Specifications
Behaverse Data Model (BDM) is a dataset structure designed to capture information relevant to the study of human behavior across a wide range of cognitive tests and questionnaires. It defines a folder structure and a set of tidy tables for organizing data from diverse behavioral experiments.
General
Part of the BDM is common to all datasets, regardless of the specific type of activity:
- Folder structure
- A BDM dataset is a folder that contains data files and all relevant information needed to make sense of that data.
- Dataset cards
- Contain metadata and information about a dataset. Dataset cards also include a codebook to describe the semantics, structure, and format of the data variables.
- Studyflows
- Describe the study protocol and the sequence of activities that subjects are exposed to.
- Instructions
- Provide additional guidance and help about tasks presented to subjects in a study.
- Questionnaires
- Describe the questions presented to subjects and the responses they provide.
Specific
Some particular structures are unique to the type of data or activity. There are three levels of data:
- Level 1: Events
- Raw data collected during the experiment. This includes both discrete events (e.g., stimuli, responses) and continuous streams sampled at regular intervals (e.g., physiological signals). This level is the most detailed and is typically sufficient to reproduce the original experiment.
- Level 2: Trials
- BDM considers trials as the core components of cognitive experiments. These are task-specific aggregates, derived from raw events, and typically designed to capture the main features of the scientific experiment.
- Level 3: Models
- Summary statistics, scores, or computational models that are derived from trials or events. They address research questions or are used for downstream analyses.