Flexible models of change: using structural equations to match statistical and theoretical models of multiple change processes

Latent Change Score Models for the Study of Development and Dynamics in Organizational Research

The empirical study of change has proven to be one of the most vexing challenges in organizational science. Fortunately, contemporary methodologies originating from developmental psychology may provide a potential solution and are consequently working their way into the literature. In particular, organizational researchers are increasingly employing variations of latent change score (LCS) models to address questions regarding change, development, and dynamics. Although these models may indeed be used to reliably study change, development, and dynamics, many studies utilizing these models—and published in premier outlets—are characterized by questionable methodological choices, improper modeling procedures, and suboptimal research designs. Thus, the purpose of the present article is to (a) provide a critical review of LCS models, (b) outline appropriate modeling procedures (with corresponding Mplus and R syntax), (c) compare and contrast LCS modeling with other analytical techniques, and (d) delineate best practices. Ultimately, we endorse the use of LCS models by organizational researchers interested in studying longitudinal phenomena. However, we also heed researchers to do so judiciously because their misuse may lead to their unwarranted rejection by the field.

The Bidirectional Association Between Physical and Cognitive Function Among Chinese Older Adults: A Mediation Analysis

This study investigated the bidirectional association between physical and cognitive function in later life and examined the mechanisms underlying the interrelationship. We employed cross-lagged panel models to analyze a sample of 4232 unique participants aged 65 years and older from three waves of the Chinese Longitudinal Healthy Longevity Survey. Physical activity and social participation were tested as potential mediators between physical and cognitive function. Our findings revealed a reciprocal relationship between physical and cognitive function and a reciprocal relationship between physical and cognitive decline. Moreover, physical activity was confirmed to mediate the bidirectional association between physical and cognitive function, whereas social participation did not seem to be a mediator. A vicious cycle linking physical and cognitive decline may exist in Chinese older adults. However, leading a physically active lifestyle could be an effective intervention to slow physical and cognitive aging, thereby toning down the vicious cycle.

Lagged Coupled Changes Between White Matter Microstructure and Processing Speed in Healthy Aging: A Longitudinal Investigation

Age-related differences in white matter (WM) microstructure have been linked to lower performance in tasks of processing speed in healthy older individuals. However, only few studies have examined this link in a longitudinal setting. These investigations have been limited to the correlation of simultaneous changes in WM microstructure and processing speed. Still little is known about the nature of age-related changes in WM microstructure, i.e., regionally distinct vs. global changes. In the present study, we addressed these open questions by exploring whether previous changes in WM microstructure were related to subsequent changes in processing speed: (a) 1 year later; or (b) 2 years later. Furthermore, we investigated whether age-related changes in WM microstructure were regionally specific or global. We used data from four occasions (covering 4 years) of the Longitudinal Healthy Aging Brain (LHAB) database project (N = 232; age range at baseline = 64–86). As a measure of WM microstructure, we used mean fractional anisotropy (FA) in 10 major WM tracts averaged across hemispheres. Processing speed was measured with four cognitive tasks. Statistical analyses were conducted with bivariate latent change score (LCS) models. We found, for the first time, evidence for lagged couplings between preceding changes in FA and subsequent changes in processing speed 2 years, but not 1 year later in some of the WM tracts (anterior thalamic radiation, superior longitudinal fasciculus). Our results supported the notion that FA changes were different between regional WM tracts rather than globally shared, with some tracts showing mean declines in FA, and others remaining relatively stable across 4 years.

Self-reported head injury symptoms exacerbated in those with previous concussions following an acute bout of purposeful soccer heading

Rates of concussion in soccer are high, especially in female players. The primary aim of this study was to examine differences in self-reported concussion-related symptoms (CRS), balance (BESS), and neurocognitive performance (ImPACT) following an acute bout of soccer heading in a group of female collegiate players with and without a history of concussion. Eighty-seven players with 0 to 3+ previous concussions participated. The measurement variables were assessed before and after heading sessions; including one linear and one rotational bout. Players with concussion histories reported more CRS than their non-concussed teammates both before and after the heading sessions. Balance and neurocognitive scores were generally unaffected. This finding should heighten our awareness to carefully monitor soccer players who have experienced concussions and be aware that they may develop concussion-like symptoms, especially after acute bouts of heading either during practice or in matches. The long-term implications of this finding remain unknown.

