Addressing Perceived Resistance to Biometric Security Systems in Airports: Exploration With a General Bayesian Network-Based Decision Support System

Introduction

In most enterprises, technology management and its associated decision-making processes have traditionally been the domain of senior staff and board members. However, in today’s competitive global market, relying solely on intuition or hierarchical decision making is no longer feasible. As data analytical models, including machine learning, have become practical, the adoption of model-based, auditable decision processes has become essential. Such a logical framework is referred to as a decision support system (DSS), an information system designed to support interactive decision making for various purposes (Gupta et al., 2007), including technology management. Modern DSS deployments increasingly incorporate machine learning with explainability features and scenario-based evaluations to enable managerial “what-if” reasoning.

DSSes also serve as valuable tools in the field of innovation studies, as they help formulate appropriate strategies for introducing new technologies to the market. By doing so, they increase the likelihood of successful consumer adoption and reduce the risk of failure (Verleye & De Marez, 2005). In some instances, a DSS can replicate the cognitive process a user undergoes when selecting a new product or service. Much like users, DSSes can be trained to evaluate whether to accept or decline an innovation. Recent evidence in biometric and authentication contexts suggests that adoption hinges on perceived risk, controllability, and trial-like exposure—factors that a DSS can surface and manipulate via scenario analysis (Wang, 2021; Yang et al., 2024; Yu et al., 2024). Typically, users who perceive benefits from an innovation promote its diffusion within the service ecosystem. Yet, adoption tells only half the story; resistance operates through its own distinct determinants and requires explicit modeling.

Innovation resistance theory (IRT) frames resistance not merely as the absence of adoption but as a proactive, autonomous response stemming from specific barriers: (a) usage issues like complexity and incompatibility, (b) value concerns such as limited relative advantage, (c) risks (functional, physical, social, and temporal), and (d) tradition or image factors (Ram, 1987; Ram & Sheth, 1989; Sheth, 1981). This perspective underscores that users may postpone, reject, or oppose innovations for reasons unrelated to diffusion speed or personal “innovativeness,” necessitating the direct modeling of resistance rather than inferring it from adoption rates.

IRT complements acceptance models like the technology acceptance model (TAM) and unified theory of acceptance and use of technology (UTAUT), which connect beliefs to intentions and actual use (Davis, 1989; Venkatesh et al., 2003). Central to these models are perceived usefulness (or performance expectancy) and perceived ease of use (or effort expectancy), which map directly onto IRT barriers. For example, functional, physical, social, and temporal risks erode perceived usefulness and heighten perceived effort; compatibility strengthens both expectancies; and trialability lowers uncertainty via low-stakes and reversible trials, boosting overall expectancy beliefs. Studies in biometric and authentication domains confirm these linkages, positioning risk, controllability, and trial experiences as pivotal determinants of acceptance (Wang, 2021; Yu et al., 2024).