Latent Profile and Transactional Analyses for Cohort Panel Data



Journal Title

Journal ISSN

Volume Title



Longitudinal data refers to research data collected from the same subjects or entities over a prolonged period of time. In longitudinal studies, data is collected at multiple time points, allowing researchers to examine and analyze changes or patterns that develop within the individuals or entities over time. The ability to continually observe and monitor the same individuals or groups is the primary property of longitudinal data, which offers insights into the emergence, development, or patterns of numerous phenomena or variables of interest. Researchers can study correlations and associations between factors as well as individual changes using this sort of data collecting. Longitudinal studies can be carried out in various fields, including psychology, sociology, education, and health sciences. They are especially useful for studying developmental processes, analyzing the long-term effects of interventions or treatments, investigating long-term outcomes, and identifying patterns of change or stability in behaviors, attitudes, or other measurable factors. Analyzing longitudinal data can be approached from both a variable-centered perspective and a person-centered perspective. The variable-centered approach focuses on examining changes and relationships among variables over time. In this approach, the emphasis is on analyzing the average patterns and associations within the sample. Common techniques used in the variable-centered analysis of longitudinal data include but are not limited to growth curve modeling, latent growth modeling, autoregressive models, and cross-lagged panel analysis. The person-centered approach focuses on locating people or subgroups of persons who display distinctive patterns of change across time. This strategy places an emphasis on the variability and uniqueness of each person's developmental pathways. Some techniques for longitudinal data analysis that focus on the individual include but are not limited to trajectory analysis, cluster analysis, and person-centered regression. It is important to remember that the research topic, the type of data, and the precise analytic goals all influence the approach that is chosen. A thorough comprehension of longitudinal data is frequently possible by combining approaches that are both variable- and person-centered. This study used many longitudinal models to explore the phycological related topics for people from different culture backgrounds and different ages. Chapter I provides a brief discerption of all models and methods that are used in the study. Chapter II explores the appetitive traits and body mass index in Chinese adolescents with unconditional latent growth curve analysis and parallel latent growth curve modeling. Chapter III explores the patterns of cognitive functions in older adults and their associations with demographic, diet, and lifestyle by using latent profile analysis, multinomial logistic regression, and latent transition analysis. Chapter IV examined the parental control in TV watching and its associations with demographic information, dietary knowledge, and physical activity among Chinese adolescents by using latent class analysis with covariates, latent class analysis, and latent transition analysis.



longitudinal study, Chinese, adolescents, older adults