Hierarchical Growth Curve Model Research Articles

Proposing a Bayesian hierarchical growth curve model (BHGCM) for tourism and hospitality research

Curvilinear growth trajectories of products/services are common in the tourism and hospitality industry. To fit nonlinear growth patterns with multilevel structures, this study proposed Bayesian hierarchical growth curve models (BHGCMs) in line with the increasing adoption of Bayesian analysis in tourism and hospitality academia. We provided the basic form of BHGCM along with its unique advantages. For an empirical test, this study applied several growth curves to approximate online reviews of U.S. hotels from August 2020 to January 2021. After selecting a Gompertz curve as a mean function of the GCM, a Bayesian hierarchical approach was employed to estimate growth parameters—namely, base and maximum volume of hotel reviews, inflection week, and relative growth rate—and identify their determinants. Our findings demonstrate the superiority of the proposed BHGCM in fitting the growth patterns of hotel reviews while revealing the effect of price and accumulated reviews on the parameters.

Life Chances, Subjective Perceptions, and Healthy Lifestyles in Older Adults: Longitudinal Evidence From China.

Previous studies and theories show that healthy lifestyles are affected by life chances and subjective perceptions. However, it remains unclear how older adults change between different lifestyle profiles as they age. We proposed the healthy lifestyle duality framework and tested it among older Chinese adults using a longitudinal design. Data were obtained from 4 Chinese Longitudinal Healthy Longevity Surveys conducted between 2008 and 2018 (N = 31,445). Latent transition analysis and hierarchical multinomial growth curve models were performed to investigate healthy lifestyle profiles and their associations with life chances and subjective perceptions over time. Four distinct healthy lifestyle profiles were identified: healthy, risky, low-standard, and mixed groups, and their changes show path dependency. Across 10 years, the proportion of the healthy group ranged from 11.16% to 16.97%. Both life chances and subjective perceptions were longitudinally associated with healthy lifestyles, with age and cohort effects observed. Our findings support the healthy lifestyle duality framework and reveal that life chances and subjective perceptions influence lifestyle changes over time. Public health policies and health intervention programs should adapt to the specific needs of different age groups and generations.

Health asymmetry as a predictor of depressive symptomatology over time among older European adults: A growth curve analysis

ObjectiveSubjective (SH) and objective health (OH) measures are associated with depressive symptomatology in older adults. We explored whether the discrepancy between SH and OH (operationalised as ‘health asymmetry’ with four categories: ‘health optimistic’, ‘health pessimistic’ and ‘good health realistic’ and ‘poor health realistic’) may also relate to depressive symptomatology 1) cross-sectionally, and 2) longitudinally, among older European adults. MethodsAdults (n = 26,520), aged 50+, from 11 European countries, were assessed over six waves of data collection (2006–2020) in the SHARE study. A hierarchical multi-level growth curve model explored whether health asymmetry was associated with depressive symptomology at baseline, and with depressive symptom trajectories across time, accounting for country of origin. ResultsAt baseline, 11.8% of older adults were classified as health pessimistic, with 15.5% being health optimistic, 42.9% being poor health realistic and 29.8% being good health realistic. A positive linear trend in depressive symptomatology was noted across 14 years of SHARE data (β = 0.11, p < .001). Health pessimists displayed higher levels of depressive symptoms than both health realistic groups and health optimists. However, health pessimists experienced a less steep increase in depressive symptoms across time (β = −0.10, p < .001), relative to good health realists. ConclusionHealth pessimists experience elevated levels of depressive symptoms, but show less growth in depressive symptomatology than expected. Further research is required to understand the underlying causes of the varying depressive symptom trajectories among these groups.

A Bayesian Multivariate Hierarchical Growth Curve Model to Examine Cumulative Socio-Economic (Dis)Advantage among Childless Adults and Parents

AbstractChildlessness and socio-economic well-being interact dynamically throughout the life course, possibly resulting in an accumulation of socio-economic (dis)advantage. Methods commonly used to investigate this hypothesis are unable to simultaneously acknowledge that childlessness entails a heterogeneous and processual ‘non-event’ which interrelates with multiple life domains. I use Bayesian multivariate hierarchical growth curve modelling to facilitate synchronous incorporation of these substantive complexities. I construct prospective interdependent life course trajectories of socio-economic well-being for eventual parents and permanently childless adults; distinguishing voluntary, involuntary, circumstantial and indecisive childlessness. Using 1970 British Cohort Study data, I find that parents and voluntarily childless adults are more satisfied with life than non-voluntarily childless adults. Voluntarily and circumstantially childless women earn most and mothers least, while fathers outearn childless men. (Dis)advantage in economic and male subjective well-being accumulates throughout the life course. Group differences in partnerships, employment, education and health, more than childbearing behaviours themselves, explain the (growth of) these socio-economic differentials. The findings have implications for policies addressing socio-economic inequalities by both parental status and gender. The proposed approach provides valuable for assessing cumulative (dis)advantage in interrelated life domains via an unclearly marked heterogeneous process, not only in the context of childlessness but also other life course topics.

