Multivariate Latent Growth Modeling Research Articles

Changes in situational and dispositional factors as predictors of job satisfaction

Arguably, job satisfaction is one of the most important variables with regard to work. When explaining job satisfaction, research usually focuses on predictor variables in terms of levels but neglects growth rates. Therefore it remains unclear how potential predictors evolve over time and how their development affects job satisfaction. Using multivariate latent growth modeling in a study with 1145 young workers over five years, we analyzed how well job satisfaction is predicted a) by levels of situational (i.e., job control) and dispositional (i.e., Core Self-Evaluations (CSE)) factors and b) by growth per year of these predictors. Results showed both intercepts and slopes to be related to each other, suggesting a joint growth of job control and CSE during early careers. Job satisfaction after five years was best predicted by the slopes of job control (β=.31, p<.001) and CSE (β=.34, p<.01). These findings provide further longitudinal evidence for the role of situational as well as dispositional factors for predicting job satisfaction. In addition, growth rates per year were better predictors than initial levels. Furthermore, a lack of change in job control or CSE went along with a drop in job satisfaction, implying that young workers need to perceive things to be improving in order to increase, or at least maintain, their level of job satisfaction. In terms of theory, the relative importance of levels versus changes deserves more attention. In terms of practical implications, our results suggest a double emphasis on job design (i.e., granting sufficient, and increasing, control) and on personal development (e.g., through training) so that people experience a match between both. Finally, negative associations between initial levels and growth rates suggest that people are quite successful in achieving a reasonable fit between their job characteristics and their needs and goals.

Multivariate Latent Growth Models for Mixed Data with Covariate Effects

The paper presents an extension of a new class of multivariate latent growth models (Bianconcini and Cagnone, 2012) to allow for covariate effects on manifest, latent variables and random effects. The new class of models combines: (i) multivariate latent curves that describe the temporal behavior of the responses, and (ii) a factor model that specifies the relationship between manifest and latent variables. Based on the Generalized Linear and Latent Variable Model framework (Bartholomew and Knott, 1999), the response variables are assumed to follow different distributions of the exponential family, with item-specific linear predictors depending on both latent variables and measurement errors. A full maximum likelihood method is used to estimate all the model parameters simultaneously. Data coming from the Data WareHouse of the University of Bologna are used to illustrate the methodology.

Math and science achievement in English language learners: Multivariate latent growth modeling of predictors, mediators, and moderators.

The effectiveness of various strategies for educating the growing U.S. population of English language learners (ELLs) has attracted a great deal of controversy. Bilingual education theory posits that retention and continued development of native language (L1) skills facilitate academic achievement through two mediating mechanisms. First, L1 proficiency promotes second language (L2) acquisition, which is required to succeed academically. Second, competence in the heritage language positively influences academic achievement by maintaining self-esteem. Both of these meditational chains were tested in a multivariate latent growth model of longitudinal data from the National Education Longitudinal Study (NELS:88/2000). L1 proficiency was specified to predict distal math and science outcomes through its association with both L2 literacy and self-esteem. In turn, academic self-concept was hypothesized to mediate the self-esteem--academic achievement relation. Home and school characteristics were included as covariates. The full model exhibited excellent fit only in the Hispanic portion of the ELL sample. Consistent with the broader literature, ethnicity appears to serve as an important moderator of the way native language proficiency relates to self-perception, L2 acquisition, and math and science achievement.

A General Multivariate Latent Growth Model With Applications to Student Achievement

The evaluation of the formative process in the University system has been assuming an ever increasing importance in the European countries. Within this context, the analysis of student performance and capabilities plays a fundamental role. In this work, the authors propose a multivariate latent growth model for studying the performances of a cohort of students of the University of Bologna. The model proposed is innovative since it is composed by (a) multivariate growth models that allow the capture of different dynamics of student performance indicators over time and (b) a factor model that allows measurement of the general latent student capability. The flexibility of the model proposed allows its applications in several fields such as socioeconomic settings in which personal behaviors are studied using panel data.

