ABSTRACT
To combat the spread of the COVID-19 virus, countries enforced quarantines, physical and social restrictions on people. These restrictions left many feeling isolated and lonely due to prolonged quarantines and lockdowns. This raises questions about using robots as social support to alleviate these symptoms, while still complying with restrictions and regulations. Since acceptance of social robots as companions has traditionally been low, an event like COVID-19 could change acceptance of robots as social companions as loneliness can influence the likelihood of anthropomorphizing nonhuman agents. Here, we aimed to see if loneliness, due to COVID-19 restrictions, influence the Uncanny Valley pattern that prior work has shown. As such, participants saw robot images that varied in physical human-likeness and were asked to evaluate them regarding trustworthiness, mind perception and likability. The measurements were obtained once before COVID-19 (in 2016) and once at the peak of the pandemic in September 2020. Results show that ratings of mind perception and likability were significantly impacted by the pandemic, with less pronounced UV patterns for those who experienced the COVID-19 pandemic. However, no differences in the UV pattern was observed on trust. Post-hoc analyses also illustrated that people were more likely to judge machinelike robots negatively, which could be due to increased loneliness/anxiety. These data suggest that loneliness attenuates UV patterns that are observed in "Uncanny” robots and that people have more favorable attitudes towards humanlike robots when feeling lonely, which provides important considerations for the use of humanlike robots as social companions. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.