). Analyses used the GABRA2 SNP, rs279858, which tags the two complementary (yin-yang) GABRA2 haplotypes. Multiple informants reported on adolescents' impulsivity and sensation seeking and adolescents self-reported their hyperactive-inattentive symptoms, conduct problems and lifetime alcohol problems. Impulsivity mediated the effect of GABRA2 on alcohol problems, hyperactive-inattentive symptoms, and conduct problems, whereas sensation seeking mediated the effect of GABRA2 on alcohol problems (AA/AG genotypes conferred risk). GABRA2 directly predicted adolescent alcohol problems, but the GG genotype conferred risk. Results suggest that there may be multiple pathways of risk from GABRA2 to adolescent externalizing problems, and suggest important avenues for future research.

A theory of states and traits--revised.

We present a revision of latent state-trait (LST-R) theory with new definitions of states and traits. This theory applies whenever we study the consistency of behavior, its variability, and its change over time. States and traits are defined in terms of probability theory. This allows for a seamless transition from theory to statistical modeling of empirical data. LST-R theory not only gives insights into the nature of latent variables but it also takes into account four fundamental facts: Observations are fallible, they never happen in a situational vacuum, they are always made using a specific method of observations, and there is no person without a past. Although the first fact necessitates considering measurement error, the second fact requires allowances for situational fluctuations. The third fact implies that, in the first place, states and traits are method specific. Furthermore, compared to the previous version of LST theory (see, e.g., Steyer et al. 1992 , 1999 ), our revision is based on the notion of a person-at-time-t. The new definitions in LST-R theory have far-reaching implications that not only concern the properties of states, traits, and the associated concepts of measurement errors and state residuals, but also are related to the analysis of states and traits in longitudinal observational and intervention studies.

Distinguishing state variability from trait change in longitudinal data: the role of measurement (non)invariance in latent state-trait analyses.

Researchers analyzing longitudinal data often want to find out whether the process they study is characterized by (1) short-term state variability, (2) long-term trait change, or (3) a combination of state variability and trait change. Classical latent state-trait (LST) models are designed to measure reversible state variability around a fixed set-point or trait, whereas latent growth curve (LGC) models focus on long-lasting and often irreversible trait changes. In the present article, we contrast LST and LGC models from the perspective of measurement invariance testing. We show that establishing a pure state-variability process requires (1) the inclusion of a mean structure and (2) establishing strong factorial invariance in LST analyses. Analytical derivations and simulations demonstrate that LST models with noninvariant parameters can mask the fact that a trait-change or hybrid process has generated the data. Furthermore, the inappropriate application of LST models to trait change or hybrid data can lead to bias in the estimates of consistency and occasion specificity, which are typically of key interest in LST analyses. Four tips for the proper application of LST models are provided.

Comparing the neuropsychiatric inventory and the revised memory and behavior problems checklist for associations with caregiver burden and depressive symptoms.

BACKGROUND: Few empirical studies compare the ability of prominent measures of behavioral and psychological symptoms of dementia (BPSD) to explain key caregiver outcomes. We compared the respective abilities of the Neuropsychiatric Inventory (NPI) and the Revised Memory and Behavior Problems Checklist (RMBPC) to detect associations between BPSD and caregiver depressive symptoms. Our results may facilitate measurement decisions for researchers and clinicians. METHODS: The NPI and the RMBPC, which measure BPSD frequency and corresponding caregiver appraisals, were administered to 164 caregivers of persons with dementia to compare the measures' ability to explain caregiver depressive symptoms. Depressive symptoms were measured using the Geriatric Depression Scale (GDS), and caregiver burden was measured (using the Zarit Burden Interview) as a mediator between BPSD frequency/appraisal and caregiver depressive symptoms. Path analysis using Mplus facilitated the comparison between the RMBPC and the NPI. RESULTS: Significant indirect associations were present when NPI frequency, NPI appraisal, RMBPC frequency, and RMBPC appraisal were modeled separately with burden and depressive symptoms, although indirect relationships are not associated with increases in R(2). Only RMBPC appraisal produced both a significant direct association with depressive symptoms and a significant increase in R(2) when modeled separately (ß = 0.24, p < 0.01; ΔR(2) = 0.04, p < 0.05). When all independent variables were modeled together, only RMBPC appraisal demonstrated significant direct (ß = 0.23, p < 0.01) and indirect associations. CONCLUSION: The RMBPC might be more suitable than the NPI in studies measuring BPSD to explain key caregiver outcomes such as depressive symptoms.

Functional and self-rated health mediate the association between physical indicators of diabetes and depressive symptoms.

