Loading...

Media is loading
 

Description

Algorithms are increasingly offered for human decision-making processes to support individuals with sophisticated data-driven insights in a variety of decision domains. Despite this promising potential, prior findings on the perceptions of algorithmic decision-making are ambiguous. This has led to a vivid discussion regarding the comprehensibility and fairness perceptions associated with human and algorithmic decision agents that also emphasized the role that decision-related factors can play in determining the evaluation of a certain decision. In a preregistered online experiment with 400 participants, we find that differences in the perceptions of decision transparency and fairness can be rather ascribed to the processes and rules applied to arrive at the decision and the decision outcome's sentiment than to the decision agent. However, being confronted with a negative decision outcome in a situation that is characterized by uncertainty, leads to a preference for human decision agents.

Share

COinS
 
Jan 17th, 12:00 AM

Just like you like it - The effects of transparency and decision outcome on the evaluation of human and algorithmic decision-making

Algorithms are increasingly offered for human decision-making processes to support individuals with sophisticated data-driven insights in a variety of decision domains. Despite this promising potential, prior findings on the perceptions of algorithmic decision-making are ambiguous. This has led to a vivid discussion regarding the comprehensibility and fairness perceptions associated with human and algorithmic decision agents that also emphasized the role that decision-related factors can play in determining the evaluation of a certain decision. In a preregistered online experiment with 400 participants, we find that differences in the perceptions of decision transparency and fairness can be rather ascribed to the processes and rules applied to arrive at the decision and the decision outcome's sentiment than to the decision agent. However, being confronted with a negative decision outcome in a situation that is characterized by uncertainty, leads to a preference for human decision agents.