Well-done and well-used: State-of-the-art & AI-based approaches for optimizing the production and utility of meta-analyses in consumer research
(A2024-119446)

AUTHORS

Gratiana Pol, Hyperthesis; Martin Eisend, European University Viadrina; Joseph Riley, Europa University Viadrina; Dominika Niewiadomska, Europa University Viadrina

ABSTRACT

Meta-analyses are an important synthesis method in consumer research (Grewal et al., 2018), whose volume has been growing exponentially, consistent with similar trends in the behavioral sciences (Sharpe & and Poets, 2020). This development poses several challenges. On the research production front, synthesizing increasingly larger bodies of literature is painstaking and error-prone (Schmidt et al., 2020), often deterring researchers who wish to conduct a meta-analysis—not surprisingly, we find that 84% of authors who publish a consumer research-focused meta-analysis never publish a second. On the research utilization front, it is nearly impossible to keep track of all meta-analyses and effectively parse and integrate their findings into meaningful conclusions—which may explain why results from consumer-focused meta-analyses rarely make their way into practice. While most academic marketing efforts around meta-analyses have focused on statistical methods education and software for running meta-analyses (Borenstein et al., 2021, Coopar et al., 2019), this session focuses on state-of-the-art approaches and insights related to less emphasized aspects of the meta-analysis creation and utilization process: extracting/coding the raw data (Presentation 1) and integrating/comparing meta-analytical finding (Presentations 2, 3). We also discuss opportunities and pitfalls associated with the use of generative AI-based chatbots—such as ChatGPT—for generating and integrating meta-analytical knowledge in consumer research. Current Data Extraction Practices in Meta-Analysis Research and How to Improve Them with the Help of AI (Gratiana Pol, Dominika Niewiadomska, Joseph Riley, Martin Eisend, Jade Winn, Olga Koz, Rick Wedgeworth, Jude Calvillo). We introduce a generative AI-based data extraction approach, meant to make the process of conducting meta-analyses more accessible, efficient, and precise. A Visual Synthesis Approach for Enhancing the Utility of Meta-Analyses for Consumer Researchers and Practitioners (Gratiana Pol, Martin Eisend, Joseph Riley, Dominika Niewiadomska, Abhishek Borah, Deborah MacInnis, Calvillo, Rick Wedgeworth, Roy Nijhof, Luciano Silvi). We introduce an innovative, award-winning approach for visually synthesizing findings from consumer-focused meta-analyses, via interactive causal maps that provide conceptual overviews of the literature and can facilitate hypothesis generation. How Much Have We Learned About Consumer Research? A Meta-Meta-Analysis (Martin Eisend, Gratiana Pol, Joseph Riley, Dominika Niewiadomska, Rick Wedgeworth). We provide high-level insights about knowledge development in consumer research and practical suggestions for consumer researchers, derived from a consumer-focused meta-meta-analysis conducted in conjunction with the approach discussed in Presentation 1.