BACKGROUND: Depression is common among persons with diabetes and associated with adverse health outcomes. To date, little is known about the causal mechanisms that lead to depression in diabetes. The aim of the present study was to examine to which extent functional and self-rated health mediate the association between physical health and depressive symptoms in diabetes. METHODS: Data of n = 3222 individuals with type 2 diabetes were analyzed cross-sectionally and longitudinally at three measurement occasions using path analysis. Indicators of physical health were glycemic control, number of comorbid somatic diseases, BMI, and insulin dependence. Furthermore, functional health, self-rated health and depressive symptoms were assessed. RESULTS: The effects of physical health on depressive symptoms were largely mediated by functional health and self-rated health. There was only a weak indirect effect of physical health on depressive symptoms. In contrast, self-rated health was a strong direct predictor of depressive symptoms. Self-rated health in turn depended strongly on patients' functional health. CONCLUSIONS: The way individuals perceive their health appears to have a stronger effect on their depressive symptoms than objective physical indicators of diabetes. Therefore practitioners should be trained to pay more attention to their patients' subjective health perceptions.

Sex hormones and mental rotation: an intensive longitudinal investigation.

The present study used an intensive longitudinal design to examine whether mental rotation performance varies according to a monthly cycle in both males and females and whether these variations are related to variations in progesterone, estradiol, and testosterone levels. We collected reaction time and accuracy data for 10 males and seven females each workday over eight weeks using 136 pairs of mental rotation stimuli/day, and measured sexual hormones concentrations in the saliva twice a week. A mixed linear model statistical analysis revealed that all females and seven males showed significant cycle effects in mental rotation performance. The female cycle showed an amplitude that was twice as large compared with the amplitude found in males. For males and females, estradiol and testosterone were significantly linearly and quadratically related to interindividual variation in performance at the beginning of the study (progesterone was linearly related to performance for females). The association between testosterone and performance differed across sexes: for males, it had an inverse U-shape, for females it was U-shaped. Towards the end of the study, none of the hormones were significantly related to performance anymore. Thus, the relationship between hormones and mental rotation performance disappeared with repeated testing. Only estradiol levels were significantly elevated at the lowest point of the cycle in mental rotation performance in females. In conclusion, in this intensive longitudinal study spanning two months, a monthly cycle in mental rotation performance was found among both males and females, with a larger cycle's amplitude for females.

On the Meaning of the "P Factor" in Symmetrical Bifactor Models of Psychopathology: Recommendations for Future Research From the Bifactor-(S-1) Perspective.

Symmetrical bifactor models are frequently applied to diverse symptoms of psychopathology to identify a general P factor. This factor is assumed to mark shared liability across all psychopathology dimensions and mental disorders. Despite their popularity, however, symmetrical bifactor models of P often yield anomalous results, including but not limited to nonsignificant or negative specific factor variances and nonsignificant or negative factor loadings. To date, these anomalies have often been treated as nuisances to be explained away. In this article, we demonstrate why these anomalies alter the substantive meaning of P such that it (a) does not reflect general liability to psychopathology and (b) differs in meaning across studies. We then describe an alternative modeling framework, the bifactor-(S-1) approach. This method avoids anomalous results, provides a framework for explaining unexpected findings in published symmetrical bifactor studies, and yields a well-defined general factor that can be compared across studies when researchers hypothesize what construct they consider "transdiagnostically meaningful" and measure it directly. We present an empirical example to illustrate these points and provide concrete recommendations to help researchers decide for or against specific variants of bifactor structure.

Telephone counseling for patients with chronic heart failure: results of an evaluation study.

BACKGROUND: The effectiveness of a secondary prevention program for patients suffering from chronic heart failure (CHF) was evaluated. PURPOSE: The program aimed at improving participants' perceived health and actual physical symptoms. Insurants of a German health insurance company participated in a telephone counseling program with four modules focusing on dietary habits, physical activity, fluid intake, and medication compliance. METHOD: Multilevel analyses were conducted to analyze changes in health related outcome variables over time in N = 259 participants who completed the program in about 6 months. RESULTS: The results showed an improvement of perceived health status, physical symptoms, and somatic impairment. Furthermore, differential change was found when comparing "finishers" compared to "non-finishers" of specific modules indicating specific module effects. CONCLUSION: The results are auspicious and, if sustained, are expected to bring about long-term health benefits for our study's participants. The program proved to be applicable and well accepted in the sample of older, severely impaired CHF patients and effective in changing perceived health.

Consistency and source specificity of symptom ratings for Child and Adolescent Behavior Inventory Scales with mother, father, and teacher ratings of Spanish children.

