# Introduction

An **experiment** defines the parameters of a study, including the type of study and the data to be evaluated.
For example, a [MUSHRA](/experiments/audio/mushra/) study evaluating compressed audio.

## Jobs and sessions


```mermaid
  graph RL
   Job --> Experiment
   Session --> Job
   Slate --> Session
   Stimulus -- "Rating" --> Slate
```

Each experiment can have multiple **jobs**.
Jobs determine *who* is going to evaluate the data, that is, who is going to participate in the study.
For example, you may wish to run a pilot study with project members before running a larger study with
crowd-sourced raters. This can be achieved by creating two jobs for the same experiment.

When a human rater participates in a study, this creates a **session** associated with the job.
Each session consists of multiple *slates* that represent a group of evaluated *stimuli*.
The stimuli typically correspond to media files from the dataset.