Jiangen He, Wanqi Zhang, Jessica Barfield
7 min
Abstract
As artificial agents increasingly integrate into professional environments, fundamental questions have emerged about how societal biases influence human-robot selection decisions. We conducted two comprehensive experiments (N = 1,038) examining how occupational contexts and stereotype activation shape robotic agent choices across construction, healthcare, educational, and athletic domains. Participants made selections from artificial agents that varied systematically in skin tone and anthropomorphic characteristics. Our study revealed distinct context-dependent patterns. Healthcare and educational scenarios demonstrated strong favoritism toward lighter-skinned artificial agents, while construction and athletic contexts showed greater acceptance of darker-toned alternatives. Participant race was associated with systematic differences in selection patterns across professional domains. The second experiment demonstrated that exposure to human professionals from specific racial backgrounds systematically shifted later robotic agent preferences in stereotype-consistent directions. These findings show that occupational biases and color-based discrimination transfer directly from human-human to human-robot evaluation contexts. The results highlight mechanisms through which robotic deployment may unintentionally perpetuate existing social inequalities.
Alex: So robots could end up reinforcing the same job divides we see in people...
Sam: Precisely. In the first study, over 400 people from different backgrounds each picked robots for four job situations. They used drawings of robots where everything stayed the same except skin color and how human-like the shape was—pulled from a database of real faces to make tones realistic: light, medium, brown, dark, plus plain silver and teal for comparison. The robot shapes went from basic blocks with no arms to full human outlines with bendy limbs.
Alex: So they controlled all the looks except color and humanness to pinpoint what swayed choices?
Sam: Yes. This setup let them see if more human looks made biases stronger. For analysis, they used a statistical tool that groups choices by individual to spot patterns fairly, adjusting for personal habits—a multilevel logistic regression.
Alex: Okay, so what did that reveal about how job spots changed color picks?
Sam: In hospitals, light-colored robots got picked far more at the most human-like level, while dark ones barely registered. Construction and sports flipped it, favoring darker tones. Non-skin colors like silver won overall, but skin choices split sharply by job.
Alex: Huh, so the more robot-like a person, the more our job biases kick in on color.
Sam: The second study built on that by adding human worker images before choices—half matching common stereotypes, half not—like Latino in construction but White in hospital. Picks shifted strongly toward colors echoing those humans, boosting stereotype matches substantially. It shows cues from people automatically steer robot assignments, even without intent.
Alex: But with all those grouped choices, did participant backgrounds matter in the patterns?
Sam: Participant race did interact, with groups showing varied strengths in these shifts. Asian participants consistently leaned toward lighter robots in care tasks. Black participants favored darker ones more in construction and sports. White participants showed less preference for light in hospitals but higher for dark in building work. Latino picks spread more evenly, often going for non-skin colors like teal.
Alex: Huh. So even within the same job, people's own backgrounds nudged their robot color choices differently.
Sam: The researchers ran models that account for each person's repeated choices across jobs, weighing factors like task type and race. Task context had the strongest pull—for instance, hospital scenarios raised the chances of light robot selection about four-and-a-half times compared to construction.
Alex: Okay, so jobs drove the biggest splits, but who you are fine-tunes it—like Black folks more open to dark bots across labor roles.
Sam: Yes, and those models showed good reason for grouping by person, since choices varied a lot between individuals. The stats confirmed context and primes drove most of the effect.
Alex: That could mean hospitals end up with mostly light-skinned bots interacting with patients...
Sam: The study points to that risk. When participants saw human professionals matching common job stereotypes—like White in hospitals or Black in sports—they became about six times more likely to pick robots with matching skin tones compared to mismatched cues. This held across robot designs.
Alex: But in real hospitals or sites, would managers actually do that, or is there something holding it back?
Sam: The experiments were lab-based, using drawings and scenarios rather than actual deployments, so they show tendencies but not guaranteed real-world outcomes. Notably, non-skin colors like silver and teal got picked most often overall, hinting people might avoid racialized choices by going neutral when possible.
Alex: Right, so while biases show up clearly here, folks lean toward safe, non-human colors to sidestep them.
Sam: That avoidance could be a natural check, but the priming shifts still raise concerns for equitable robot use unless designs prioritize neutral options or policies ignore appearance. The work highlights how social cues transfer to technology.
Alex: Well put, Sam. Thanks for breaking it down so clearly—it's given me a lot to think about on how our assumptions shape the future of work with robots. That's our look at this research for today. Thanks for listening to ResearchPod.