Positive Affect Modulates Early Valuation and Conflict Processing in Social Decision-Making

This study demonstrates that incidental positive affect induced by music modulates social decision-making in the Ultimatum Game by accelerating reaction times, amplifying early neural conflict detection and late valuation signals, and shifting the underlying brain network topology from executive control to integrative processing, thereby enhancing the accuracy of predicting decision outcomes.

The Gist

Imagine your brain is a busy, high-stakes courtroom. Every day, you have to make decisions: Should I accept this deal? Is this offer fair? Should I say "yes" or "no"?

Usually, this courtroom is run by a strict, calculating judge (your logical brain) who weighs every penny and checks every rule. But what happens if you walk into that courtroom while listening to your favorite upbeat, happy song? Does the judge change their mind? Does the trial run faster?

This study asked exactly that question. The researchers wanted to see how happy music changes the way our brains make social decisions, specifically in a game called the Ultimatum Game.

The Setup: The "Fairness" Game

In this game, one person (the "Proposer") offers you a split of money (e.g., "I keep $9, you get $1"). You have two choices:

1. Accept: You both get the money.
2. Reject: You both get nothing (because you're too angry to take the unfair deal).

Usually, people get angry at unfair offers and reject them. But the researchers wondered: Does listening to happy music make us more forgiving, faster, or just different?

The Experiment: Two Groups, Two Sounds

They took 56 students and split them into two groups:

• The Happy Group: Listened to a cheerful, uplifting song for one minute.
• The Control Group: Listened to the sound of rain (neutral, calming, but not "happy").

Then, both groups played the money game while wearing a special hat with 64 sensors (EEG) that could read their brain activity down to the millisecond. It's like having a high-speed camera filming the thoughts inside their heads.

The Big Discoveries

Here is what happened, explained with some simple metaphors:

1. The "Speed Run" Effect

The Finding: The happy music group made their decisions much faster.
The Analogy: Imagine the Control Group is a hiker carefully checking a map, looking at every rock and tree before taking a step. The Happy Group is a runner who sees the path and just goes. The positive mood didn't just make them happy; it made their brains move faster, cutting out the hesitation.

2. The "Volume Knob" on Brain Signals

The Finding: When the happy group saw an offer, their brains lit up more intensely at two specific moments:

• Moment 1 (The "Whoa!" moment): About 220 milliseconds after seeing the offer, their brains reacted strongly to the conflict (Is this fair?).
• Moment 2 (The "Okay, let's think" moment): About 520 milliseconds later, their brains were very active in evaluating the value.
  The Analogy: Think of the Control Group's brain as a radio with the volume set to "medium." The Happy Group's brain turned the volume up. They didn't ignore the unfair offer; they actually felt the "conflict" signal louder and clearer, but then they processed it differently.

3. The "Decoupling" Mystery

The Finding: In the Control Group, a stronger brain signal meant a slower decision (they were thinking hard). In the Happy Group, stronger signals did NOT mean slower decisions.
The Analogy:

• Control Group: Like a calculator. The more complex the math, the longer it takes to print the answer.
• Happy Group: Like a seasoned chef. Even if the ingredients are complicated, the chef chops them instantly. The happy brain stopped the "thinking = slow" link. They processed the signal deeply but acted immediately.

4. The "Brain Map" Switch (The Most Important Part)

The Finding: This is the coolest part. The researchers used AI to look at how the brain connected different parts to make the decision.

• Control Group: Used the Frontal Lobe (the "CEO" of the brain). This is the part that does strict math, logic, and self-control. They were calculating: "Is this worth it? No? Reject."
• Happy Group: Switched to the Temporal Lobe (the "Storyteller" and "Socializer" part of the brain). This area handles meaning, emotions, and social connection.
  The Analogy:
• The Control Group was acting like a Lawyer: "The contract is unfair. I reject it based on the rules."
• The Happy Group was acting like a Friend: "Hey, it's not perfect, but let's just go with it and keep the vibe good."

The happy music didn't just make them "nicer"; it literally rewired the path the decision took. Instead of a long, logical debate in the front of the brain, the decision flowed through the emotional and social centers, making the "Accept" choice feel more natural and integrated.

The Takeaway

This study shows that happiness isn't just a feeling; it's a tool.

When you are in a good mood (like listening to happy music), your brain doesn't just ignore problems. Instead, it changes its strategy. It stops trying to be a strict, calculating robot and starts acting like a flexible, social human. It processes information faster, holds onto the decision more stably, and relies on social intuition rather than cold calculation.

So, the next time you have to make a tough social decision, maybe put on some upbeat music. It might just help your brain switch from "Lawyer Mode" to "Friend Mode," helping you make a choice that feels right, fast, and easy.

Technical Summary

1. Problem Statement

Social decision-making is not a purely rational process but relies on dynamic interactions between cognition and affect. While previous research has established that incidental positive emotions (unrelated to the decision itself) can bias economic choices (e.g., increasing acceptance rates in the Ultimatum Game), the neural mechanisms underlying this modulation remain unclear. Specifically, it is unknown how positive affect alters the millisecond-scale temporal dynamics of decision formation, the neural separability of decision states (Accept vs. Reject), and the functional network topology that predicts behavioral outcomes. This study aims to bridge the gap between behavioral effects and the underlying neuro-computational architecture using high-temporal-resolution EEG and advanced machine learning techniques.

