[Show abstract][Hide abstract] ABSTRACT: In this chapter, we provide a tutorial on how to test for moderated effects of home–school relationships using different statistical approaches. We first provide a brief review of current research on moderated effects of home–school relationships to illustrate the theoretical and practical value of statistical testing of moderated effects. Next, with a detailed multilevel data example, we demonstrate different statistical methods, including: moderation models with composite scores, latent factors, and the reliability-adjusted composite scores as well as the use of the latent moderated structural equations (LMS) to test the moderating effect of a child characteristic (i.e., ADHD symptoms) on home–school relationship. We have used both SPSS Mixed and Mplus to analyze the models and included the corresponding annotated syntax of both programs in this chapter. We also discuss some related issues on testing moderation effects such as centering variables and handling missing data to provide guidance to researchers.
[Show abstract][Hide abstract] ABSTRACT: In a sample of 527 academically at-risk youth, we investigated trajectories of friends' and parents' school involvement across ages 12–14 and the joint contributions of these trajectories to adolescents' age 15 school engagement and academic achievement. Girls reported higher levels of friends' and parents' school involvement than boys. Both parents' and friends' school involvement declined across ages 12–14. Combined latent growth models and structural equation models showed effects of the trajectories of friends' and parents' school involvement on adolescents' age 15 school engagement and academic achievement, over and above adolescents' prior performance. These effects were additive rather than interactive. Strategies for enhancing parent involvement in school and students' affiliation with peers who are positively engaged in school are discussed.
No preview · Article · Jan 2016 · Journal of Research on Adolescence
[Show abstract][Hide abstract] ABSTRACT: We address two challenges in data analysis of group research. First, data may be clustered (i.e., responses of individual group members are correlated). Second, some dependent variables may consist of integer counts of number of occurrences of an event. Familiar ANOVA and regression models provide nonoptimal analyses in both cases. Standard multilevel (mixed) models yield accurate inference for clustered normally distributed data. Generalized linear models (GLMs), specifically Poisson regression and related models, yield accurate inference for nonclustered count data. New generalized linear mixed models (GLMMs) integrate GLMs with multilevel models, addressing both challenges and yielding accurate inferences for grouped count outcomes. To provide the necessary background for understanding GLMMs, we first introduce GLMs, with detailed coverage in an example of Poisson regression. We then introduce multilevel models. Finally, we develop GLMMs and illustrate in an example their application to clustered count data. Group research may benefit from the flexibility provided by GLMMs.
No preview · Article · Apr 2015 · Group Processes & Intergroup Relations
[Show abstract][Hide abstract] ABSTRACT: This chapter describes several best practices for collecting, storing, analyzing, and writing up data generated from biobehavioral research. The goal is to provide a foundation for researchers who wish to consider more advanced topics of experimental design and statistical inference in their work. The principles considered here are general and apply to basic biobehavioral research using a variety of different model systems and animals. Common issues and pitfalls are discussed, and considerations draw on both traditional principles and newer ideas.
[Show abstract][Hide abstract] ABSTRACT: Mediation analysis, or more generally models with direct and indirect effects, are commonly used in the behavioral sciences. As we show in our illustrative example, traditional methods of mediation analysis that omit confounding variables can lead to systematically biased direct and indirect effects, even in the context of a randomized experiment. Therefore, several definitions of causal effects in mediation models have been presented in the literature (Baron & Kenny, 19861.
Baron, R.M., & Kenny, D.A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. doi:10.1037/0022-3522.214.171.1243[CrossRef], [PubMed], [Web of Science ®], [CSA]View all references; Imai, Keele, & Tingley, 201013.
Imai, K., Keele, L., & Tingley, D. (2010). A general approach to causal mediation analysis. Psychological Methods, 15, 309–334. doi:10.1037/a0020761[CrossRef], [PubMed], [Web of Science ®]View all references; Pearl, 201232.
Pearl, J. (2012). The causal mediation formula: A guide to the assessment of pathways and mechanisms. Prevention Science, 13, 426–436. doi:10.1007/s11121-011-0270-1[CrossRef], [PubMed], [Web of Science ®]View all references). We illustrate the stochastic theory of causal effects as an alternative foundation of causal mediation analysis based on probability theory. In this theory we define total, direct, and indirect effects and show how they can be identified in the context of our illustrative example. A particular strength of the stochastic theory of causal effects are the causality conditions that imply causal unbiasedness of effect estimates. The causality conditions have empirically testable implications and can be used for covariate selection. In the discussion, we highlight some similarities and differences of the stochastic theory of causal effects with other theories of causal effects.
