A single evidence accumulation process informs perceptual choices and subsequent confidence reports

Imagine your brain is a busy courtroom where a jury is trying to decide if a blurry image is a cat or a dog. This paper asks a fascinating question: How does that same jury decide how sure they are about their verdict?

There are two main theories about how this works, and they disagree on two big things:

1. One Jury or Two? Does the brain use one single "jury" to make the choice and the confidence rating, or does it have a second, separate jury that only handles confidence?
2. When do they stop? Does the jury stop gathering evidence when they hit a specific "certainty threshold" (like a gavel banging), or do they just stop after a set amount of time (like a clock running out)?

The Experiment: A Game of "Blurry Dots"

The researchers asked people to play a game where they had to spot a direction of movement in a cloud of random dots (like watching a school of fish swim in a specific direction). After making their choice, the people had to say how confident they were.

The researchers then built computer models to see which theory matched the human behavior best.

The Findings: One Brain, Two Rules

1. The "When to Stop" Mystery
When looking only at the people's choices and confidence scores, the models showed that the brain acts like a threshold-based judge. It doesn't just wait for a timer to run out; it keeps gathering evidence until it feels "sure enough" to stop and make a report.

2. The "One vs. Two" Showdown
This is where it gets interesting.

• The "Two-Jury" Theory: Suggests that after you pick "cat," a different part of your brain starts a fresh investigation to figure out how confident you are.
• The "One-Jury" Theory: Suggests it's the same jury that never stops. It just keeps listening to the evidence. Once they decide "Cat," they keep listening a bit longer until they feel confident enough to say, "I'm 90% sure it's a cat."

The Behavior Test: When the researchers only looked at the game scores, the data was too messy to tell which theory was right. Both the "One-Jury" and "Two-Jury" models could explain the choices and confidence ratings almost equally well.

The Brain Scan Test: However, when they looked at the actual electrical signals in the brain (the "neural signatures"), the answer became crystal clear. The "One-Jury" model was the winner.

The brain signals showed that the evidence was being accumulated in a single, continuous stream. The most surprising part? The brain's "confidence signal" was already building up hundreds of milliseconds before the person even announced their choice. It was as if the jury was already whispering, "We're pretty sure," while they were still debating the verdict.

The Conclusion

The paper concludes that your brain doesn't need a second team to check your confidence. Instead, it uses one single process of gathering evidence.

Think of it like a thermometer:

• The mercury rises as it gathers heat (evidence).
• When it hits a certain line, it triggers the "Choice" (e.g., "It's hot").
• But the mercury keeps rising! The height it reaches after that first line is what determines the "Confidence" (e.g., "It's very hot").

The study proves that our choices and our confidence in those choices are just different stages of the same continuous process, governed by a single stream of evidence that keeps flowing even after we've made up our minds.