> ## Documentation Index
> Fetch the complete documentation index at: https://docs.myresearchlab.app/llms.txt
> Use this file to discover all available pages before exploring further.

# Variants

> Split participants across version-level arms (A/B and factorial).

**Variants** define experimental **arms** at the version level: participants are randomly assigned across them, so you can compare, say, two framings of a stimulus or a 2×2 factorial design.

## Variants vs conditions

* [**Conditions**](/builder/conditions) branch one participant's path *within* a version based on their answers.
* **Variants** assign *different participants* to *different versions* of the study and balance assignment across them. This is the A/B (and factorial) mechanism.

## Setting up

1. Open the **Variants** section in the Builder.
2. Define one or more **factors**, each with two or more **levels** (e.g. factor *Headline framing* with levels *Neutral* / *Alarming*).
3. The study's blocks reference the active level, so each arm shows the right content. Participants are split across the resulting cells.

## Wording per arm

Many blocks let you reword their copy per variant, so an arm can show different text without duplicating the whole block.

## Analysis

Each response is tagged with the arm the participant saw, and the exported dataset includes the variant/condition columns so you can compare arms directly.

<Tip>
  Keep the number of cells small relative to your sample — every extra factor level multiplies the participants you need for adequate power.
</Tip>