Evaluating model fit for growth curve models: Integration of fit indices from SEM and MLM frameworks.

Wu, Wei; West, Stephen G.; Taylor, Alasdair

doi:10.1037/a0015858

Cited by 204 publications

(167 citation statements)

References 70 publications

Supporting

Mentioning

164

Contrasting

Unclassified

Order By: Relevance

“…For all models, omnibus Wald's tests indicated that the predictors had an overall effect. For final unconditional and conditional models, fit was assessed based on several indices, and all had acceptable fit such that SRMR < .05; CFI > .95; and RMSEA < .07 (Wu, West, & Taylor, 2009).…”

Section: Plan Of Analysesmentioning

confidence: 99%

Mapping Developmental Precursors of Cyber-Aggression: Trajectories of Risk Predict Perpetration and Victimization

Modecki

Barber

Vernon

2012

J Youth Adolescence

View full text Add to dashboard Cite

Technologically mediated contexts are social arenas in which adolescents can be both perpetrators and victims of aggression. Yet, there remains little understanding of the developmental etiology of cyber aggression, itself, as experienced by either perpetrators or victims. The current study examines three-year latent within-person trajectories of known correlates of cyber-aggression, problem behavior, (low) self-esteem, and depressed mood, in a large and diverse sample of youth (N = 1364; 54.6% female; 12-14 years old at T1 ). Findings demonstrate that developmental increases in problem behavior across grades 8-10 predict both cyber-perpetration and victimization in grade 11. Developmental decreases in self-esteem also predicted both grade 11 perpetration and victimization. Finally, early depressed mood predicted both perpetration and victimization later on, regardless of developmental change in depressed mood in the interim. Our results reveal a clear link between risky developmental trajectories across the early high school years and later cyber-aggression and imply that mitigating trajectories of risk early on may lead to decreases in cyber-aggression at a later date.

show abstract

Section: Plan Of Analysesmentioning

confidence: 99%

Mapping Developmental Precursors of Cyber-Aggression: Trajectories of Risk Predict Perpetration and Victimization

Modecki

Barber

Vernon

2012

J Youth Adolescence

View full text Add to dashboard Cite

show abstract

“…For example, it is more straightforward in a multilevel model to accommodate individual variation in the timing of measurements at a given occasion and to allow for further levels of clustering. (Further discussion of the relative strengths of MLM and SEM approaches to growth curve modelling can be found in Ghisletta & Lindenberg (2004), MacCallum et al (1997), Steele (2008) and Wu, West, & Taylor (2009). )…”

Section: Bivariate Growth Curve Modelsmentioning

confidence: 99%

A multilevel simultaneous equations model for within-cluster dynamic effects, with an application to reciprocal parent–child and sibling effects.

Steele¹,

Rasbash²,

Jenkins³

2013

Psychological Methods

View full text Add to dashboard Cite

LSE has developed LSE Research Online so that users may access research output of the School. Copyright © and Moral Rights for the papers on this site are retained by the individual authors and/or other copyright owners. Users may download and/or print one copy of any article(s) in LSE Research Online to facilitate their private study or for non-commercial research. You may not engage in further distribution of the material or use it for any profit-making activities or any commercial gain. You may freely distribute the URL (http://eprints.lse.ac.uk) of the LSE Research Online website.This document is the author's final accepted version of the journal article. There may be differences between this version and the published version. You are advised to consult the publisher's version if you wish to cite from it. given the phenomena under study, reciprocal influences are also investigated. Thus we can examine the extent to which a mother influences her child, as well as the way the child influences the mother's behavior over time. WITHIN-CLUSTER DYNAMIC EFFECTSOne of the challenges in the interpretation of correlational data is the isolation of influences that help us draw conclusions about causal mechanisms. For instance, the correlation in behavior between parents and children may be due to genes, environmental experiences, or a mixture of the two (Kaffman & Meaney, 2007;Plomin & Davis, 2009 There are several goals to this article. First, we briefly outline the methods that have been used previously for examining reciprocal processes. Such methods are based on growth curve and autoregressive cross-lagged models. We discuss the limitations of these methods for conclusions about causal mechanisms. Second, we introduce an analytic framework that allows for a more comprehensive and flexible examination of reciprocal influences than has been presented to date. The model allows for the inclusion of social groups of different sizes (including groups with only one dyad). We differentiate between reciprocal influences between individuals in a power hierarchy (parents and their multiple children), as well as individuals at the same level of power (e.g. siblings). We allow effects to depend on characteristics of individuals with different roles (e.g. parent and child). Effects of unmeasured factors at the occasion, individual and group levels are included. Third, we illustrate our approach with an analysis of the reciprocal effects of maternal depression and children's behavioral problems, using a dataset that includes up to three children per family.Our approach can be implemented in existing multilevel modelling software. WITHIN-CLUSTER DYNAMIC EFFECTS 5 A Review of Methods for Estimation of Reciprocal Parent-Child and Sibling EffectsPrevious studies into how the behavior of one family member affects another have been restricted to dyadic (or pairwise) relationships involving a parent-child (e.g. Elgar, Curtis, ). These studies have involved separate analyses for parent-child and sibling relationships using growth curve m...

