Laboratory experiments have many benefits and serve as a powerful tool for social psychology research. However, relying too heavily on laboratory experiments leaves the entire discipline of social psychology vulnerable to the inherent limitations of laboratory research. We discuss the benefits of integrating archival research into the portfolio of tools for conducting social psychological research. Using four published examples, we discuss the benefits and limitations of conducting archival research. We also provide suggestions on how social psychological researchers can take advantage of the benefits while overcoming the weaknesses of archival research. Finally, we provide useful resources and directions for utilizing archival data. We encourage social psychologists to increase the robustness of this scientific literature by supplementing laboratory experiments with archival research.

Social psychology has a long and respected tradition of conducting laboratory experiments. There are clear benefits to conducting such experiments. Most notably, laboratory experiments include the elements of contextual control and random assignment to treatment and control groups that when utilized properly allow researchers to draw causal inferences (Cook & Campbell, 1979). Delineation of causality allows for the generation and refinement of psychological theories, and aids in the understanding of how to influence psychological phenomena. Laboratory experiments are tailor made to facilitate these inferences, making them an extremely powerful and useful tool for conducting social psychological research (Falk & Heckman, 2009). Despite their many strengths, laboratory experiments have important limitations. Artificial settings may miss important elements of real world contexts (Kerlinger, 1986), and demand characteristics in such artificial settings can distort construct relationships (Klein et al., 2012). Laboratory experiments are often conducted with relatively small samples, which may lead to unstable parameter estimates and invalid inferences (Hollenbeck, DeRue, & Mannor, 2006), and undermine the reliability of replications (Fraley & Vazire, 2014; Open Science Collaboration, 2015). Some of these limitations contribute to what some are calling a “crisis of confidence” in psychology (Baumeister, 2016; Hales, 2016; Pashler & Wagenmakers, 2012). Also, impracticality of random assignment of some characteristics potentially narrows the range of topics that can be studied in laboratory experiments (Doss, Rhoades, Stanley, & Markman, 2009; Sbarra, Emery, Beam, & Ocker, 2014). Archival research has the potential to address many of these limitations and is therefore a promising complementary research approach to the traditional laboratory experiments. Archival research entails analyzing data that were stored other than for academic research purposes1 . This research approach has frequently been utilized in other fields (e.g., economics, sociology, and developmental psychology; Cherlin, 1991; Shultz, Hoffman, & Reiter-Palmon, 2005), but remains severely underutilized in social psychology. A search of the published articles in three top social-psychology journals (Journal of Personality and Social Psychology, Psychological Science, and Journal of Experimental Social Psychology) in 1996, 2006, and 2016 reveals that archival studies were used in < 1% of the published studies across three decades, meaning that only a small subset of the social psychology literature uses archival research. This underrepresentation of archival research is evident in spite of the high-impact archival studies that have been done in the field, such as Cohn, Mehl, and Pennebaker’s (2004) study on linguistic markers of psychological change after the September 11 attacks, Alter and Oppenheimer’s research (2008) on the effects of fluency, and Sales’ (1973) investigation of threat as a cause of authoritarianism. Considering that the digital universe will more than double every two years from 2013 to 2020—from 4.4 trillion to 44 trillion gigabytes (International Data Corporation, 2014), archival research can be a fruitful and robust methodology for social psychologists to investigate social phenomena. Yet despite the vast amount of data available, only half of 1% of newly created digital data have been analyzed (MIT Technology Review, 2013). In recent years, tools for the assembly of relevant datasets have become widely available to researchers, including notable examples such as Google Trends, Twitter tags, and online marketplace bidding logs. Clearly, the “Big Data” revolution is beginning to alter the research landscape by turning archival research into a promising methodological option for research.