Time series analysis as an emerging method for researching L2 affective variables.

In language studies, marked by a myriad of psychological and social, and linguistic factors, linear modeling fails to represent the creativity, irregularity and emergent patterns of behavior. To adequately represent the dynamicity and complexity of psychological or affective variables, time-sensitive non-linear modeling is needed, especially the time series analysis (TSA), which accommodates incompatibility over time. TSA is a mathematical framework that can effectively show whether and to what degree the measured time series represent nonlinear variation through time. TSA makes prediction or retrodiction of complex and dynamic phenomena possible in future or past and, thus, can significantly contribute to the unraveling of the nuanced changes in the progress of different learner-related constructs during learning a new language. The present paper, at first, offers an introductory overview of the TSA, and then pinpoints its technical features and procedures. Exemplary works of research in language studies will be reviewed next, followed by useful conclusive remarks about the subject. Finally, suggestions will be made for further investigation of language-related affective variables using this innovative method.