The Information Systems research field is inherently dynamic. New technologies, new standards, new legislation, and changing user expectations are some of the reasons why topics of interest to the IS field remain in flux. As researchers, we seek to uncover and explain relationships among variables, but due to the dynamism of IS phenomena, these relationships are apt to change over time. For example, the effect of informational features such as product diagnosticity or seller reputation on the price of an electronic commerce transaction is likely to change over time as users become more comfortable with online trading. This paper describes several statistical methods to model these changes in relationships. Specifically, we discuss methods to investigate time-varying coefficients in regression models, including rolling regression, “parameterizing” the coefficients using process functions, and testing for structural change. Importantly, we describe how the structure of many of the data sets used in IS research differs from that of data sets often used in other fields such as finance, economics, or marketing. This has implications for the investigation of time-based effects. We illustrate each method using a data set gathered from the wholesale automotive market, which not only helps us explain the methods, but also allows us to investigate the evolution of market practice in one empirical context. Thus, we address both methodological and substantive issues. Given that our field is inherently dynamic, an understanding of how effects change over time should be central to the overall IS research agenda. This paper is designed to familiarize IS researchers with methods available for this purpose.