1. 発達研究における縦断データの解析手法: 成長曲線モデルと潜在クラス成長分析.
- Author
-
西村 倫子
- Abstract
I Research on child development considers it important to understand each childʼs developmental process as well as that of the entire population. Therefore, child development scholars must be proficient in two methodologies—longitudinal research and longitudinal analysis—with the latter based on drawing developmental trajectories. This study focuses on describing the populationʼs average developmental trajectory while capturing individual deviations from the average. To this end, this study introduces the growth curve model and latent class growth analysis, highlighting findings from the Hamamatsu Birth Cohort (HBC) Study. The growth curve model introduces the mixed-effects and latent class approaches using an example question of whether an individualʼs birth weight affects their expressive language development. Latent class growth analysis emphasizes the parallel-process approach, which processes multiple domains in parallel and the joint model, which can be used to examine links between the developmental trajectories of two outcomes. [ABSTRACT FROM AUTHOR]
- Published
- 2022