Time series analysis for psychological research: examining and forecasting change.

Andrew T Jebb,Wei Wang,Qiming Huang,Louis Tay

doi:10.3389/fpsyg.2015.00727

Abstract

Psychological research has increasingly recognized the importance of integrating temporal dynamics into its theories, and innovations in longitudinal designs and analyses have allowed such theories to be formalized and tested. However, psychological researchers may be relatively unequipped to analyze such data, given its many characteristics and the general complexities involved in longitudinal modeling. The current paper introduces time series analysis to psychological research, an analytic domain that has been essential for understanding and predicting the behavior of variables across many diverse fields. First, the characteristics of time series data are discussed. Second, different time series modeling techniques are surveyed that can address various topics of interest to psychological researchers, including describing the pattern of change in a variable, modeling seasonal effects, assessing the immediate and long-term impact of a salient event, and forecasting future values. To illustrate these methods, an illustrative example based on online job search behavior is used throughout the paper, and a software tutorial in R for these analyses is provided in the Supplementary Materials.

Highlights

The first step in autoregressive integrative moving average (ARIMA) modeling is to visually examine a plot of the series’ autocorrelation function (ACF) to see if there is any autocorrelation present that can be used to improve the regression model—or else the analyst may end up adding unnecessary terms
The first step in ARIMA modeling is to visually examine a plot of the series’ ACF to see if there is any autocorrelation present that can be used to improve the regression model—or else the analyst may end up adding unnecessary terms
7The use of additional fit indices, such as the Akaike information criterion (AIC) c and Bayesian information criterion (BIC) is recommended, but we focus on the AIC here for simplicity

Summary

A Note on Software Implementation

Conceptual expositions of new analytical methods can often be undermined by the practical issue of software implementation (Sharpe, 2013). Interrupted Time Series Analysis Overview we are interested in describing the underlying trend within the Google time series as a function of time, we are interested in the effect of a critical event, represented by the following question: “Did the 2008 economic crisis result in elevated rates job search behaviors?” In psychological science, many research questions center on the impact of an event, whether it be a relationship change, job transition, or major stressor or uplift (Kanner et al, 1981; Dalal et al, 2014). Interrupted time series analysis requires that events be discrete, this conceptual problem can be managed in practice; selecting a point of demarcation that generally reflects when the event occurred will still allow the statistical model to assess the impact of the event on the level and trend of the series. In describing ARIMA modeling, the following sections take the form of those discussing regression methods: Conceptual and mathematical treatments are provided in complement in order to provide the reader with a more holistic understanding of these methodologies

Introduction

Procedure

Findings

Conclusion