Abstract
With the increasing use of information and communication technologies, there is a wealth of longitudinal data available, which open up new research directions. This availability necessitates special analytical tools, namely time series analysis methods. The paper focuses on Auto Regressive Integrated Moving Average (ARIMA) modeling and provides an outline of how it can be used in social sciences to study dynamic social processes. It provides a typology of dynamics of social processes, using the distinctions between stability vs. fluctuation of a communication process and exogenous vs. endogenous changes. Five distinct types of dynamics of social processes are outlined: stability; linear trend; different attractors; permanent effect; and not permanent effect. Further, the paper examines how these types can be analyzed with the use of ARIMA modeling, and what this means for understanding of the underlying social process. Conclusions are drawn for the use of ARIMA in social sciences, and for understanding of dynamics of social processes.
Notes
1. This was preferred over the widely used Durbin-Watson statistic, which only tests for first-order autocorrelation and is not valid when the model includes AR terms.