Methodological Advances in Marketing & Consumer Research
(A2025-125742)

AUTHORS

Hauke Roggenkamp, University of St. Gallen; Antonia Krefeld-schwalb, Rotterdam School of Management, Erasmus University; Felix Eggers, Copenhagen Business School; Constant Pieters, Copenhagen Business School

ABSTRACT

Rigorous marketing research requires validity: experiments should reflect reality, findings must generalize, models should make accurate predictions, and distinct constructs should be captured by distinct measures. This special session focuses on methodological advances that address each of these challenges, introducing new frameworks and tools for enhancing ecological validity in experiments, strengthening research generalizability, improving predictive accuracy, and establishing measurement validity of constructs. Overall, this session provides the following methodological perspectives on how to assess and enhance validity. “DICE: Advancing Social Media Research Through Digital In-Context Experiments” by Hauke Roggenkamp (University of St. Gallen), Johannes Boegershausen (Erasmus University Rotterdam), and Christian Hildebrand (University of St. Gallen) We introduce Digital In-Context Experiments (DICE), a research paradigm with accompanying app that enables researchers to create scrollable social media feeds while maintaining experimental control. Through unobtrusive behavioral tracking, DICE captures individual-level engagement data for each social media post. Two case studies demonstrate how DICE enables research questions around context effects and attention mechanisms. “From Reproducibility Crisis to Generality Revolution” by Antonia Krefeld-Schwalb (Erasmus University), Xuwen Hua, Eli Sugerman, and Eric Johnson (Columbia University) In response to the criticism of consumer research’s lack of replicability, researchers have started to collect larger samples, and more studies. For that, researchers typically rely on selected online panels, that systematically differ on relevant moderators affecting the generalizability of results. We make 6 recommendations to overcome that problem. “Empirical Generalizations about Predictive Accuracy in Choice Experiments” by Felix Eggers (Copenhagen Business School) and John Hauser (MIT) Hit rates are widely used as a measure of validity in choice experiments. To explore drivers of predictive accuracy, we conduct a meta-analysis of holdout hit rates, examining variables related to research context, experimental design, estimation methods, and implementation quality. The results offer potential validation benchmarks and insights into factors that affect validity. “A Proposed Framework for Discriminant Validation of Multi-Item Marketing Scales” by Constant Pieters (Copenhagen Business School), Hans Baumgartner (Pennsylvania State University), and Rik Pieters (Tilburg University) It was unclear which fundamental requirements discriminant validation criteria should satisfy, which motivated a conceptual and empirical analysis of discriminant validation criteria. Based on that analysis, this paper proposes a decision-theoretic framework for discriminant validation of multi-item marketing scales. Case studies, annotated example code, and an online application illustrate the benefits of using the proposed framework.