Laura Castro-Schilo

Distinguishing ordinal and disordinal interactions.

Re-parameterized regression models may enable tests of crucial theoretical predictions involving interactive effects of predictors that cannot be tested directly using standard approaches. First, we present a re-parameterized regression model for the Linear × Linear interaction of 2 quantitative predictors that yields point and interval estimates of 1 key parameter-the crossover point of predicted values-and leaves certain other parameters unchanged. We explain how resulting parameter estimates provide direct evidence for distinguishing ordinal from disordinal interactions. We generalize the re-parameterized model to Linear × Qualitative interactions, where the qualitative variable may have 2 or 3 categories, and then describe how to modify the re-parameterized model to test moderating effects. To illustrate our new approach, we fit alternate models to social skills data on 438 participants in the National Institute of Child Health and Human Development Study of Early Child Care. The re-parameterized regression model had point and interval estimates of the crossover point that fell near the mean on the continuous environment measure. The disordinal form of the interaction supported 1 theoretical model-differential-susceptibility-over a competing model that predicted an ordinal interaction.

Determinants of Mental Health and Self-Rated Health: A Model of Socioeconomic Status, Neighborhood Safety, and Physical Activity

OBJECTIVES We investigated the underlying mechanisms of the influence of socioeconomic status (SES) on mental health and self-rated health (SRH), and evaluated how these relationships might vary by race/ethnicity, age, and gender. METHODS We analyzed data of 44 921 adults who responded to the 2009 California Health Interview Survey. We used a path analysis to test effects of SES, neighborhood safety, and physical activity on mental health and SRH. RESULTS Low SES was associated with greater neighborhood safety concerns, which were negatively associated with physical activity, which was then negatively related to mental health and SRH. This model was similar across different racial/ethnic and gender groups, but mean levels in the constructs differed across groups. CONCLUSIONS SES plays an important role in SRH and mental health, and this effect is further nuanced by race/ethnicity and gender. Identifying the psychological (neighborhood safety) and behavioral (physical activity) factors that influence mental health and SRH is critical for tailoring interventions and designing programs that can improve overall health.

Gender differences in the relationship between emotional intelligence and right hemisphere lateralization for facial processing

The present study examined relationships between emotional intelligence, measured by the Mayer-Salovey-Caruso Emotional Intelligence Test, and right hemisphere dominance for a free vision chimeric face test. A sample of 122 ethnically diverse college students participated and completed online versions of the forenamed tests. A hierarchical regression was performed to test for the hypothesized interaction between gender and EI on the right hemisphere bias score. No significant main effects were found for gender or total EI score. However, when entered into the model, the interaction term contributed an additional 4.5% of the variance in right hemisphere dominance for the processing of facial emotions. Descriptively, men with greater EI were associated with higher right hemisphere dominance in the free vision test, while no association was observed for women.

Neglect the Structure of Multitrait-Multimethod Data at Your Peril: Implications for Associations With External Variables

In 1959, Campbell and Fiske introduced the use of multitrait–multimethod (MTMM) matrices in psychology, and for the past 4 decades confirmatory factor analysis (CFA) has commonly been used to analyze MTMM data. However, researchers do not always fit CFA models when MTMM data are available; when CFA modeling is used, multiple models are available that have attendant strengths and weaknesses. In this article, we used a Monte Carlo simulation to investigate the drawbacks of either using CFA models that fail to match the data-generating model or completely ignore the MTMM structure of data when the research goal is to uncover associations between trait constructs and external variables. We then used data from the National Institute of Child Health and Human Development Study of Early Child Care and Youth Development to illustrate the substantive implications of fitting models that partially or completely ignore MTMM data structures. Results from analyses of both simulated and empirical data show noticeable biases when the MTMM data structure is partially or completely neglected.

Preliminary Detection of Relations Among Dynamic Processes With Two-Occasion Data

Most novel analytic methods for longitudinal data are applicable to studies spanning three time-points of data at a minimum, whereas methods for two-occasion data have garnered comparatively little attention. Here, we address this limitation by introducing the two-wave latent change score (2W-LCS) model, a technique appropriate for preliminary detection of relations among dynamic processes with two-occasion data. The 2W-LCS model is well suited for the investigation of hypotheses in which changes in a construct are posited as predictors of changes in another construct. In an empirical illustration using data of elderly Hispanics from the Health and Retirement Study, we demonstrate how the 2W-LCS model provides the best match to theories rooted in changes, and highlight the advantages of this approach over other modeling alternatives (i.e., Little, Preacher, Selig, & Card, 2007; Selig & Preacher, 2009).