Context Stability in Habit Building Increases Automaticity and Goal Attainment.

In this paper, we investigate the effects of context stability on automaticity and goal attainment in intentional habit building. We used hierarchical growth curve modeling and multilevel mediation to test our hypotheses on two datasets. In Study 1, N = 95 university students (N = 2,482 habit repetitions) built new study habits over a period of 6 weeks with manipulated context stability. One group was instructed to constantly vary the context of their habit repetitions by changing rooms and times and the other group was instructed to keep the context of habit performance stable. In Study 2, N = 308 habits (N = 2,368 habit repetitions) from N = 218 users of a published habit building app were analyzed without manipulating but measuring context stability. We found the same pattern in both datasets: Context stability predicted more automaticity and higher habit repetition goal attainment. We also found that the effect of context stability on habit repetition goal attainment was partially mediated by automaticity in both datasets. These results show that context does not only act as a trigger for habit instigation but also has an ongoing effect on habit execution.

A two-stage approach for the spatio-temporal analysis of high-throughput phenotyping data

High throughput phenotyping (HTP) platforms and devices are increasingly used for the characterization of growth and developmental processes for large sets of plant genotypes. Such HTP data require challenging statistical analyses in which longitudinal genetic signals need to be estimated against a background of spatio-temporal noise processes. We propose a two-stage approach for the analysis of such longitudinal HTP data. In a first stage, we correct for design features and spatial trends per time point. In a second stage, we focus on the longitudinal modelling of the spatially corrected data, thereby taking advantage of shared longitudinal features between genotypes and plants within genotypes. We propose a flexible hierarchical three-level P-spline growth curve model, with plants/plots nested in genotypes, and genotypes nested in populations. For selection of genotypes in a plant breeding context, we show how to extract new phenotypes, like growth rates, from the estimated genotypic growth curves and their first-order derivatives. We illustrate our approach on HTP data from the PhenoArch greenhouse platform at INRAE Montpellier and the outdoor Field Phenotyping platform at ETH Zürich.

Testing the adolescent social reorientation model during self and other evaluation using hierarchical growth curve modeling with parcellated fMRI data

Adolescence is characterized as a period when relationships and experiences shift toward peers. The social reorientation model of adolescence posits this shift is driven by neurobiological changes that increase the salience of social information related to peer integration and acceptance. Although influential, this model has rarely been subjected to tests that could falsify it, or studied in longitudinal samples assessing within-person development. We focused on two phenomena that are highly salient and dynamic during adolescence—social status and self-perception—and examined longitudinal changes in neural responses during a self/other evaluation task. We expected status-related social information to uniquely increase across adolescence in social brain regions. Despite using hierarchical growth curve modeling with parcellated whole-brain data to increase power to detect developmental effects, we didn’t find evidence in support of this hypothesis. Social brain regions showed increased responsivity across adolescence, but this trajectory was not unique to status-related information. Additionally, brain regions associated with self-focused cognition showed heightened responses during self-evaluation in the transition to mid-adolescence, especially for status-related information. These results qualify existing models of adolescent social reorientation and highlight the multifaceted changes in self and social development that could be leveraged in novel ways to support adolescent health and well-being.

Using automated feedback to develop writing proficiency

• We examined shape and rate of growth when revising, and unaided and aided transfer. • Growth rate in writing quality decelerated when revising successive drafts. • No gains in first-draft quality across the school year (no unaided transfer). • Gains in revision efficiency and efficacy (aided transfer) associated with AWE. • Implications for implementing AWE within writing instruction are discussed. The present study examined growth in writing quality associated with automated feedback provided by an automated writing evaluation (AWE) software called MI Write and whether prolonged usage of MI Write was associated with gains in students’ independent performance ( i.e ., unaided transfer), measured as first-draft writing quality, and gains in students’ ability to efficiently and effective revise their writing with the aid of automated feedback ( i.e ., aided transfer). Data were collected from 431 students in Grades 4–5 who used MI Write to complete two essays, one in Fall and one in Spring, and revised each essay a minimum of two times. Three-level hierarchical growth curve modeling revealed a non-linear, decelerating shape of growth in writing performance across successive revisions to the Fall and Spring essays. In addition, two-level hierarchical gain score modeling showed that there were no unaided transfer effects for average initial draft scores, but an aided transfer effect was evidenced in students’ gain in growth rates. Based on these findings, we discuss the educational implications of using automated feedback and AWE to improve writing proficiency.

