Effect of Cognitive Bias and Gender on Digital Voice Assistant Shopping
(A2024-119445)

AUTHORS

Gaia Rancati, Middle Tennessee State University; Carsten Schultz, University of Hagen; Maurizio Mauri, Università Cattolica del Sacro Cuore

ABSTRACT

Digital voice assistants (DVAs) have evolved into recommendation agents assisting users in their online shopping by supporting their decision-making. These assistants can also lead to gender stereotyping and potentially cause cognitive biases that influence customers’ purchasing behavior. This study examines the potential of DVAs as recommendation agents in terms of gender bias and social proof. To answer this research question, we created a multidisciplinary research protocol involving a scenario-based survey and a neuroscientific experiment. In the survey scenario of the male DVA, prior experience is more important for women than men. For the female DVA scenario, perceived usefulness has a greater impact on purchase intention for men than women, whereas women’s attitudes are more relevant than men’s. Implicit association measures reveal that female DVA elicits more short-term associations related to gender stereotypes compared to male DVA. Facial expression analysis shows higher attention during product selection. The skin conductance arousal also supports an opposite attraction hypothesis for DVAs’ and participants’ gender. Most importantly, DVA recommendations outweigh traditional social proof heuristics in customer product selection.