Bayesian Versus Maximum Likelihood Estimation of Multitrait–Multimethod Confirmatory Factor Models

This article compares maximum likelihood and Bayesian estimation of the correlated trait–correlated method (CT–CM) confirmatory factor model for multitrait–multimethod (MTMM) data. In particular, Bayesian estimation with minimally informative prior distributions—that is, prior distributions that prescribe equal probability across the known mathematical range of a parameter—are investigated as a source of information to aid convergence. Results from a simulation study indicate that Bayesian estimation with minimally informative priors produces admissible solutions more often maximum likelihood estimation (100.00% for Bayesian estimation, 49.82% for maximum likelihood). Extra convergence does not come at the cost of parameter accuracy; Bayesian parameter estimates showed comparable bias and better efficiency compared to maximum likelihood estimates. The results are echoed via 2 empirical examples. Hence, Bayesian estimation with minimally informative priors outperforms enables admissible solutions of the CT–CM model for MTMM data.

Using residualized change versus difference scores for longitudinal research

Researchers interested in studying change over time are often faced with an analytical conundrum: Whether a residualized change model versus a difference score model should be used to assess the effect of a key predictor on change that took place between two occasions. In this article, the authors pose a motivating example in which a researcher wants to investigate the effect of cohabitation on pre- to post-marriage change in relationship satisfaction. Key features of this example include the likely self-selection of dyads with lower relationship satisfaction to cohabit and the impossibility of using experimentation procedures to attain equivalent groups (i.e., cohabitants vs. not cohabitants). The authors use this example of a nonrandomized study to compare the residualized change and difference score models analytically and empirically. The authors describe the assumptions of the models to explain Lord’s paradox; that is, the fact that these models can lead to different inferences about the effect under investigation. They also provide recommendations for modeling data from nonrandomized studies using a latent change score framework.

Augmenting the Correlated Trait–Correlated Method Model for Multitrait–Multimethod Data

We introduce an approach for ensuring empirical identification of the correlated trait–correlated method (CT–CM) model under a variety of conditions. A set of models are referred to as augmented correlated trait–correlated method (ACT–CM) models because they are based on systematically augmenting the multitrait–multimethod matrix put forth by Campbell and Fiske (1959). We show results from a Monte Carlo simulation study in which data characteristics lead to an empirically underidentified standard CT–CM model, but a well-identified fully augmented correlated trait–correlated method (FACT–CM) model. This improved identification occurs even for a model in which equality constraints are imposed on loadings on each trait factor and loadings on each method factor—a specific case shown to lead to an empirically underidentified CT–CM model.

Developmental Outcomes of School Attachment Among Students of Mexican Origin

We used a longitudinal community study of 674 grade school children (Grades 5, 6, 7, and 8; 337 males, 337 females) of Mexican origin to examine outcomes of school attachment. Attachment to school is important in this population given the high level of school dropout rates of Mexican-origin students. Results indicated that, on average, school attachment from fifth to sixth grade remains stable, but declines from sixth to eighth grade. Boys had lower levels of school attachment at fifth grade but followed similar patterns of change as girls did. Attachment to teachers, peer competence, school aspirations and expectations, and substance-use cognitions emerged as longitudinal outcomes of level or changes in school attachment. Gender moderated associations of school attachment.

Toward Improving the Modeling of MTMM Data: A Response to Geiser, Koch, and Eid (2014)

Geiser, Koch, and Eid (2014) expressed their views on an article we published describing findings from a simulation study and an empirical study of multitrait–multimethod (MTMM) data. Geiser and colleagues raised concerns with (a) our use of the term bias, (b) our statement that the correlated trait–correlated method minus one [CT–C(M–1)] model is not in line with Campbell and Fiske’s (1959) conceptualization of MTMM data, (c) our selection of a data-generating model for our simulation study, and (d) our preference for the correlated trait–correlated method (CT–CM) model over the CT–C(M–1) model. Here, we respond to and elaborate on issues raised by Geiser et al. We maintain our position on each of these issues and point to the interpretational challenges of the CT–C(M–1) model. But, we clarify our opinion that none of the existing structural models for MTMM data are flawless; each has its strengths and each has its weaknesses. We further remind readers of the goal, findings, and implications of our recently published article.