Predictors of Interpersonal Growth in Volunteer Tourism: A Latent Curve Approach

Volunteer travel has become a substantial market segment in the tourism industry. Case studies have documented the effects of volunteer travel on participants and host communities. The purpose of this quantitative study was to determine the impact of volunteer travel experiences on the openness, civic attitudes, and wisdom of college participants and to elucidate predictors of positive growth in these intended outcomes. A multivariate latent growth model was tested to determine the nature of growth trajectories. Results indicate that the program had positive immediate impacts on all dependent variables. Evidence of continued growth was found one month after the experience in a follow-up assessment. Participants involved in leadership roles and those who engaged in regular personal reflection demonstrated stronger long-term growth.

Transitioning to adolescence: How changes in child personality and overreactive parenting predict adolescent adjustment problems

The present study examined how changes in child Big Five personality characteristics and overreactive parenting during the transition from childhood to adolescence predict adolescent adjustment problems. The sample included 290 children, aged 8-9 years. At three moments, with 2-year intervals, mothers, fathers, and a teacher reported on the child's personality, and mothers and fathers reported on their parenting behavior. At the third measurement moment, mothers, fathers, and children reported on the child's adjustment problems. Rank-order stability of the personality dimensions and overreactive parenting were high. Univariate latent growth models revealed mean-level decreases for extraversion, conscientiousness, and imagination. Mean levels of benevolence, emotional stability, and overreactive parenting were stable. Multivariate latent growth models revealed that decreases in extraversion and emotional stability predicted internalizing problems, whereas decreases in benevolence, conscientiousness, and emotional stability predicted externalizing problems. Increases in overreactive parenting predicted externalizing, but not internalizing problems. The associations were similar for boys and girls. The results indicate that changes in child personality and overreactive parenting during the transition to adolescence are associated with adolescent adjustment problems. Overall, child personality was more important than overreactive parenting, and children were more likely to "act out" than to "withdraw" in reaction to overreactive parenting.

A Comparison of Latent Growth Models for Constructs Measured by Multiple Items

Univariate latent growth modeling (LGM) of composites of multiple items (e.g., item means or sums) has been frequently used to analyze the growth of latent constructs. This study evaluated whether LGM of composites yields unbiased parameter estimates, standard errors, chi-square statistics, and adequate fit indexes. Furthermore, LGM was compared with 2 alternatives: LGM with fixed error variances and the curve-of-factors model (McArdle, 1988), which is a multivariate latent growth model. It was found that the 2 univariate models only yield adequate results when the items are essentially tau-equivalent and there is strict factorial invariance. The curve-of-factors model was found to produce adequate results in all conditions, but it usually requires large sample sizes.

PLACING PEER RATINGS IN CONTEXT: SYSTEMATIC INFLUENCES BEYOND RATEE PERFORMANCE

Performance evaluation research indicates that variance in ratings may be attributable to systematic sources beyond the actual performance of the ratee. However, the majority of prior work compares ratings across sources and uses ratings from a single rating event. Using confirmatory factor analysis and multivariate latent growth modeling (MLGM), we specifically examine peer ratings from 740 participants on 5 performance dimensions across 3 distinct performance situations for systematic sources of variance beyond ratee performance. Results demonstrate that both ratee performance and the performance context have systematic effects, with contextual effects varying by how “strong” or “weak” the situation is for a given performance dimension. Furthermore, MLGM results suggest that the influence of performance dynamism is only meaningfully interpreted when contextual effects are modeled.

Change in cognitive capabilities in the oldest old: the effects of proximity to death in genetically related individuals over a 6-year period.

Change in cognitive abilities was assessed over a 6-year period in a sample of monozygotic and same-sex dizygotic twin pairs (N = 507 individuals), aged 80 and older (mean age = 83.3 years: SD = 3.1). who remained nondemented over the course of the study. Latent growth models (LGMs) show that chronological age and time to death are consistent predictors of decline in measures of memory, reasoning, speed, and verbal abilities. Multivariate LGM analysis resulted in weak and often negative correlations among rates of change between individuals within twin pairs, indicating greater differential change within twin pairs than occurs on average across twin pairs. These findings highlight several challenges for estimating genetic sources of variance in the context of compromised health and mortality-related change.

Qualitative and quantitative shifts in adolescent problem behavior development: A cohort-sequential multivariate latent growth modeling approach.