Self-Efficacy in Habit Building: How General and Habit-Specific Self-Efficacy Influence Behavioral Automatization and Motivational Interference.

In this article, we investigate the role of self-efficacy (SE) in intentional habit building. We analyzed event sampling data from a habit building app we created that helps define and track habit data. We used hierarchical growth curve modeling and multilevel mediation to test our hypotheses. In a first study, N = 91 university students built new study habits over a period of 6 weeks in a controlled study. We found that the trait-like (Level 2) general self-efficacy predicted automaticity (i.e., habit strength) but not the experience of motivational interference (MI). In a second study with real user data, N = 265 idiographic habits have been analyzed. The specific SE associated with these habits – habit-specific self-efficacy (Level 1, HSE) – was measured during habit formation. We found that lagged HSE predicted automaticity and that lagged automaticity predicted HSE, indicating a positive feedback mechanism in habit building. Furthermore, we found that lagged HSE predicted less MI during habit performance. A multilevel mediation analysis showed significant effects of lagged HSE (Level 1) and aggregated HSE (Level 2) on MI, which were both partially mediated by automaticity. These results show the importance of defining the specificity of SE beliefs and how they interact with automaticity in the habit building process.

County-Level Factors That Influenced the Trajectory of COVID-19 Incidence in the New York City Area.

More than a century of research has shown that sociodemographic conditions affect infectious disease transmission. In the late spring and early summer of 2020, reports of the effects of sociodemographic variables on the spread of COVID-19 were used in the media with minimal scientific proof attached. With new cases of COVID-19 surging in the United States at that time, it became essential to better understand how the spread of COVID-19 was varying across all segments of the population. We used hierarchical exponential growth curve modeling techniques to examine whether community socioeconomic characteristics uniquely influence the incidence of reported COVID-19 cases in the urban built environment. We show that as of July 19, 2020, confirmed coronavirus infections in New York City and surrounding areas—one of the early epicenters of the COVID-19 pandemic in the United States—were concentrated along demographic and socioeconomic lines. Furthermore, our data provides evidence that after the onset of the pandemic, timely enactment of physical distancing measures such as school closures was essential to limiting the extent of the coronavirus spread in the population. We conclude that in a pandemic, public health authorities must impose physical distancing measures early on as well as consider community-level factors that associate with a greater risk of viral transmission.

THE EFFECTS OF PROCEDURE DURATION AND ATIPAMEZOLE ADMINISTRATION ON HYPERKALEMIA IN TIGERS (PANTHERA TIGRIS) AND LIONS (PANTHERA LEO) ANESTHETIZED WITH α-2 AGONISTS.

A retrospective analysis of 40 extended (>2 hr) anesthetic events in eight lions (Panthera leo) and 32 tigers (Panthera tigris) was performed using a hierarchical linear growth curve model to assess the effects of anesthetic time, α-2 adrenoreceptor agonist dosages, administration of atipamezole, and biochemical parameters on rising plasma K+ concentrations. Hyperkalemia was first noted at a mean time of 187 min (range: 131-226 min), with time under anesthesia as a statistically significant predictor of K+ concentration (P < 0.0001). A significant two-way interaction between time and atipamezole administration (P = 0.0082) for rising K+ concentrations was demonstrated, indicating that administration of atipamezole can mitigate the rise in K+ concentrations. Administration of atipamezole beyond 150 min of anesthetic time was less effective in reducing K+ concentrations than if administered earlier. Electrocardiographic abnormalities were noted in eight animals, including three hyperkalemic individuals. Lions developed significantly greater plasma K+ concentrations than tigers (P = 0.0009) during anesthesia. No biochemical parameter was identified as a significant indicator of which individuals will develop hyperkalemia. Clinicians anesthetizing any large nondomestic felid should monitor electrolytes regularly during anesthetic events; consider early, partial- to full-dose reversal of α-2 agonists; and be prepared to correct potentially life-threatening electrocardiographic abnormalities resulting from hyperkalemia.

