擬人化效應和投資人自我控制感對理財機器人使用意願的影響：以對理財機器人的信任為中介變數、以投資人專業度為調節變數

Abstract

While the adoption of robo-advising services has expanded exponentially over the past decade, there are still very few empirical studies that have investigated how investors feel towards such robo-advisors. Prior research has explored people’s trust towards automated algorithms and anthropomorphized technology within the medical, automobile, and service industries, but has not yet examined the effects of humanizing A.I. products in the realm of financial services. To address the gap, this paper investigates how investors’ trust are influenced by varying levels of humanization perceived in robo-advisors as well as investors’ sense of control felt over robo-advisors. This paper studies the role of humanization in robo-advisors, investors’ sense of control, the moderating roles of investors’ expertise in terms of (1) past experience, an objective metric, and (2) familiarity, a subjective assessment, and their effects on investors’ trust in robo-advisors. It also extends the research framework to examine whether trust fully or partially mediates the relationship between investors’ adherence to advisors’ recommendation and the two independent variables, perceived humanization and sense of control. Study 1 empirically investigates all the aforementioned variables except sense of control, whereas Study 2 incorporates all the variables in the experiment. Both studies observed a positive correlation between perceived humanization and trust (H1), trust and adherence to recommendation (H2), and a positive moderating effect between perceived humanization and experience on adherence (H3a-2), whereas Study 2 saw a positive correlation between sense of control and trust (H4). Neither studies detected a moderation effect between familiarity and perceived humanization on trust (H3b-1) or adherence (H3b2), perhaps because familiarity was subject to individual bias and skewed scales across participants. These findings make theoretical contributions to the psychology, management, and marketing papers in numerous ways. It provides more insight into the mechanisms behind investors’ decision-making processes when dealing with humanized and unhumanized robo-advisors. Furthermore, this paper provides several managerial interventions to increase consumer acceptance of A.I. financial advisors: (1) robo-advisors should be appropriately humanized, especially when dealing with experienced investors, (2) financial companies should enhance investors’ illusion of control over the process of seeking financial advice, (3) firms can manipulate such sense of control by offering investors several options of robo-advisors to choose from and by giving investors the final say over their investment decisions. However, this research does present some limitations and suggestions for fruitful avenues in future work: (1) identify the specific type(s) of humanization that optimizes perceived levels of humanization across all investors, (2) explore other variables that could affect investors’ trust and reliance on robo-advisors, (3) develop more sensitive scales to measure and capture investors’ trust as well as both aspects of expertise, and (4) investigate how trust will be affected if investors observe errs or subpar performances from humanized robo-advisors.