show abstract

“…Many steps are needed to properly structure the data, and the SEM code quickly becomes unwieldy. In contrast, the HLM approach allows for simpler model specification, is computationally more efficient, and can easily be expanded to higher level growth models for manifest variables (Curran, 2003;Wu, West, & Taylor, 2009). A detailed comparison between HLM and SEM can be seen in Bauer (2003) and Curran (2003).…”

mentioning

confidence: 99%

“…In contrast, there is some agreement on the cut-off criteria of conventional fit indices based on the likelihood ratio test in SEM, such as RMSEA, CFI, and NNFI (TLI) (e.g., Hu & Bentler, 1999). However, since in SEM-based LGM the factor loadings are usually fixed at time points rather than freely estimated, and the fit of the model to the mean structure should be reflected as well, assessment of model fit by using conventional SEM-based fit indices should be cautious (Mehta & Neale, 2005;Wu et al, 2009). When every individual is observed at the same fixed set of time points (called "balanced") with no missing values (called "complete"), ML estimation is used; otherwise, FIML estimation is used (Wu et al, 2009).…”

mentioning

confidence: 99%

“…However, since in SEM-based LGM the factor loadings are usually fixed at time points rather than freely estimated, and the fit of the model to the mean structure should be reflected as well, assessment of model fit by using conventional SEM-based fit indices should be cautious (Mehta & Neale, 2005;Wu et al, 2009). When every individual is observed at the same fixed set of time points (called "balanced") with no missing values (called "complete"), ML estimation is used; otherwise, FIML estimation is used (Wu et al, 2009). With FIML estimation, the model-implied means and covariances are computed for each individual, and the maximum likelihood chi-square fit function is obtained by summing −2 log likelihood across all of the individual data vectors (Bovaird, 2007).…”

mentioning

confidence: 99%

See 1 more Smart Citation

Using SAS PROC CALIS to fit Level-1 error covariance structures of latent growth models

Ding

Jane

2012

Behav Res

View full text Add to dashboard Cite

In the present article, we demonstrates the use of SAS PROC CALIS to fit various types of Level-1 error covariance structures of latent growth models (LGM). Advantages of the SEM approach, on which PROC CALIS is based, include the capabilities of modeling the change over time for latent constructs, measured by multiple indicators; embedding LGM into a larger latent variable model; incorporating measurement models for latent predictors; and better assessing model fit and the flexibility in specifying error covariance structures. The strength of PROC CALIS is always accompanied with technical coding work, which needs to be specifically addressed. We provide a tutorial on the SAS syntax for modeling the growth of a manifest variable and the growth of a latent construct, focusing the documentation on the specification of Level-1 error covariance structures. Illustrations are conducted with the data generated from two given latent growth models. The coding provided is helpful when the growth model has been well determined and the Level-1 error covariance structure is to be identified.Keywords Error covariance structure . Latent growth model . Structural equation modeling The latent growth model (LGM) plays an important role in repeated measure analysis over a limited number of occasions in large samples (e.g., Meredith & Tisak, 1990;Muthén & Khoo, 1998; Preacher, Wichman, MacCallum, & Briggs, 2008, p. 12). The model can not only characterize intraindividual (within subjects) change over time, but also examine interindividual (between subjects) difference by means of random growth coefficients, and is a typical application of hierarchical linear modeling (HLM). The within-subjects errors over time and the between-subjects errors are conventionally referred to as "Level-1" and "Level-2" errors, respectively. LGM can also be handled by using structural equation modeling (SEM) (e.g., Bauer, 2003;Bollen & Curran, 2006;Chan, 1998;Curran, 2003;Duncan, Duncan, & Hops, 1996; Mehta & Neal 2005;Meredith & Tisak, 1990; Willet & Sayer, 1994). SEM and HLM stem from different statistical theory, and each has developed its own terminology and standard ways of framing research questions. However, there exists much overlap between the two methodologies under some circumstances. Typically, when a two-level data structure arises from the repeated observations of a variable over time for a set of individuals (so that time is hierarchically nested within each individual), SEM is analytically equivalent

show abstract

Evaluating model fit for growth curve models: Integration of fit indices from SEM and MLM frameworks.

Cited by 204 publications

References 70 publications

Mapping Developmental Precursors of Cyber-Aggression: Trajectories of Risk Predict Perpetration and Victimization

Mapping Developmental Precursors of Cyber-Aggression: Trajectories of Risk Predict Perpetration and Victimization

A multilevel simultaneous equations model for within-cluster dynamic effects, with an application to reciprocal parent–child and sibling effects.

Using SAS PROC CALIS to fit Level-1 error covariance structures of latent growth models

Contact Info

Product

Resources

About