The Toronto Concussion Study: a longitudinal analysis of balance deficits following concussion in community-dwelling adults

ABSTRACT Objective To characterize recovery of balance deficits in community-dwelling adults with concussion. Hypothesis Balance measures will improve 2 weeks after injury and persist over 12 weeks. Design Prospective longitudinal observational study. Methods Assessments included the Balance Error Scoring System (BESS) and quiet standing during eyes open, eyes closed, and a cognitive dual task. Recovery was determined using a Hierarchical Growth Curve Model (HGCM) at Week1 (n = 61), Week2 (n = 58), Week4 (n = 53), Week8 (n = 51), and Week12 (n = 39) post-injury. Within-individual follow-up analysis was conducted using the coefficient of variation (quiet standing measures) and a reliable change index (BESS) on 28 individuals with concussion assessed at all 5 time points. Results Self-reported symptom score recovered between Week 4–8. Anteroposterior COP velocity (eyes closed) was the only variable to show statistically significant (p < .05) recovery in the HGCM. The within-individual analysis identified fewer than 43% (12/28) of participants recovered by Week 12, relative to their own Week 1 assessment. Conclusions While recovery of balance deficits was observed in 1 variable over 12 weeks, less than half of the participants included in all assessments demonstrated improvement in balance outcomes. Future research and clinical practice should focus on the unique characteristics of community-dwelling adults with concussion to optimize recovery in this cohort.

Factors Associated With Health Service Use Among Older Adults in a Mobile Veterans Program.

The current retrospective, longitudinal study applied Andersen's Behavioral Model of Health Services Use to examine how demographic characteristics (age), available resources (e.g., a caregiver, the Mobile Veterans Program [MVP]) and health needs (e.g., cognitive and physical functioning, depressive symptoms) affect hospitalization and institutionalization outcomes among older adults using the MVP. Fifty-four Veterans (age 55 to 95) participating in the MVP for up to 2 years were examined using hierarchical linear modeling (HLM) and growth curve model. In the final HLM model, each 1-point increase in depressive symptoms was associated with 76% (p ≤ 0.05) greater risk of institutionalization and 40% (p ≤ 0.01) greater risk of hospitalization. Each 1-point increase in cognitive functioning was associated with 24% (p ≤ 0.05) lower risk of institutionalization. The relationship between caregiver burden and hospitalization was attenuated by frequency of MVP visits. Services focused on reducing depressive symptoms may influence health service use and reduce caregiver burden in this population. [Journal of Gerontological Nursing, 46(5), 15-22.].

Estimating the impact of an integrated CCTV program on crime

Prior research has identified that CCTV cameras play some role in deterring crime in public places. However, increasingly, CCTV surveillance is being integrated with proactive monitoring and police response. This study uses a quasi-experimental design to determine the impact of an integrated CCTV program on crime in Detroit. Data on the first 87 businesses in Project Green Light Detroit was collected, along with a matched comparison group of 201 other businesses. A Bayesian hierarchical linear growth curve model was employed to determine the effect of CCTV camera surveillance at Green Light businesses over the course of 2 years. Findings indicated that Green Light businesses experienced an immediate, but transient increase in property crime reports, as well as more consistent proactive police activity post-intervention. Over time, property crimes and disorder crimes decreased faster relative to the matched controls. There was little evidence of any impact on violent crimes. Integrated CCTV programs may increase the reporting of minor crimes which may have not been reported before. However, these programs may not likely impact violent crimes. In addition, tying proactive response and patrol to participating businesses places higher manpower costs on departments. This has implications to other cities considering implementing wide-spread integrated CCTV programs like Project Green Light Detroit.

Parental Anger and Trajectories of Emotional Well-Being from Adolescence to Young Adulthood.

This study assesses the relationship of parental anger on children's anger and self-derogation from adolescence to young adulthood. We examine a life-course perspective and incorporate theories of emotion regulation and self-referent behavior. Using structural equation models and hierarchical growth curve models with the Kaplan Longitudinal and Multigenerational data, the results indicate that parental anger leads to anger in adolescence, supporting theories of learning through modeling in adolescence. In young adulthood, as predicted by self-referent behavior theory, parental anger contributes to increases in self-derogation. Interestingly, maternal anger, but not paternal anger, is correlated with child's anger and self-derogation.

Estimating loss reserves using hierarchical Bayesian Gaussian process regression with input warping

In this paper, we visualize the loss reserve runoff triangle as a spatially-organized data set. We apply Gaussian Process (GP) regression with input warping and several covariance functions to estimate future claims. We then compare our results over a range of product lines, including workers’ comp, medical malpractice, and personal auto. Even though the claims development of the lines are very different, the GP method is very flexible and can be applied to each without much customization. We find that our model generally outperforms the classical chain ladder model as well as the recently proposed hierarchical growth curve models of Guszcza (2008) in terms of point-wise predictive accuracy and produces dramatically better estimates of outstanding claims liabilities.