Modeling Latent Change Score Analysis and Extensions in Mplus: A Practical Guide for Researchers

Many developmental and life course researchers are interested in modeling dynamic developmental processes. Latent change score (LCS) modeling is a potentially powerful modeling technique that can be used to assess complex life course processes, as well as the direction of longitudinal bivariate associations. Advances in modeling software, like Mplus, as well as widening adoption of software by researchers has made LCS modeling simpler. Thus, in the present paper, we provide 1) a theoretical overview of LCS analysis, 2) information on the interpretation of these models, 3) a practical guid7e for estimating these models in Mplus (including example syntax), 4) illustrative examples of LCS analysis, and 5) potential caveats for researchers.

Bivariate Developmental Relations between Calculations and Word Problems: A Latent Change Approach

The relation between 2 forms of mathematical cognition, calculations and word problems, was examined. Across grades 2-3, performance of 328 children (mean starting age 7.63 [SD=0.43]) was assessed 3 times. Comparison of a priori latent change score models indicated a dual change model, with consistently positive but slowing growth, described development in each domain better than a constant or proportional change model. The bivariate model including change models for both calculations and word problems indicated prior calculation performance and change were not predictors of subsequent word-problem change, and prior word-problem performance and change were not predictors of subsequent calculation change. Results were comparable for boys versus girls. The bivariate model, along with correlations among intercepts and slopes, suggest calculation and word-problem development are related, but through an external set of overlapping factors. Exploratory supplemental analyses corroborate findings and provide direction for future study.

Modeling bivariate change in individual differences: Prospective associations between personality and life satisfaction

A number of structural equation models have been developed to examine change in 1 variable or the longitudinal association between 2 variables. The most common of these are the latent growth model, the autoregressive cross-lagged model, the autoregressive latent trajectory model, and the latent change score model. The authors first overview each of these models through evaluating their different assumptions surrounding the nature of change and how these assumptions may result in different data interpretations. They then, to elucidate these issues in an empirical example, examine the longitudinal association between personality traits and life satisfaction. In a representative Dutch sample (N = 8,320), with participants providing data on both personality and life satisfaction measures every 2 years over an 8-year period, the authors reproduce findings from previous research. However, some of the structural equation models overviewed have not previously been applied to the personality-life satisfaction relation. The extended empirical examination suggests intraindividual changes in life satisfaction predict subsequent intraindividual changes in personality traits. The availability of data sets with 3 or more assessment waves allows the application of more advanced structural equation models such as the autoregressive latent trajectory or the extended latent change score model, which accounts for the complex dynamic nature of change processes and allows stronger inferences on the nature of the association between variables. However, the choice of model should be determined by theories of change processes in the variables being studied.

Longitudinal associations among change in overweight status, fear of negative evaluation, and weight-related teasing among obese adolescents

OBJECTIVE

To examine longitudinal bidirectional associations between changes in adolescents' weight status and psychosocial constructs.

METHOD

118 obese adolescents aged 13-16 years participated in a behavioral weight control intervention. Percent overweight (OW), fear of negative evaluation (FNE), and frequency of weight-related teasing (WRT) were collected at baseline, end of intervention, and 12 and 24 months post-randomization. 3 multivariate latent change score models were estimated to examine longitudinal cross-lagged associations between: (1) OW and FNE; (2) OW and WRT; and (3) FNE and WRT.

RESULTS

Decreases in OW were prospectively associated with subsequent decreases in both FNE and WRT; however, changes in FNE and WRT were not prospectively associated with subsequent change in OW. Decreases in FNE were prospectively associated with subsequent decreases in WRT.

CONCLUSION

Moderate weight loss in the context of a behavioral weight control intervention has positive long-term implications for obese adolescents' peer relations.