Paper Type
Short
Paper Number
PACIS2025-1716
Description
This study introduces LRNS, a large language model (LLM)-based reciprocal nudging system designed to improve the choices made by the Choice Architect (Nudger) during the design of a digital nudge While current implementations influence user choices, they often overlook how the Nudger interprets behavioral feedback from users (Nudgee), which may lead to biased design choices. Grounded in Social Cognitive Theory and Reciprocal Determinism, this research reconceptualizes existing digital nudge feedback loop designs by exploring the integration of reciprocal interaction mechanisms with an LLM-based system. This research proposes a Design Science Research approach to develop LRNS, utilizing recent advances in LLMs to effectively “Nudge the Nudger”. By analyzing how reciprocal interactions influence digital nudge design, the study proposes a replicable LLM-based design. Ultimately, this study aims to demonstrate how the concept of reciprocal nudging can improve feedback interpretation, enabling the Nudger to design more informed and effective digital nudge interventions.
Recommended Citation
Boyce, James; Phan, Thuy Linh (Isabella); Namvar, Morteza; Akhlaghpour, Saeed; and Risius, Marten, "Who Nudges the Nudger? Conceptualizing an LLM-Based Reciprocal Nudging System (LRNS)" (2025). PACIS 2025 Proceedings. 11.
https://aisel.aisnet.org/pacis2025/hci/hci/11
Who Nudges the Nudger? Conceptualizing an LLM-Based Reciprocal Nudging System (LRNS)
This study introduces LRNS, a large language model (LLM)-based reciprocal nudging system designed to improve the choices made by the Choice Architect (Nudger) during the design of a digital nudge While current implementations influence user choices, they often overlook how the Nudger interprets behavioral feedback from users (Nudgee), which may lead to biased design choices. Grounded in Social Cognitive Theory and Reciprocal Determinism, this research reconceptualizes existing digital nudge feedback loop designs by exploring the integration of reciprocal interaction mechanisms with an LLM-based system. This research proposes a Design Science Research approach to develop LRNS, utilizing recent advances in LLMs to effectively “Nudge the Nudger”. By analyzing how reciprocal interactions influence digital nudge design, the study proposes a replicable LLM-based design. Ultimately, this study aims to demonstrate how the concept of reciprocal nudging can improve feedback interpretation, enabling the Nudger to design more informed and effective digital nudge interventions.
Comments
HCI