Modeling language fluency using latent growth curve models

Abstract

Latent growth curve modeling (LGCM) is a technique that is used to examine the changes occurring in a variable over time. LGCMs can be formulated using a structural equation modeling approach, the latter being a technique used to analyse the relationships between one or more independent variables and one or more dependent variables, where variables may be either observed or latent. Latent variables are variables that cannot be measured directly, while observed variables in a LGCM are repeated measures used to represent these latent variables. In using a structural equation model (SEM) formulation, the process towards attaining a LGCM is broken down into five main stages: model specification, identification, estimation, testing, and if necessary, respecification. This thesis provides the theoretical framework involved in fitting a SEM and how it translates to fitting LGCMs from the perspective of a SEM. The LGCM procedure is applied to a real-life longitudinal dataset provided by a local English language school, with the primary focus being placed on students’ fluency over time. In addition to modeling the changes in fluency across time points, further analysis shall be conducted to determine whether variables such as age, gender, and students’ class affect the level of fluency.