Preview · Article · Sep 2014 · Multivariate Behavioral Research
[Show abstract][Hide abstract] ABSTRACT: Aging is associated with progressive changes in learning and memory. A potential approach to attenuate age-related cognitive decline is cognitive training. In this study, adult male and female rats were given either repeated exposure to a T-maze, or no exposure to any maze, and then tested on a final battery of cognitive tasks. Two groups of each sex were tested from 6-18 months old on the same T-maze; one group received a version testing spatial reference memory, and the other group received only the procedural testing components with minimal cognitive demand. Groups three and four of each sex had no maze exposure until the final battery, and were comprised of aged or young rats. The final maze battery included the practiced T-maze and two novel tasks, one with a similar, and one with a different, memory type to the practice task. The fifth group of each sex was not maze tested, serving as an aged control for the effects of maze testing on neurotrophin protein levels in cognitive brain regions. Results showed that adult intermittent cognitive training enhanced performance on the practice task when aged in both sexes, that cognitive training benefits transferred to novel tasks and cognitive domains only in females, and that cognitive demand was necessary for these effects since rats receiving only the procedural testing components showed no improvement on the final maze battery. Further, for both sexes, rats that showed faster learning when young demonstrated better memory when aged. Age-related increases in neurotrophin concentrations in several brain regions were revealed, which was related to performance on the training task only in females. This longitudinal study supports the tenet that cognitive training can help one remember later in life, with broader enhancements and associations with neurotrophins only in females.
Full-text · Article · Jun 2014 · Neurobiology of Aging
[Show abstract][Hide abstract] ABSTRACT: Treatment noncompliance in randomized experiments threatens the validity of causal inference and the interpretability of treatment effects. This article provides a nontechnical review of 7 approaches: 3 traditional and 4 newer statistical analysis strategies. Traditional approaches include (a) intention-to-treat analysis (which estimates the effects of treatment assignment irrespective of treatment received), (b) as-treated analysis (which reassigns participants to groups reflecting the treatment they actually received), and (c) per-protocol analysis (which drops participants who did not comply with their assigned treatment). Newer approaches include (d) the complier average causal effect (which estimates the effect of treatment on the subpopulation of those who would comply with their assigned treatment), (e) dose-response estimation (which uses degree of compliance to stratify participants, producing an estimate of a dose-response relationship), (f) propensity score analysis (which uses covariates to estimate the probability that individual participants will comply, enabling estimates of treatment effects at different propensities), and (g) treatment effect bounding (which calculates a range of possible treatment effects applicable to both compliers and noncompliers). The discussion considers the areas of application, the quantity estimated, the underlying assumptions, and the strengths and weaknesses of each approach. (PsycINFO Database Record (c) 2014 APA, all rights reserved).
[Show abstract][Hide abstract] ABSTRACT: A propensity score is the probability that a participant is assigned to the treatment group based on a set of baseline covariates. Propensity scores provide an excellent basis for equating treatment groups on a large set of covariates when randomization is not possible. This article provides a nontechnical introduction to propensity scores for clinical researchers. If all important covariates are measured, then methods that equate on propensity scores can achieve balance on a large set of covariates that mimics that achieved by a randomized experiment. We present an illustration of the steps in the construction and checking of propensity scores in a study of the effectiveness of a health coach versus treatment as usual on the well-being of seriously ill individuals. We then consider alternative methods of equating groups on propensity scores and estimating treatment effects including matching, stratification, weighting, and analysis of covariance. We illustrate a sensitivity analysis that can probe for the potential effects of omitted covariates on the estimate of the causal effect. Finally, we briefly consider several practical and theoretical issues in the use of propensity scores in applied settings. Propensity score methods have advantages over alternative approaches to equating groups particularly when the treatment and control groups do not fully overlap, and there are nonlinear relationships between covariates and the outcome. (PsycINFO Database Record (c) 2014 APA, all rights reserved).
