Home
/
Visual Statistical Learning (Shape Patterns)

Visual Statistical Learning (Shape Patterns)

In review

Platform: HTML · English · verified 79d ago

Description

Reimplemented from the LoL battery's Shape-Patterns mini-game (Brain & Learning Lab, UNIGE). Participants passively watch a continuous stream of abstract shapes built from a fixed set of hidden triplets — within-triplet conditional probability is 1.0, between-triplet ~0.33 — for ~10 minutes. Two test phases follow: 34 two-alternative recognition trials (pick the familiar 3-shape pattern from a foil) and 8 pattern-completion trials (pick the missing shape from a row of options). Accuracy above chance indexes implicit learning of the embedded transitional probabilities. The original 16 abstract-shape stimuli, the 4 hidden triplets, and the 42 question files are reused verbatim from the LoL battery; only the harness (PHP/jQuery) was reimplemented in jsPsych for stand-alone browser use.

Contact & Lab

Lab
Brain and Learning Lab
unige.ch
Principal investigator
Daphné Bavelier
University of Geneva
Task contact
daphne.bavelier@unige.ch

Contributors

Daphné Bavelier
University of Geneva
Supervision
Benoît Bediou
University of Geneva
Conceptualization
Supervision
Hugo Farajallah
University of Geneva
Software
Data Curation

Citation

  1. Fiser, J., & Aslin, R. N. (2002). Statistical learning of new visual feature combinations by infants. Proceedings of the National Academy of Sciences, 99(24), 15822–15826. https://doi.org/10.1073/pnas.232472899
  2. Turk-Browne, N. B., Jungé, J. A., & Scholl, B. J. (2005). The automaticity of visual statistical learning. Journal of Experimental Psychology: General, 134(4), 552–564. https://doi.org/10.1037/0096-3445.134.4.552

Credits

Adapted from the Brain and Learning Lab (UNIGE) LoL Gamer Cognitive Study battery.

Domains

Learning

Target Population

Age Categories
young adults
middle age
Clinical Status
non-clinical
Additional Notes
Originally validated with adult gamers (18+). Long passive-viewing phase — verify attention with seated, uninterrupted setup.

Languages

English

Run under

Engine
HTML / JavaScript
Format
Online + in-lab

Modalities

behavioral

Task Outputs

What it measures

Measure implicit learning of transitional probabilities in a visual stream via two-alternative recognition and pattern-completion tests.

Output formats
csv
One CSV per participant, one row per test trial (familiarization is passive).
Experimental design
Trial-based
42 trials
Within-subject: phase
Conditions
phase
recognition
completion
— Test phase. 34 recognition + 8 completion trials.
Key measures for analysis
  • Recognition accuracy (% above 25–50% chance, depending on number of options per item)
  • Completion accuracy (% above ~33% chance — 3 options per item)
  • Combined statistical-learning index
Recorded variables
VariableRoleTypeDescriptionUnit
trial_id
metadata
number
Test-trial index within phase (0-based) -
phase
independent
string
Test sub-phase
Values: recognition, completion
-
question_file
metadata
string
Source question file (e.g., questions/12.txt) -
options
metadata
string
Shuffled option list as JSON (sequence indices for recognition, shape numbers for completion) -
response
dependent
string
Selected option (1-based row label for recognition, shape filename index for completion) -
correct
dependent
boolean
Whether the response matches the embedded triplet -
rt
dependent
number
RT from question onset to response ms

Provenance

Source: nccr-seed

Fetched: 2026-04-29

Last verified: 2026-04-29 — passed

Quick Actions

Preview
Download Task
Web build
Suggest Edit

Additional Information

Created Date
2026-04-29
Last Modified
2026-04-29