Framework

The Peekbank data framework consists of three components:

1. processing raw experimental datasets
2. publishing them as a versioned release
3. providing an interface to the data

The code used to process raw datasets is hosted in peekbank-data-import, which holds one import script per dataset, converting each lab’s raw eye-tracking data into the common schema. The conversion and validation helpers those scripts rely on ship as part of peekbankr.

Each release is published to Redivis: the tables themselves in the peekbank dataset, and the raw data, processed data and per-dataset READMEs in peekbank_files. Every Redivis version holds one named Peekbank release, such as 2026.1, so an analysis can be pinned to the data it was run against.

The peekbankr library (R) provides an application programming interface, or API, that offers high-level abstractions for accessing that data. Other languages can read the tables from Redivis directly.

The interactive analyses provide a tool for exploring the data in the browser.