No preview · Article · Apr 2014 · Journal of Consulting and Clinical Psychology
[Show abstract][Hide abstract] ABSTRACT: The Adolescent Motivation for Educational Attainment Questionnaire is a 32-item questionnaire (we drew 20 items from 3 subscales of the Educational Motivation Questionnaire; Murdock, 1999) that was developed to measure multiple potential dimensions of adolescents' motivation to complete high school and enroll in post-secondary education, including competence and effort beliefs; perceived value of education; and peer, teacher, and parent support for educational attainment. We assessed a multiethnic sample (N = 569) of low-achieving students who started 1st grade together in 1 urban and 2 small city school districts. Participants were assessed over 2 consecutive years (Grades 8 and 9 given prior grade retention, or Grades 9 and 10 if not retained). Exploratory factor analyses identified 4 correlated dimensions underlying the questionnaire responses. Subsequent confirmatory factor analyses provided support for a bifactor model, which includes a general factor of students' basic educational motivation, and specific factors of (a) teacher educational expectations, (b) peer aspirations, and (c) value of education. Measurement invariance of the bifactor model was established across students' gender and ethnicity (Caucasian, African American, and Hispanic) and year of testing. Criterion-related validity of the general and specific factors with students' school belonging, student-teacher warmth and conflict, disciplinary practices, letter grade, conduct problems, and behavioral engagement was examined. Practical implications of the measure are discussed. (PsycINFO Database Record (c) 2014 APA, all rights reserved).
Full-text · Article · Mar 2014 · Psychological Assessment
[Show abstract][Hide abstract] ABSTRACT: This study demonstrates how a variant of growth curve modeling known as longitudinal parallel-process modeling can yield an understanding of the effect of symptoms on quality of life (QOL). A two-level hierarchical linear model with random intercepts and slopes was implemented within a structural equation modeling approach. The data (N = 367) comes from a large database of persons with HIV-associated illness. Twenty-three symptoms based on the Sign and Symptom Checklist for Persons with HIV disease and items measuring QOL from the general health status scales were used. Each respondent completed from 1 to 11 questionnaires. The number of reported symptoms had a significant association with patient QOL over time. These findings suggest that appropriate symptom management has the potential to improve patient QOL. This study demonstrates how a state-of-the-art longitudinal modeling technique evaluates the relationship between concurrent rates of change in measurements of two relevant variables.
No preview · Article · Nov 2013 · Western Journal of Nursing Research
[Show abstract][Hide abstract] ABSTRACT: In this article, the Society for Personality and Social Psychology (SPSP) Task Force on Publication and Research Practices offers a brief statistical primer and recommendations for improving the dependability of research. Recommendations for research practice include (a) describing and addressing the choice of N (sample size) and consequent issues of statistical power, (b) reporting effect sizes and 95% confidence intervals (CIs), (c) avoiding "questionable research practices" that can inflate the probability of Type I error, (d) making available research materials necessary to replicate reported results, (e) adhering to SPSP's data sharing policy, (f) encouraging publication of high-quality replication studies, and (g) maintaining flexibility and openness to alternative standards and methods. Recommendations for educational practice include (a) encouraging a culture of "getting it right," (b) teaching and encouraging transparency of data reporting, (c) improving methodological instruction, and (d) modeling sound science and supporting junior researchers who seek to "get it right."
No preview · Article · Nov 2013 · Personality and Social Psychology Review
[Show abstract][Hide abstract] ABSTRACT: Maternal sensitivity is a fundamental parenting construct and a determinant of positive child outcomes and healthy parent-child relationships. Few longitudinal studies have investigated determinants of sensitive parenting, particularly in a population of children at risk for developmental delay.
This study modeled trajectories of maternal sensitivity observed in two independent parenting contexts at child ages 3-, 4-, and 5-years. The sample included N = 247 mother-child dyads, with n = 110 children classified as at risk for developmental delays. Predictors included maternal distress, child anger proneness, and developmental risk status.
Maternal sensitivity changed during more demanding parenting tasks over the 3-year period but not during a low-demand task. Mothers of children with developmental risk, relative to mothers of typically developing children, and mothers of boys relative to mothers of girls, showed less sensitivity during more demanding parenting tasks.
Early developmental risk and child gender contribute to the nature of maternal sensitivity over time, but their contributions depend on the situational demands of the interaction. This contextualized view of sensitivity provides further evidence in support of parenting as a dynamic developmental process.
[Show abstract][Hide abstract] ABSTRACT: This study examined the performance of 4 correlation-based fit indexes (marginal and conditional pseudo R 2s; average and conditional concordance correlations) in detecting misspecification in mean structures in growth curve models. Their performance was also compared to that of 4 traditional SEM fit indexes. We found that the marginal pseudo R 2 and average concordance correlation were able to detect misspecification in the marginal mean structure (average change trajectory). The conditional pseudo R 2 and concordance correlation could detect misspecification when it occurred in the conditional mean structure (individual change trajectory) or in both mean structures. Compared to the SEM fit indexes, the correlation-based fit indexes were more robust to sample size but were less robust to data properties such as magnitude of population mean and measurement error. Theoretical and practical implications of the results and directions for future research are discussed.
