In week two, I drafted a second version of the experimental protocol and reviewed literature to find theories supporting my hypotheses. I began questioning my research based on my experience in corporate recruiting, where leadership pushed to implement HireVue’s on-demand video interview software. This technology aimed to expedite the process and support DE&I efforts by allowing candidates to record their responses at their convenience, potentially increasing the candidate pool. However, many candidates found the experience cold and one-sided, leading to low completion rates and unnatural responses. Recruiters also struggled with reminders and promoting the technology to hiring managers.

This experience made me wonder if video interviewing negatively impacts candidate performance, potentially causing us to overlook qualified individuals due to the poor setup. With advances in GPT-4.0, two-way realistic conversations could soon be incorporated into virtual interviews, potentially improving the experience. However, it remains unclear if candidate performance would match that in human interviews.

I delved into the literature on AI interview agents, virtual interviews, and related theories such as Media Richness Theory, Social Presence Theory, Impression Management, technology-mediated interviews, social-emotional interaction, and anthropomorphism. Additionally, I explored online research participant platforms like MTurk and Prolific to develop my recruitment strategy and experimental procedures, focusing on differentiating the two experimental groups.