The Shrinking Imagination: When Did Human-Robot Interaction Become Synonymous with Psychology?

Cognitive science and psychology tend to study artificial tasks in artificial environments. The findings are difficult

to generalise to the messy, real world.

Donald Norman, The Design of Everyday Things

A science of behaviour is not yet practical until

it goes beyond the laboratory.

BF Skinner, Beyond Freedom and Dignity

Human-Robot Interaction (HRI) is meant to be an interdisciplinary field—one that integrates robotics, design, engineering, computer science, social sciences, and the humanities. Yet in recent years, it has increasingly taken on the structure and sensibilities of applied psychology in some research collectives.

In such research environments and labs, psychological methods are not simply influential; they are treated as foundational.

This dominance of psychology has narrowed the scope of inquiry, discouraged creativity, and contributed to a formulaic body of knowledge that often fails to reflect the real-world potential—and messiness—of how humans and robots relate.

Instead, we spend too much time worrying about whether survey scales are valid, whether an experiment had a control group, and whether a robot’s gaze behaviour was statistically significant. If your study didn’t involve a lab setting, questionnaires, or a p-value, it was often considered anecdotal, soft, or—worse—unscientific.

It didn’t take long to realise something was off, and why so much of the work being published felt… predictable.

Well-controlled, yes. But also formulaic, unimaginative, and disconnected from how robots might matter in the world.

HRI’s Shrinking Imagination

Psychological models have undoubtedly contributed useful insights into HRI, including how humans perceive trust, social presence, or emotional cues in robotic systems. However, the widespread adoption of standardised psychological instruments and methods has come at a cost.

Rather than encouraging exploration and innovation, the prevailing academic culture rewards replication, variable control, and statistical validation. Research agendas are shaped not by open-ended curiosity but by what is most easily measurable within established paradigms.

This has created a research environment that prioritises safe, publishable studies over bold, imaginative thinking.

The Tyranny of the Lab

Much research is conducted in tightly controlled lab settings. The robot’s actions were pre-scripted. Participants followed instructions. The environment was stripped of distraction.

They're not studying interaction. Rather, they're studying reactions to a stimulus—like Skinner with a shinier box.

These setups are excellent for controlling variables, but terrible for revealing the messy, emergent dynamics that happen when robots enter the real world.

People don’t interact with robots like lab rats. They interact like… people.

Yet, there was little space—or support— for exploratory and design-led inquiry. If you couldn’t plug your data into SPSS, it barely counted.

Measurement Over Meaning

One of the most significant limitations of psychology-led HRI is its preoccupation with quantifying phenomena that are qualitative in nature. Trust is reduced to Likert scales. Social presence becomes a numeric score. Anthropomorphism is measured in blink rates or voice pitch.

The richness of human experience—the hesitation before touching a robot, the discomfort of being watched, the ambivalence in laughter—does not lend itself easily to variable control or survey design. And yet, these aspects are often the most critical in shaping the success or failure of robot integration.

Optimising robots to perform well on quantifiable metrics risks missing what truly matters in interaction: meaning, context, ambiguity, and emotional resonance.

We're essentially building robots that could pass social tests, not ones that could carry social meaning. Too often, it feels like the goal is publishability, not possibility.

Human-Centred ≠ Psychology-Led

Many HRI projects claim to be “human-centred,” yet the term is often interpreted narrowly—as a synonym for user studies grounded in psychological constructs. This interpretation obscures the fact that human-centred design is a design-led imperative, not a psychological one.

Importantly, user-centred design principles originated in engineering disciplines, particularly in human-computer interaction (HCI), systems engineering, and product-design engineering. They emerged to help engineers create technologies that are usable, safe, and contextually appropriate—not generalisable to the population at large, but fit for purpose.

Where psychology often seeks abstraction and statistical confidence, design thrives on specificity, iteration, and real-world engagement. Human-centred HRI should reflect that. It should include methods like ethnography, participatory design, speculative prototyping, and field testing—all of which remain marginalised in favour of lab-based validation.

Where HRI Could Go

HRI holds immense potential to shape not only how robots function, but how they are ethically, socially, and symbolically understood in society. But as long as the field remains tethered to psychology as its methodological anchor, that potential will remain constrained.

Current orthodoxy discourages:

• Exploratory and speculative inquiry

• Interdisciplinary collaboration beyond behavioural science

• Design-led and participatory approaches

• Engagement with context, ethics, and power

Reclaiming the field’s imagination requires decentring psychology. Not discarding it—just placing it alongside other tools again, rather than on a pedestal.

Rebalancing the HRI Toolkit

The future of HRI depends on expanding what counts as legitimate knowledge. That includes welcoming:

• Design research – with its emphasis on iteration, prototyping, and storytelling.

• Ethnographic and sociological methods – that capture lived experience and context.

• Philosophical and critical thinking – to interrogate meaning, ethics, and power.

• Engineering mindsets – returning to roots where user design was a tangible imperative, not a psychological abstraction.

HRI shouldn’t just study interaction—it should shape it. That means building tools, narratives, and relationships that reflect the full diversity of human life, not just what’s measurable in the lab.

A Final Thought

A field that aspires to help humans and robots coexist must make space for the complexity, ambiguity, and imagination that coexistence demands.

If HRI is to fulfil its promise, it must expand its methodological vocabulary, broaden its intellectual lens, and reclaim its capacity to invent new ways of being.