The study determined consistency relative to a reference source (convergent validity for nonreference sources with a reference source) and specificity for nonreference sources in true score variance in Child and Adolescent Behavior Inventory (CABI) ADHD-inattentive (IN), ADHD-hyperactive/impulsive (HI), sluggish cognitive tempo (SCT), oppositional defiant disorder (ODD), limited prosocial emotions (LPE), depression, anxiety, academic impairment, and social impairment scale scores. A multiple indicator by single trait by correlated (methods-1) model with indicator-specific trait factors was used to evaluate CABI scale scores with mother, father, and teacher ratings of 2,142 Spanish children (49.49% girls; ages 8-13). Mother ratings served as the reference source with father and teacher ratings serving as nonreference sources. Father with mother ratings showed substantial convergent correlations for each CABI scale (.73 to .90). Teacher with mother ratings showed a substantial convergent correlation for academic impairment (.72), strong convergent correlations for ADHD-IN (.57), ADHD-HI (.46), SCT (.52), and depression (.46), and small-to-moderate convergent correlations for ODD (.33), LPE (.24), and anxiety (.22) scales. Source independent and context-specific measures with appropriate content validity for home and school settings are required to better understand the greater variability in consistency from teacher to mother ratings. These results provide new information for the use of the CABI with multiple sources. In addition, the study indicates how a multiple indicator by single trait by correlated (methods-1) model with indicator-specific trait factors can be used to better understand multiple source ratings of child and adolescent psychopathology and adjustment. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

The Strengths and Difficulties Questionnaire: An examination of factorial, convergent, and discriminant validity using multitrait-multirater data.

This study examined the factorial, convergent, and discriminant validity of scores on the Strengths and Difficulties Questionnaire (SDQ), a tool for screening children's psychological adjustment. Data were collected from a community sample of 346 children (46% girls, M age = 54.07 months), using teacher, mother, and father reports. Results from confirmatory factor analyses indicated that the SDQ's hypothesized 5-factor structure fit the data well and partial strict measurement invariance was established across raters. Using teachers' reports as reference method, a correlated trait-correlated method minus 1 model (Eid et al., 2008) was fitted to investigate convergent and discriminant validity. The convergent validity of parents' ratings relative to teachers' ratings was modest. Mothers and fathers had a unique perspective on children's behavior above and beyond their partial overlap with teacher reports. Results indicated good discriminant validity between most of the traits measured by the SDQ. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Correction to: Promises and Pitfalls of Latent Variable Approaches to Understanding Psychopathology: Reply to Burke and Johnston, Eid, Junghänel and Colleagues, and Willoughby.

The original version of this article unfortunately contained a mistake. The supplementary material was not captured to our commentary article.

Application of the Bifactor S - 1 Model to Multisource Ratings of ADHD/ODD Symptoms: an Appropriate Bifactor Model for Symptom Ratings.

The symmetrical bifactor model is often applied to attention-deficit/hyperactivity disorder (ADHD)-hyperactive/impulsive (HI), ADHD-inattentive (IN), and oppositional defiant disorder (ODD) symptoms, but this model frequently yields anomalous or inadmissible results. An alternative model, the bifactor S - 1 model, is more appropriate for examining the hierarchical structure of ADHD/ODD symptoms. Both models were applied to ADHD-HI, ADHD-IN, and ODD symptom ratings by mothers, fathers, and teachers for 2142 Spanish children (49.49% girls; ages 8-13 years). The symmetrical bifactor model yielded the typical anomalous loadings, with a weakly defined ADHD-HI specific factor and difficult to interpret associations of general and specific factors with correlates. In contrast, the bifactor S - 1 model with ADHD-HI symptoms as general reference factor produced clearly interpretable results. For mothers and fathers, slightly more than 50% of true score variance in ADHD-IN and ODD symptoms represented specific residual variance not shared with the general ADHD-HI reference factor. For teachers, approximately 69% and 39% of true score variance in ADHD-IN and ODD symptoms, respectively, represented specific residual variance not shared with the general ADHD-HI reference factor. The general ADHD-HI reference factor and specific ADHD-IN and ODD residual factors showed convergent and discriminant validity across sources, along with unique associations with peer rejection, social impairment, and academic impairment factors. The bifactor S - 1 model also yielded results consistent with predictions from trait-impulsivity theory of ADHD/ODD development. Researchers should use the bifactor S - 1 model rather than the symmetrical bifactor model if hypotheses involve the latent hierarchical structure of ADHD/ODD symptoms.

Promises and Pitfalls of Latent Variable Approaches to Understanding Psychopathology: Reply to Burke and Johnston, Eid, Junghänel and Colleagues, and Willoughby.

The commentaries by Burke and Johnston (this issue), Eid (this issue), Junghänel et al. (this issue), and Willoughby (this issue) on Burns et al. (this issue) provide useful context for comparing three latent variable modeling approaches to understanding psychopathology-the correlated first-order syndrome-specific factors model, the bifactor S - 1 model, and the symmetrical bifactor model. The correlated first-order syndrome-specific factors model has proven useful in constructing explanatory models of psychopathology. The bifactor S - 1 model is also useful for examining the latent structure of psychopathology, especially in contexts with clear theoretical predictions. Joint use of correlated first-order syndrome-specific model and bifactor S - 1 model provides leverage for explaining psychopathology, and both models can also guide individual clinical assessment. In this reply, we further clarify reasons why the symmetrical bifactor model should not be used to study the latent structure of psychopathology and also discuss a restricted bifactor S - 1 model that is equivalent to the first-order syndrome-specific factors model.