Cambridge Gambling Task
| Field | Value |
|---|
| Name | Cambridge Gambling Task |
| Version | v0.2.2-dev |
| URL / Repository | https://github.com/TaskBeacon/T000029-cambridge-gambling |
| Short Description | Explicit-risk decision task with red/blue probability judgment and proportional betting. |
| Created By | TaskBeacon |
| Date Updated | 2026-03-19 |
| PsyFlow Version | 0.1.9 |
| PsychoPy Version | 2025.1.1 |
| Modality | Behavior |
| Language | Chinese |
| Voice Name | zh-CN-YunyangNeural (voice disabled by default) |
1. Task Overview
This implementation follows the Cambridge Gambling Task (CGT) structure: participants view explicit red/blue box ratios (10 boxes total), choose the color most likely to hide a token, then choose a betting percentage of current points.
Unlike MID-style templates, this task has no cue-target reaction stage. The core logic is probability judgment + bet selection + point outcome update.
2. Task Flow
Block-Level Flow
| Step | Description |
|---|
| 1. Block context | Controller sets block-level bet order (ascending or descending) and tracks points. |
| 2. Trial execution | Each trial samples a box ratio, majority color side, and token outcome probability. |
| 3. Block summary | Reports score, quality-of-decision, win rate, mean bet, RTs, timeout counts, delay-aversion proxy. |
| 4. Task summary | Final metrics shown with ascending vs descending bet comparison. |
Trial-Level Flow
| Step | Description |
|---|
| Fixation | Short jittered fixation (+) from timing.fixation_duration. |
| Color choice | Show 10 red/blue boxes + ratio text; participant chooses F=红 or J=蓝. |
| Bet choice | Show five bet options (5/25/50/75/95) ordered by block context; participant selects with 1-5. |
| Feedback | Reveal token color, applied stake, point delta, and updated score. |
| ITI | Short jittered fixation before next trial from timing.iti_duration. |
Controller Logic
| Component | Description |
|---|
| Ratio sampler | Samples one pair from [[9,1],[8,2],[7,3],[6,4]] each trial. |
| Outcome sampler | Token color sampled by visible probability (P(red)=red_boxes/10). |
| Bet order | ascending block uses [5,25,50,75,95]; descending reverses this order. |
| Score update | bet_amount = round(points_before * bet_percent / 100); win adds, loss subtracts. |
| Timeout policy | Color timeout -> no bet and no score change; bet timeout -> auto-select last displayed percentage. |
Runtime Context Phases
| Phase Label | Meaning |
|---|
fixation | Inter-stage fixation. |
color_choice | Explicit probability judgment stage. |
bet_choice | Percentage betting stage. |
feedback | Outcome/score update display. |
iti | Inter-trial interval fixation. |
3. Configuration Summary
a. Subject Info
| Field | Meaning |
|---|
subject_id | 3-digit participant identifier. |
b. Window Settings
| Parameter | Value |
|---|
size | [1280, 720] |
units | pix |
screen | 0 |
bg_color | black |
fullscreen | false |
monitor_width_cm | 35.5 |
monitor_distance_cm | 60 |
c. Stimuli
| Name | Type | Description |
|---|
trial_prompt, score_text, ratio_text, color_key_hint | text | Color-choice stage envelope with explicit ratio and key mapping. |
box_token_template | rect | Ten red/blue boxes generated per trial from sampled ratio and side assignment. |
bet_prompt, bet_key_hint | text | Bet-choice instruction and key legend. |
bet_option_template | text | Five on-screen bet options mapped to keys 1-5 in block-specific order. |
feedback_outcome, feedback_auto_bet, feedback_color_timeout | text | Outcome and timeout-specific feedback messages. |
fixation, block_break, good_bye, instruction_text | text | Shared trial separators and task envelope screens. |
d. Timing
| Phase | Duration |
|---|
| fixation | jittered (0.3-0.6 s human) |
| color choice | 3.0 s deadline |
| bet choice | 3.5 s deadline |
| feedback | 1.0 s |
| ITI | jittered (0.3-0.6 s human) |
e. Triggers
| Trigger | Code | Description |
|---|
choice_onset | 30 | Color-choice screen onset |
choice_red / choice_blue | 31 / 32 | Color response events |
color_timeout | 33 | No color response |
bet_onset | 40 | Bet-choice screen onset |
bet_key_1..5 | 41..45 | Bet key responses |
bet_timeout | 46 | No bet response |
feedback_onset | 50 | Outcome screen onset |
4. Methods (for academic publication)
Participants completed an explicit-risk gambling paradigm in which probabilities were represented by visible red/blue box ratios (total = 10). On each trial, participants first selected the color judged most likely to hide a token, then selected a stake as a percentage of current points.
The task manipulated bet-order context (ascending vs descending) at block level and recorded key CGT outputs: majority-color choice quality, betting magnitude, outcome sensitivity, and response latency. Score updates were proportional to selected stake and outcome congruence.
All phases were instrumented with phase-level triggers and trial context metadata, enabling synchronized behavioral and acquisition-aligned analyses across human, QA, and simulation modes.
