# Habit Task — Version 1

## Overview

First pilot of the habit formation experiment testing whether social conformity-based learning produces stronger habits than individual reward-based learning. Tests whether subjects continue choosing originally-rewarded shapes after a silent reward reversal.

## Design

- **Subjects:** n=24
- **Structure:** Within-subjects, alternating social and individual blocks
- **Trial count:** 144 total trials, 12 blocks × 12 trials per block
- **Phases:** Training phase (blocks 1-6) followed by test phase (blocks 7-12)
- **Reversal:** Silent reward reversal at the training-to-test transition. No instructions or notification to subjects that contingencies have changed.

## Context Manipulation

- **Background colors** used to dissociate social vs. individual contexts:
  - Orange backgrounds for social blocks
  - Yellow backgrounds for individual blocks
- Color served as a contextual cue tied to S-R associations

## Probe Trials

- Probe trials interspersed within each block
- Probes presented forced choices between an originally-high-reward shape and an originally-low-reward shape
- **No social information shown during probes** — even in social blocks, probes had no avatars or other player choices
- No reward feedback given on probes (pure preference measure)

## Key Dependent Variable

P(choose originally-high shape) on probe trials:
- During training: indexes learning of the reward structure
- During test (after silent reversal): indexes habit (persistence of S-R association despite reward change)

## Hypothesis

Social blocks produce stronger S-R habits than individual blocks because imitation-based learning attaches "do what they did" to the available shapes without requiring slow reward integration. After silent reversal, social-context habits should persist longer than individual-context habits.