The development of academic procrastination in first-year secondary education students: The link with metacognitive self-regulation, self-efficacy, and effort regulation

Academic procrastination has been widely recognized as a problematic but common phenomenon in education. A growing body of literature construes procrastination as a situational, dynamic construct. Yet, little is known about its development in young secondary education students. The current study aims to elucidate this issue while exploring the relation with metacognitive self-regulation, self-efficacy, and effort regulation. Hierarchical growth curve modeling of the development of students' procrastination (566 students, 20 Mathematics/English grade-1 secondary education classes) revealed positive linear trajectories at class level but showed greater variability at student level. All mentioned predictors were negatively associated with procrastination and declined over time, with effort regulation revealing the strongest association. Interaction effects with time revealed a stable procrastination - effort regulation association, whereas the association with metacognitive self-regulation and self-efficacy diminished over time. The findings support the view on academic procrastination as a dynamic construct and highlight the importance of early intervention.

Population Aging and Health Trajectories at Older Ages

Objective This study examines relationships between municipal age structure and two types of self-rated health: general (SRH) and comparison with similar-aged peers (C-SRH). Methods Using a national sample of almost 5,000 Japanese older adults over two decades, we employ hierarchical growth curve models to estimate health trajectories. For municipal age structure, we consider both the relative prevalence of elderly adults in the local population and the pace of aging over time. Results Living in the oldest municipalities was generally associated with worse health, particularly between the ages of 70 and 80 years. For SRH, the speed of municipal population aging was also independently associated with worse health. For C-SRH, worse health in older areas was partially explained by less favorable economic conditions in those municipalities. Results also suggest that higher levels of employment and social integration among older adults living in the oldest municipalities operate in the opposite direction. That is, these attributes partially "protect" individuals from other factors that contribute to worse health. Discussion Relative differences in municipal age structure and the pace of population aging are largely unexplored and potentially important correlates of older adult health. This line of research is increasingly salient in a world with substantial and growing regional variation in population aging.

Population Aging and Health Trajectories at Older Ages.

This study examines relationships between municipal age structure and two types of self-rated health: general (SRH) and comparison with similar-aged peers (C-SRH). Using a national sample of almost 5,000 Japanese older adults over two decades, we employ hierarchical growth curve models to estimate health trajectories. For municipal age structure, we consider both the relative prevalence of elderly adults in the local population and the pace of aging over time. Living in the oldest municipalities was generally associated with worse health, particularly between the ages of 70 and 80 years. For SRH, the speed of municipal population aging was also independently associated with worse health. For C-SRH, worse health in older areas was partially explained by less favorable economic conditions in those municipalities. Results also suggest that higher levels of employment and social integration among older adults living in the oldest municipalities operate in the opposite direction. That is, these attributes partially "protect" individuals from other factors that contribute to worse health. Relative differences in municipal age structure and the pace of population aging are largely unexplored and potentially important correlates of older adult health. This line of research is increasingly salient in a world with substantial and growing regional variation in population aging.

Daily cortisol activity, loneliness, and coping efficacy in late adolescence

Many late adolescents who transition to the college environment perceive changes in psychosocial stress. One such stressor, loneliness, has been associated with numerous health problems among adolescents and adults. The hypothalamic–pituitary–adrenal axis is one mechanism through which loneliness may affect health. Guided by a risk and resilience framework, the present study investigated the association between longitudinal changes in loneliness from high school to college and diurnal cortisol activity (waking levels, cortisol awakening response, diurnal slope) by sampling saliva intensively 5 times a day for 3 weekdays in a US sample of late adolescents in their first semester of college (N = 70; Mage = 18.49, SD = 0.38). The present study also explored how the link between loneliness and cortisol might depend on coping efficacy—one’s belief in successfully coping with future stressors or novel situations. Results from hierarchical linear growth curve models demonstrated that an increase in loneliness across this contextual transition was associated with steeper cortisol slopes in college. Coping efficacy at baseline (in high school) significantly moderated the relation between changes in loneliness and diurnal slopes, such that late adolescents with low levels of coping efficacy who reported increased loneliness across the transition exhibited significantly flatter diurnal slopes in college. Higher levels of coping efficacy at baseline also significantly predicted lower waking cortisol levels during the first semester of college. These results suggest that coping efficacy may serve as a protective factor by contributing to regulation of daily physiological stress activity for late adolescents as they struggle with loneliness across the transition to college.