Algorithmic bias: review, synthesis, and future research directions

As firms are moving towards data-driven decision making, they facing an emerging problem, namely, algorithmic bias. Accordingly, systems can yield socially-biased outcomes, thereby compounding inequalities in the workplace and society. This paper reviews, summarises, synthesises current literature related to bias makes recommendations for future information research. Our analysis shows that most studies have conceptually discussed ethical, legal, design implications of bias, whereas only a limited number empirically examined them. Moreover, mechanisms through which technology-driven biases translate into decisions behaviours been largely overlooked. Based on reviewed papers drawing theories such as stimulus-organism-response theory organisational justice theory, we identify explicate eight important theoretical concepts develop research model depicting relations between those concepts. The proposes affect fairness perceptions technology-related machine-generated recommendation acceptance, algorithm appreciation, system adoption. also contextual dimensions (i.e., individual, task, technology, organisational, environmental) influence perceptual behavioural manifestations These propositions highlight significant gap provide roadmap studies.