Publication Date

2020

Document Type

Thesis

Committee Members

Ion Juvina, Ph.D. (Advisor); Kevin A. Gluck, PhD. (Committee Member); Joseph Houpt, Ph.D. (Committee Member); Valerie L. Shalin, Ph.D. (Committee Member)

Degree Name

Master of Science (MS)

Abstract

Social interactions are complex and constantly changing decision making environments. Prior research (Mayer, Davis, & Schoorman, 1995) has found that people use their trust in others as a criterion for decision making during social interactions. Trust is not only relevant for human-human interaction, but has also been found to be important for human-machine interaction as well, which is becoming a growing feature in many work domains (De Visser et al., 2016). Prior research on trust has attempted to identify the behavioral characteristics an individual (trustor) uses to assess the trustworthiness of another (trustee) to determine the trustor's level of trust. Experimental findings have been used to develop into various models of trust (Mayer et al., 1995; Juvina, Collins, Larue, Kennedy & de Mello, 2019) to explain how a trustor comes to trust a trustee. An aspect of trust that has not been investigated is how or if trust changes when a trustor attempts to interact with a trustee, but cannot interact with the trustee. Under such situations Juvina et al.’s (2019) trust model makes the novel prediction that trust will decrease. To assess the prediction of Juvina et al. (2019) model, a new experimental design (the multi-arm trust game) was developed to evaluate how trust is affected under conditions where an individual variably interacts with multiple trustees. Additionally, the identity the trustee (human and machine) was manipulated to examine differences between human-human and human-machine trust. Before data were collected, the model made ex-ante predictions of the participants’ behavior. The accuracy of these predictions was then evaluated after the data were collected. The results from our experiment found that our model was able to predict general characteristics of the data confirming the necessity of the model’s discounting mechanism, while also highlighting model limitations that are areas for future research.

Page Count

133

Department or Program

Department of Psychology

Year Degree Awarded

2020


Share

COinS