Video-call glitches trigger uncanniness and harm consequential life outcomes
Summary
This Nature paper (Brucks, Rifkin & Johnson) shows that common video-call glitches — freezes, audio drops and timing mismatches — make people feel a sense of uncanniness, and that feeling reduces trust and willingness to engage in consequential real-world decisions. Across multiple experiments (including a main study, a conceptual replication and multi-wave follow-ups; sample sizes range into the thousands), glitches made interlocutors like each other less, feel less heard and perceive less shared reality. The effect was mediated by reported uncanniness and translated into lower interest in working with a health professional or financial advisor. Small behavioural fixes (a joke) softened the hit a little; explicit acknowledgements of glitches sometimes made things worse. The authors note implications for telehealth, remote courts, hiring, education and other domains where video remains common. Data and analysis code are available on ResearchBox (links in the paper).
Key Points
- Technical glitches on video calls reliably increase perceived uncanniness and reduce social connection (liking, feeling heard, shared reality).
- Uncanniness mediates the drop in willingness to work with professionals — e.g. 16 percentage points fewer people opted to meet a health professional after a glitchy call.
- Effects replicate across contexts: healthcare, financial advice and broader conversational samples (N across studies ranges from hundreds to >1,600).
- Small social repairs matter: a light joke after a glitch slightly improves outcomes; simply acknowledging the glitch can sometimes make impressions worse.
- Authors provide a “Glitchionary” and show glitches are common in high-stakes domains (telehealth, virtual courts, interviews), raising equity and policy concerns where remote encounters affect life outcomes.
- Data (except sensitive parole transcripts) and all custom code are publicly archived on ResearchBox for transparency and replication.
Content summary
The research combines lab experiments, conceptual replications and observational analyses. Participants watched or took part in video interactions that either ran smoothly or included brief, realistic glitches. Measured outcomes included liking, perceived shared reality, feeling heard and willingness to continue or choose the partner for consequential tasks (e.g. meeting a provider, hiring, financial advice).
Statistical results show consistent, small-to-medium effect sizes: glitches reduced liking and shared reality (Cohen’s d ≈ -0.2 to -0.5 across outcomes) and produced substantial drops in behavioural choices (for example, a 16 percentage-point reduction in willingness to meet a health professional). Mediation analyses indicate that the elevated sense of uncanniness explains much of the behavioural change. The authors also tested simple conversational repairs and found jokes can attenuate harm while acknowledgements often exacerbate it.
Supplementary materials include extended figures and tables (e.g. glitch subtypes, Tele-sales attitudes, Glitchionary) and extended data on video-call prevalence across sectors. Sensitive materials (parole and interview transcripts) are withheld for privacy; other data and code are available via ResearchBox links in the paper.
Context and relevance
Video calling remains widespread in healthcare, courts, recruitment and education. This paper is important because it links everyday technical problems to measurable social and consequential harms — not only momentary annoyance but lower trust and worse real-world choices. The findings intersect with concerns about digital inequality (unequal connection quality), procedural fairness (virtual courts, parole boards), telehealth uptake and remote hiring bias. Designers of teleconferencing platforms, practitioners using video for high-stakes decisions, and policymakers should note that small technical disruptions can bias outcomes and that simple interaction strategies can shift, but not fully remove, that bias.
Why should I read this?
If you run telehealth, manage remote hiring, sit on virtual courts or build video platforms — this is one to skim. The paper neatly explains how tiny tech hiccups make people feel ‘weird’ about the person on the other end, and that weirdness changes real choices. The researchers also point to quick social fixes (and pitfalls) you can use straightaway. We’ve saved you the trouble of wading through the experiments — read this if you care about fairness and outcomes when decisions move online.