Research Talk by Anocha Aribarg, University of Michigan (US)

Anocha Aribarg's MRSS session tackled a question that feels increasingly hard to ignore: why do some sociopolitical influencers manage to shape public opinion while others, working on similar issues, barely register? Together with her coauthors, she argues that persona is a big part of the story, and in particular the contrast between aspirational cues — signals of ambition, purpose, and future-oriented vision — and relatable ones, which are more about closeness and accessibility. The empirical backbone is a four-year panel of 4,396 YouTube videos from 98 influencers, analyzed with automated face and speaker recognition to build multi-modal measures across verbal, visual, and vocal signals. The results consistently point to aspiration as the stronger driver of engagement, with nonverbal cues like higher vocal pitch and direct eye contact adding meaningful lift. There are also striking gender differences: female influencers seem to benefit from non-smiling expressions and lower vocal pitch, both of which read as aspirational signals. A planned experimental phase using AI-generated influencer videos will push the causal side of the story further.

The part I found most exciting about this talk was the methodological ambition of combining large-scale panel data with computer vision, audio analysis, and eventually AI-generated stimuli in an experimental setting. It is the kind of research design that would have been genuinely difficult to run just a few years ago, and it opens up a lot of possibilities for studying communication in ways that go well beyond text. I also appreciated how carefully the authors distinguished aspiration from relatability, since these two ideas often get lumped together in more casual conversations about influencer culture. I walked out thinking about how much of persuasion online might be driven by cues we never consciously register.