BIDS Stats Models
=================

*A specification for neuroimaging statistical models.*

*BIDS Stats Models* describe how to fit *statistical models for neuroimaging data* 
using *a machine-readable JSON document*.  
Its purpose is prescriptive (as in a recipe), 
rather than descriptive (as in summary of already fit models). 

A *BIDS Stats Models* — in combination with a preprocessed BIDS dataset — 
is sufficient to execute statistical models, 
with minimal configuration and intervention required from the user.

The **mission** of *BIDS Stats Models* is to promote scientific best practice 
by making neuroimaging models *transparent and reproducible*, 
as well as encouraging the development of *standardized and automated* statistical analysis pipelines.

::::{grid} 1 1 1 2
:gutter: 3 3 3 3

:::{grid-item-card}
:shadow: md
:class-header: bg-light
**📖 Learn more**
^^^
**[](motivation.md)**: Why do we need *BIDS Stats Models*?

**[](walkthrough-1.md)**: Step-by-step introduction

**[Specification](reference.md)**: Dive into the technical reference
:::

:::{grid-item-card}
:shadow: md
:class-header: bg-light
**🚀 Get started**
^^^
**[](model-zoo.md)**: View example models on public datasets, and modify them to your own data

**[Default model](default_model.md)**: How to create a simple model
:::

::::

## Acknowledments

BIDS Stats Models is supported by by NIH award R01MH109682, 
and is part of the [Brain Imaging Data Standard (BIDS)](https://bids.neuroimaging.io/index.html) community.