The aim of this study was to examine the nature of problem behavior development from late childhood through adolescence, to assess the quantitative development of problem behavior (alcohol use, marijuana use, deviance, academic failure) as well as potential qualitative shifts in problem behavior over time. Multivariate latent growth curve modeling (LGM) analyses and a cohort-sequential design were employed. Data were from the National Youth Survey and included 770 youth from four cohorts (11, 12, 13, 14 years old), assessed annually for 5 years. Results showed significant growth in problem behavior from ages 11 to 18. Alcohol use and marijuana use contributed most, and academic failure contributed least to the problem behavior latent construct. Results of the variant model revealed that the contribution of all four behaviors to the overall problem behavior construct increased similarly as children aged.

Modeling Intraindividual Changes in Children's Social Skills at Home and at School: A Multivariate Latent Growth Approach to Understanding Between-Settings Differences in Children's Social Skill Development

Multivariate latent growth modeling was used to conceptualize and analyze intraindividual changes in children's social skills and interindividual differences in these changes in home and school settings. Parent and teacher ratings assessing children's social skills at home and school settings, respectively, were obtained for a sample of 378 children at 4 time points spaced at approximately 12-month intervals over a 4-year period from Kindergarten to Grade 3. Results showed that, in initial status at Kindergarten, there were significant individual differences in social skills in both home and school settings and a significant positive association between initial status in social skills in the two settings. Systematic between-settings differences in children's social skill development were found. Social skills development at home was best described with a nonlinear trajectory in which skills increased from Kindergarten to Grade 2 with a substantially larger increase from Grade 1 to Grade 2 than from Kindergarten to Grade 1, and then remained relatively constant from Grade 2 to Grade 3. In contrast, social skills development at school was best described with a negative linear trajectory in which skills decreased at a constant rate from Kindergarten to Grade 3. The differences in social skills development may derive form the fact that different teachers with different expectations regarding social skills provided ratings each year while the same parent was the source of at-home social skills ratings. There were significant individual differences in growth rates in the school as well as the home setting. Evidence of between-settings differences in social skills development were obtained from differential patterns of associations between growth parameters (initial status and growth rate) and individual predictors (family income, parent education, child verbal skills) across settings.

Risk and protective factors influencing adolescent problem behavior: A multivariate latent growth curve analysis

This study examined the dynamic relations between adolescent problem behaviors (alcohol, marijuana, deviance, academic failure) over time and predictors of these behaviors. Data from the National Youth Survey (1) included 1,044 adolescents (53.5% male; mean age at year 1 = 13.20). Dependent measures were adolescent alcohol use, marijuana use, deviance, and academic failure, assessed annually over 4 years. Independent measures included age, gender, marital status, income, family time, family support, time with friends, friend deviance, knowledge of friends, activities, and neighborhood problems. An associative latent growth modeling (LGM) analysis showed significant increases and relations between the four behaviors in both initial status and development. Second-order multivariate LGM analyses indicated that the four behaviors could be modeled by a higher-order problem behavior construct. Significant effects on the common problem behavior intercept or slope included time with friends, deviant friends, age, marital status, family time, and support. Additional effects were found to be specific to the initial status and slopes of individual problem behaviors. Overall, results indicate the importance of assessing the relations between adolescent problem behaviors as they change over time and identifying the risk and protective factors that have both common and individual influences on these behaviors.

Contributions of the social context to the development of adolescent substance use: a multivariate latent growth modeling approach

This article demonstrates a latent growth curve methodology for analyzing longitudinal data of adolescent substance use. Hypotheses concerning the form of growth in alcohol, cigarette, and marijuana use, and covariates influencing the form of growth, were tested. Participants were male and female adolescents ( n=664) assessed at three time points. A common trajectory existed across the developmental period with significant increases in all three substances. Second-order multivariate extensions of the basic latent growth modeling framework suggested that associations among the individual differences parameters, representing growth or change in the various substance use behaviors, could be adequately modeled by a higher-order substance use construct. Inept parental monitoring, parent-child conflict, peer deviance, academic failure, gender, and age, were significant predictors of initial levels and the trajectory of substance use.