When reflecting back on the special issues I have gravitated toward over the years, those on innovative methods and statistics immediately come to mind. These types of special issues tend to stimulate interesting conversations and debates, often crossing disciplinary boundaries and providing fruitful insights into new processes and procedures. For the current special issue on advances in methods and statistics in Communication Monographs, I invited some of the most respected scholars within and across disciplines to contribute articles. What is particularly noteworthy about these methodological pieces is that they are inherently tied to advances in theory. A trademark of Communication Monographs is its contribution to communication theory – that manuscripts significantly change and progress the way scholars think about a body of work, concept, or theory related to communication. The manuscripts in the current special issue do just that; they provide innovative ideas across a variety of methods and statistical procedures, while simultaneously addressing larger theoretical issues.
In this special issue, I asked the authors to “push the envelope” in terms of their thoughts about innovative methods and statistics, the state of their field or area of specialization, potential problems with the status quo, and what the future holds. They were urged to provide arguments that not only challenged the readers’ thinking but also challenged their own thinking. Some of the most prominent scholars in their respective areas provide insights on a range of topics, including artificial intelligence (AI) and Big Data, the “brain as predictor” approach in fMRI, rhetorical field methods, critical cultural research in interpersonal communication, Bayesian statistics, a new approach for testing mediation with multiple moderators using Hayes’ PROCESS macro, and a discussion of the benefits, limitations, and ethical uses of Mechanical Turk (mTurk) for collecting data. Each of the articles offers information that not only advances methods and statistics but also provides information that could change the way scholars think about their research and the theories used to guide it.
Two of the articles in the special issue are devoted to advances in statistics. Since his seminar article “Beyond Baron and Kenny: Advances in mediation analysis for the new millennium” published in 2009 in Communication Monographs, Hayes has become the leading expert in mediation analysis. In his article in the current special issue, Hayes extends his Conditional Process Analysis approach to mediation by articulating how scholars can examine multiple moderators in mediation. Because of the complexity of human behavior, it is common for social scientists to examine multiple conditions that might influence mediation. But, these types of statistical models can become overly complex and difficult to analyze. Hayes provides an understandable and parsimonious way for researchers to examine multiple moderators in mediation using his PROCESS macro. As an expert in Bayesian statistics, Rouder et al. also discuss the importance of adopting a Bayesian approach in the field of Communication. Traditional significance testing has long been debated across disciplines. Bayesian model comparison is an approach that other disciplines have increasingly adopted to address the critiques of traditional significance testing. For example, many scholars in the field of Communication rely on small sample sizes (e.g., for observational coding of dyads or groups; clinical samples) that often limit their power to detect significant differences. The Bayesian approach provides a unique set of assumptions and tests that can help with some of these (among many other) sample constraints. On the other end of the spectrum, Bayesian models are also used with Big Data because Bayesian methods use prior knowledge and history to make sophisticated inferences about unknown properties. Rouder et al. provide a brief overview of Bayesian methods so that Communication scholars can determine their utility for their own research.
Elish and boyd’s article on “Situating methods in the magic of Big Data and AI” is an excellent illustration of scholars challenging existing thoughts on a phenomenon. In recent years, universities and commercial markets alike have been enamored with what Elish and boyd refer to as the “magic” of Big Data and AI, with people blindly following and adopting data-driven technologies. Unfortunately, results from machine learning are often void of theory and removed from the context in which they were generated. In this article, Elish and boyd provide a history of Big Data and AI and a propose a framework for conducting such analyses. Specifically, they argue that scholars need to re-envision machine learning as a form of “computational ethnography,” to better situate and ground the computations in the context in which they exist.
Two of the articles challenge scholars’ understanding of communication through critical cultural approaches. For example, even though a critical cultural approach is one of the staples of the field of Communication, it has not been readily adopted in subfields like Interpersonal Communication. Even though family communication scholars are increasingly using critical methods and theories, Suter examines why interpersonal communication scholars might be hesitant to adopt a critical approach, the disservice it is doing to scholars’ understanding of human communication by not using it, and then provides examples of two methods – contrapuntal and intersectional methods – that can be used to analyze data from a critical lens. In her explanation of these methods, Suter discusses four shifts in thinking that characterize a critical paradigm in her critical interpersonal and family communication (CIFC) framework. Taking a different approach, Pezzullo and de Onis discuss the importance of rhetorical field methods in their article. They note that rhetorical field methods are not new, but the way scholars are incorporating them into their scholarship, and with increasing regularity as a way to amplify marginalized voices, is novel. They use environmental rhetoric to illustrate how scholars are using a variety of methodological tools to understand the relationship between human agency and the planet. As they mention in their abstract, “this essay explores how rhetorical field methods have provided and might further offer a compelling set of principles and practices for resisting structures of ecological and social precarity for life on Earth.” They organize these principles and practices around three primary assumptions: culture, interconnection, and voice.
The article by Weber et al. on the advances of fMRI research offers a prime example of the important connection between method and theory. As Weber et al. note, the amount of research using fMRI has increased exponentially over the past 25 years. They discuss the method–theory synergy in fMRI research that has crystalized in the field of Communication, allowing for a “brain-as-predictor” approach whereby researchers can predict and explain population-level patterns from individual-level data with neurons. They discuss how the combination of brain imaging, behavioral and self-report data, content analysis, and the theories that explain human behavior provide more predictive power than traditional methodologies used in isolation.
Finally, researchers have increasingly relied on companies that supply affordable, community-based participation pools in their attempt to reduce their reliance on college student samples. Perhaps as a reaction to receiving an enormous amount of manuscripts using mTurk and wanting more established guidelines for its use, I asked one of the most well-known scholars who publishes articles on mTurk – Dr. Kim Sheehan – to write an article describing it, how the data are generated, how the data are similar or different to the larger U.S. population and data gathered from other platforms, its strengths and limitations, and ethical considerations when using it and similar platforms. The goal of this particular article is to provide scholars with ethical guidelines for using mTurk for their research.
Not only are the articles in this special issue thought-provoking and innovative methodologically but also they are innovative theoretically. The arguments presented in these articles push researchers to think “outside the box” in their methods and procedures and, in so doing, encourages them to question their long-held assumptions about why certain findings might exist in the first place. Finally, the manuscripts also provide practical guidelines for researchers to “do” their research better. Hopefully, this special issue encourages current and future generations to consider new approaches to conducting their research.
Disclosure statement
No potential conflict of interest was reported by the author.