Quantitative Approaches for Analyzing Longitudinal Data in Second Language Research

Abstract

This article discusses methods used in second language (L2) research to analyze quantitative longitudinal data. Longitudinal studies are experimental and nonexperimental studies that collect repeated measures of the same variable(s) from the same participant(s) at two or more time points. Three challenging areas in longitudinal L2 research are first discussed: study design, measurement, and data analysis and modeling. Next, various traditional and recent quantitative approaches for analyzing longitudinal data are discussed, including difference or gain scores, repeated measures univariate and multivariate analysis of variance (RM ANOVA, MANOVA), multilevel modeling (MLM), autoregressive models and latent growth curve modeling (LGCM) within the structural equation modeling (SEM) framework, item response theory (IRT), single-case research designs and time series analysis (TSA), and event history analysis (EHA). Longitudinal L2 studies published in the last 10 years are reviewed to identify trends and patterns in the use of different quantitative approaches for analyzing longitudinal L2 data, describe best data analysis practices in such research, and provide recommendations for future longitudinal L2 studies. It is argued that, when feasible and appropriate, recent approaches (e.g., MLM, LGCM) have several conceptual, methodological, and practical advantages and can stimulate the development and empirical examination of more complex questions and models concerning L2 development over time than is possible with traditional techniques.