Amreeta Leddy
Algorithmic bias is when users receive unequal outcomes when using an algorithm. This can include functional aspects of algorithmic bias which allow users to filter content to their preference, but also problematic bias e.g. unequal outcomes when users are underrepresented in training data. This study investigated whether an educational video could impact efficacy beliefs towards algorithmic bias in organisational and individual use algorithms.
Algorithms play an important part in daily life, often making decisions and functioning in the background. As psychology continues to adapt, it is important to explore whether prior theory can be generalised to modern contexts. Efficacy beliefs or people's confidence towards algorithmic bias is an understudied topic, despite being highly relevant. The current study aims to fill a gap in the literature by exploring the impact of prior experience, via an educational video, on the efficacy beliefs of participants towards algorithmic bias within organisational and individual-use algorithms. This study recognises that algorithm bias is both a risk and a functional aspect (e.g. liking content to filter preferences). Thus, this study acknowledges the interactivity and nuance of algorithmic bias.
This study displayed how an educational video made the participants aware
of biased outcomes due to algorithmic bias, which significantly lowered their efficacy beliefs towards both domains of algorithmic bias. The efficacy beliefs of participants were significantly lower towards algorithmic bias in organisational algorithms. This could have been due to the omission of aspects of Sundar's (2020) action path HAII TIME theory. Both the educational video and algorithm type (organisational and individual use) had an independent effect on efficacy beliefs. The findings of this study expand on Bandura’s (1986) theory of efficacy beliefs, that emphasise contextual understanding of a task as a foundation
of efficacious techniques. The findings imply increased understanding can sometimes decrease efficacy beliefs, specifically regarding a discriminatory system. The workarounds for algorithmic bias that are presented within academia must be communicated to users in an accessible format, for example, potentially, though an educational video (Overbye-Thompson & Rice, 2025).
Hello! My name is Amreeta Leddy I have a BSc in Applied Psychology and Technology. My interests lie in the intersection of Psychology and Technology for an equitable future. I was the co-chairperson of IADT Psychological society from 2024-25 and spearheaded the planning and running of the All-Ireland National Student Congress 2025. This involved writing and moderating a panel for Equity, Diversity, and Inclusion in Research and Practice in Psychology, which was in attendance of the PSI president. I also have been a volunteer for the past 5 years in various organizations. My experiences have given me a good balance of practical and academic skills which translate into my research.
