Challenging Assumptions and Exploring New Applications of Social, Cognitive, and Evolutionary Theories of Voice Perception

Abstract

Voice perception is an integral component to social connection and communication. Using the sound of a voice we infer information about a speaker’s physical, psychological, and emotional characteristics. These impressions that are formed have the potential to influence behavioural responses to others. This thesis examines some of the fundamental assumptions of voice perception by replicating and extending their findings. In chapter 2, the assumption that exposure to voices alters how attractive voices are was tested. We did not find evidence that increased exposure to high- or low- pitched voices affected attractiveness judgements. Given that exposure to voices did not alter their perceived attractiveness, we were curious to explore if attractiveness judgements were part of first impressions people formed from voices. In chapter 3, we explored what people consciously thought about when listening to voices. We then used machine learning to organize and analyse free form descriptions of participant impressions of voices. A diverse set of topics were used when talking about voices including gender, accent, and social traits. We also confirmed that valence, dominance, and attractiveness were all important social dimensions even when participants were not prompted by researchers to evaluate traits on those domains. We followed these results by testing if the same model of dominance, trust, attractiveness, and competence applied in a practical setting. We had participants judge the voices of doctors and nurses. Low-pitched female voices were perceived as more competent sounding than male voices when they were labelled as belonging to doctors. Low-pitched voices were judged as more dominant regardless of voice sex and profession and high-pitched female voices and low-pitched male voices were judged as most attractive regardless of profession. We replicated previous findings for attractiveness and dominance perceptions and extended the work by applying it to a novel context. Our findings challenge and expand on existing assumptions of voice perception.