[Show abstract][Hide abstract] ABSTRACT: How many days per week do you exercise for 30 minutes or more? Never? Once or twice? Every other day? Most days? Every day? Coarsely grouped counts such as this are commonly used in the behavioral sciences. When these variables are used as outcome variables, they often violate the assumptions of both linear regression and models designed for categorical outcomes; there is no model designed specifically for grouped count outcomes. Many analysis approaches also ignore the unequal spacing between categories; using the mean of the count range for each category captures the unequal spacing common in grouped count outcomes. The purpose of this study was to compare the statistical performance of three common regression models (linear regression, Poisson regression, and ordinal logistic regression) that can be used when the outcome is a grouped count.
A simulation study was used to determine the power, type I error, and confidence interval (CI) coverage for these models. Mean structure, variance structure, effect size, predictor type, and sample size were included in the factorial design. Mean structure reflected either a linear or an exponential relationship between the predictor and the outcome. Since the distribution of the underlying count is unobserved, several variance options were evaluated, including homoscedastic, monotonically increasing, and increasing then decreasing variance. Zero, small, medium, and large effect sizes and sample sizes of 100, 200, 500, and 1000 were examined. A single predictor (either continuous or binary) was used to predict the grouped count outcome.
All regression models produced unbiased estimates of the regression coefficient. Ordinal logistic regression produced type I error, power, and confidence interval (CI) coverage rates that were consistently within acceptable limits. Linear regression produced type I error and power that were within acceptable limits, but CI coverage was too low in conditions with an exponential mean structure, particularly with a large effect size and/or monotonically increasing variance structure. Poisson regression displayed inflated type I error, low power, and low CI coverage rates for nearly all conditions.
Based on the statistical performance of the three models, ordinal logistic regression is the preferred method for analyzing grouped count outcomes. Linear regression also performed well, but CI coverage was too low for several conditions with an exponential mean structure; these specific conditions are of particular interest because they reflect conditions commonly observed for counts and frequencies. Comparisons of model fit and tests of model assumptions (e.g., the proportional odds assumption for ordinal logistic regression) are in progress.
[Show abstract][Hide abstract] ABSTRACT: A Monte Carlo simulation was conducted to investigate the robustness of four latent variable interaction modeling approaches (Constrained Product Indicator [CPI], Generalized Appended Product Indicator [GAPI], Unconstrained Product Indicator [UPI], and Latent Moderated Structural Equations [LMS]) under high degrees of non-normality of the observed exogenous variables. Results showed that the CPI and LMS approaches yielded biased estimates of the interaction effect when the exogenous variables were highly non-normal. When the violation of non-normality was not severe (normal; symmetric with excess kurtosis < 1), the LMS approach yielded the most efficient estimates of the latent interaction effect with the highest statistical power. In highly non-normal conditions, the GAPI and UPI approaches with ML estimation yielded unbiased latent interaction effect estimates, with acceptable actual Type-I error rates for both the Wald and likelihood ratio tests of interaction effect at N ≥ 500. An empirical example illustrated the use of the four approaches in testing a latent variable interaction between academic self-efficacy and positive family role models in the prediction of academic performance.
No preview · Article · Nov 2012 · Multivariate Behavioral Research
[Show abstract][Hide abstract] ABSTRACT: This study investigated the effects of retention or promotion in first grade on growth trajectories in mathematics and reading achievement over the elementary school years (grades 1-5). From a large multiethnic sample (n = 784) of children who were below the median in literacy at school entrance, 363 children who were either promoted (n = 251) or retained (n = 112) in first grade could be successfully matched on 72 background variables. Achievement was measured annually using Woodcock-Johnson W scores; scores of retained children were shifted back one year to permit same-grade comparisons. Using longitudinal growth curve analysis, trajectories of math and reading scores for promoted and retained children were compared. Retained children received a one year boost in achievement; this boost fully dissipated by the end of elementary school. The pattern of subsequent retention in grades 2, 3 and 4 and placement in special education of the sample during the elementary school years is also described and their effects are explored. Policy implications for interventions for low achieving children are considered.
No preview · Article · Aug 2012 · Journal of Educational Psychology