2. Methodology

Experimental Design & Participants

• Participants: 56 healthy, right-handed undergraduate students (28 per group).
• Groups:
  • Happy Music Group: Induced positive affect by listening to a validated happy music excerpt (CAMS No. HAPPY01) for 1 minute.
  • Active Control Group: Induced neutral affect by listening to rain sounds.
• Task: The Ultimatum Game (UG). Participants acted as "Responders" facing offers from a computer-generated "Proposer."
  • Stimuli: 200 trials total, including Fair (5:5), Moderately Unfair (6:4, 7:3), and Extremely Unfair (8:2, 9:1) splits.
  • Procedure: Sequential design where mood induction preceded the UG task to prevent dual-task interference. Mood was verified via self-report (Likert scale) before and after induction.
• Data Acquisition: 64-channel EEG recorded at 1000 Hz.

Data Processing & Analysis Pipeline

The study employed a multi-modal analytical approach:

1. Preprocessing: Visual inspection, bad channel interpolation, bandpass filtering (0.5–20 Hz), notch filtering (45–55 Hz), and artifact removal (EOG/EMG). Re-referencing using REST.
2. Univariate ERP Analysis:
   • Averaged waveforms from frontal electrodes (Fz, FC1, FCz, FC2, Cz).
   • Cluster-based permutation tests to identify significant amplitude differences between groups in specific time windows.
   • Neuro-behavioral correlation: Pearson correlation between ERP amplitudes and Reaction Times (RTs).
3. Multivariate Pattern Analysis (MVPA):
   • Classifier: Linear Support Vector Machine (SVM) trained to discriminate "Accept" vs. "Reject" trials based on single-trial spatial topographies at each time point.
   • Temporal Generalization: A classifier trained at time $t$ was tested on all other time points $t'$ to assess the temporal stability of neural representations (distinguishing transient vs. sustained attractor states).
4. Functional Connectivity & Machine Learning Prediction:
   • Connectivity Metric: Phase Locking Value (PLV) calculated for the early decision window (0–300 ms) across 9 Regions of Interest (ROIs).
   • Prediction Model: Support Vector Regression (SVR) used to predict individual acceptance rates based on PLV network features.
   • Feature Importance: Random Forest algorithm used to identify which network edges (connections) drove the prediction accuracy.

3. Key Results

Behavioral Findings

• Reaction Time (RT): The happy music group exhibited significantly faster RTs compared to the control group ($p < 0.05$).
• ERP–RT Correlation:
  • Control Group: Significant negative correlation between early ERP amplitude (220–280 ms) and RTs (larger neural response $\rightarrow$ faster decision).
  • Happy Music Group: This correlation was decoupled (no significant relationship), suggesting a shift in how neural signals translate to behavior.

Neural Dynamics (ERP & MVPA)

• ERP Amplitudes: The happy music group showed significantly larger amplitudes in two distinct windows compared to controls:
  • Early Conflict Detection (220–280 ms): Likely corresponding to the Medial Frontal Negativity (MFN/FRN).
  • Late Valuation (520–560 ms): Likely corresponding to the Late Positive Potential (LPP).
• MVPA Decoding:
  • Both groups could decode "Accept" vs. "Reject" states above chance (50%) in the 200–300 ms window.
  • Happy Music Group: Showed significantly higher decoding accuracy (230–440 ms) and enhanced temporal generalization. This indicates that positive affect sharpens the neural code and stabilizes the decision state over time (an "attractor state").

Network Topology & Prediction

• SVR Prediction Accuracy:
  • Happy Music: High accuracy in predicting acceptance rates ($R = 0.60$, $R^2 = 0.35$).
  • Control: Lower accuracy ($R = 0.41$, $R^2 = 0.14$).
• Topological Shift (Feature Weights):
  • Control Group: Predictions relied on Frontal–Central connections, implicating executive control and conflict resolution networks.
  • Happy Music Group: Predictions relied on Central–Temporal connections, implicating integrative processing and social-semantic integration.

4. Key Contributions

1. Millisecond-Scale Mechanism: Provides evidence that incidental positive affect intervenes at the very onset of decision-making (220 ms), modulating both early conflict detection and late valuation simultaneously.
2. Decoupling of Neural-Behavioral Links: Demonstrates that positive affect breaks the direct coupling between early neural conflict signals and reaction speed, suggesting a shift from a sequential accumulation model to a heuristic/automatic processing mode.
3. Enhanced Representational Stability: Using Temporal Generalization, the study shows that positive affect creates more stable, durable neural representations of decision states, reducing neural noise and facilitating faster decisions.
4. Network Reconfiguration: Identifies a fundamental shift in the computational strategy of decision-making:
   • Neutral State: Relies on resource-heavy Frontal-Executive Control (calculative, conflict resolution).
   • Positive State: Shifts to Central-Temporal Integration (heuristic, social-semantic integration).

5. Significance

This study offers a dynamic network-based account of the affect-cognition interaction. It challenges the view that positive emotion merely "speeds up" processing; instead, it fundamentally reconfigures the neural architecture of social decision-making.

• Theoretical Impact: Supports the Affect Infusion Model and Broaden-and-Build Theory by showing that positive mood promotes a heuristic, integrative processing style (Temporal lobes) over a critical, conflict-monitoring style (Frontal lobes).
• Practical Implications: The findings suggest that incidental positive stimuli (like music) can optimize social cooperation by engaging efficient neural pathways, potentially reducing the cognitive load required for fair decision-making.
• Methodological Advance: The integration of MVPA, temporal generalization, and SVR-based network prediction provides a robust framework for decoding complex social behaviors from EEG data, moving beyond simple amplitude comparisons to understanding the topological organization of decision networks.

Limitations: The study used a single music stimulus, a homogeneous young adult sample, and EEG (limited spatial resolution). Future work should explore diverse emotional tones, populations, and combine EEG with fMRI for better source localization.