Introduction
Social cognition research aims to understand how organisms process and respond to the presence, intentions, and actions of others. While real-world social interactions are dynamic, reciprocal, and face-to-face, traditional laboratory settings often rely on static or unidirectional stimuli, such as photos or pre-recorded videos. This creates a "space dilemma" where participants must split their attention between task-related computer screens and their social partners.
To bridge this gap, we present the Dyadic Interaction Platform (DIP), a novel experimental setup designed to facilitate real-time, face-to-face social interactions in a shared workspace. By merging the workspace and the social partner into a single line of sight using transparent, touch-sensitive displays, the DIP allows for the study of complex social behaviors with high experimental control.
The Dyadic Interaction Platform (DIP)
The core of the DIP is a transparent bidirectional display positioned between two agents. The platform has been implemented using two primary technologies:
- OLED-based: Utilizing transparent Organic LED screens to provide high-contrast stimuli, though characterized by a luminosity asymmetry between the two sides.
- Projection-based: Using transparent film and ultra-short-throw projectors to ensure a symmetric visual experience for both participants.
The DIP serves as a multimodal hub, integrating behavioral recordings (touchscreens, joysticks), hyperscanning (synchronized EEG/MEG), motion tracking (3D hand and eye tracking), and physiological measures (heart rate, skin conductance).
Multi-level Social Interactions
The DIP is uniquely suited to capture social interactions across multiple temporal and cognitive scales, ranging from split-second motor adjustments to long-term strategic shifts.
As illustrated in the figure above, the platform enables the study of:
- Immediate Interactions: Real-time adjustments in actions based on the partner's movements.
- Across-trial Interactions: Strategic choices and the development of coordination strategies over multiple trials.
- Long-term Interactions: The emergence of partner-specific behaviors and social learning across multiple sessions.
Experimental Validation and Insights
The platform's effectiveness was validated through several paradigms, including transparent economic games, competitive foraging, and social learning tasks involving humans, children, and non-human primates.
The "Transparency" Effect
A key finding is that real-time visibility of a partner’s actions fundamentally alters social strategy. In coordination games, macaques—who typically struggle with abstract coordination—successfully coordinated with human partners by monitoring their reach and adjusting their own actions mid-flight. This "transparency" removes the cognitive load of abstraction, making social tasks more intuitive and salient for both primates and young children.
Pre-empting Intentions
In adult-child interactions, the platform revealed that children do not merely follow a partner's gaze; they use the transparent setup to "pre-empt" choices, fixating on targets before the partner even initiates a touch. This demonstrates the platform's ability to capture subtle, anticipatory social cues that are often lost in non-transparent setups.
Limitations and Future Directions
While the DIP significantly advances the study of live social interactions, it faces certain constraints. The vertical screen orientation can lead to physical fatigue ("gorilla arm"), and the overlap of social cues and task stimuli presents challenges for dissociating neural signals in hyperscanning studies. Future iterations may incorporate Augmented Reality (AR) to move beyond 2D constraints and allow for the manipulation of 3D digital objects in the shared space.
Despite these challenges, the DIP provides a powerful framework for integrating economic game theory with time-continuous sensorimotor decision-making, offering a more ecologically valid approach to social neuroscience.
Summary updated