
Seeing Past the Generalization Mirage in Human-Algorithm Behavior
Status
Under Submission
Category
Participatory AI
This paper challenges prevailing assumptions in the science of human-algorithm interaction, arguing that much of the existing research fails to generalize across populations. That failure has real consequences for the AI systems built on its findings. The paper advances a framework for more rigorous, representative research that can serve as a foundation for AI systems that work for everyone, not just the populations most studied.
Fund research that
answers to communites.
CLAIR's research is only possible with sustained support. Every dollar raised goes directly toward studies that produce findings communities can use to hold AI systems accountable. If that work matters to you, here is how to support it.
Support CLAIR
→
Navigation
Supported by UC Riverside and the School of Education
Follow us

© 2026 CLAIR. All rights reserved.


Seeing Past the Generalization Mirage in Human-Algorithm Behavior
Status
Under Submission
Category
Participatory AI
This paper challenges prevailing assumptions in the science of human-algorithm interaction, arguing that much of the existing research fails to generalize across populations. That failure has real consequences for the AI systems built on its findings. The paper advances a framework for more rigorous, representative research that can serve as a foundation for AI systems that work for everyone, not just the populations most studied.
Fund research that
answers to communites.
CLAIR's research is only possible with sustained support. Every dollar raised goes directly toward studies that produce findings communities can use to hold AI systems accountable. If that work matters to you, here is how to support it.
Support CLAIR’s Research
→
Navigation
Supported by UC Riverside and the School of Education
Follow us

© 2026 CLAIR. All rights reserved.

Research
News & updates
About us
Visit AI4
Support CLAIR’s Research
→

Seeing Past the Generalization Mirage in Human-Algorithm Behavior
Status
Under Submission
Category
Participatory AI
This paper challenges prevailing assumptions in the science of human-algorithm interaction, arguing that much of the existing research fails to generalize across populations. That failure has real consequences for the AI systems built on its findings. The paper advances a framework for more rigorous, representative research that can serve as a foundation for AI systems that work for everyone, not just the populations most studied.
Fund research that
answers to communites.
CLAIR's research is only possible with sustained support. Every dollar raised goes directly toward studies that produce findings communities can use to hold AI systems accountable. If that work matters to you, here is how to support it.
Support CLAIR’s Research
→
Navigation
Supported by UC Riverside and the School of Education
Follow us
© 2026 CLAIR. All rights reserved.
