> ## Documentation Index
> Fetch the complete documentation index at: https://support.deepfield-ai.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Configure conditions and routing

> Control which questions participants see, where they go next, and how earlier answers are reused.

Question logic is configured from the **⋯** menu on a question card. Conditions use the same rule builder whether they control a question, section, option, jump, screen-out, or early end.

## Before you start

Save the source questions and place them before the questions that depend on them. Write the expected route for a matching and a non-matching answer.

## Show content conditionally

<Steps>
  <Step title="Open the condition builder">
    Open the question's **⋯** menu and select **Show only if…**. For an entire section, use its **Show section when** control instead.
  </Step>

  <Step title="Add the rule">
    Select **Add condition**, then choose an earlier question, an operator, and the value to match.
  </Step>

  <Step title="Combine conditions">
    Conditions use **AND** by default. Change the connector to **OR** or select **Add condition group** when the logic requires grouped conditions.
  </Step>

  <Step title="Review the summary">
    The card displays a **Show if** summary. Read it as a sentence and confirm it matches the intended participant route.
  </Step>
</Steps>

<Frame caption="Show the purchase-location question only when the earlier frequency answer is Every day.">
  <img src="https://mintcdn.com/deepfieldai/dK8nllvOe0f0oG_-/images/building-your-study/condition-builder.png?fit=max&auto=format&n=dK8nllvOe0f0oG_-&q=85&s=2857eca926ad5cb2a9eb07ae1f30b8f0" alt="Completed condition: the earlier cold-coffee frequency equals Every day" width="807" height="284" data-path="images/building-your-study/condition-builder.png" />
</Frame>

Use **is one of** or **is not one of** to match several values in one rule. Numeric questions also support comparisons such as **greater than** and **less than**. **AI checks** evaluates an open-text answer against an instruction you provide.

Place a shared condition on the section or block when every item inside it follows the same rule. Individual options can also use **Only show this option if…**.

## Jump to a later question or section

Open **⋯ → Jump to…**, select a destination later in the questionnaire, and add the condition that activates the jump. The question card shows a **Jump to … if …** summary.

Jumps only move forward. A destination inside a randomized block is not accepted because its position can differ between participant routes. Select a destination outside the block or remove that block's randomization.

Use conditional display when the requirement is to hide a set of questions without moving to a named destination.

## End the survey after a qualifying route

Use **⋯ → End survey when…** to complete the response when a condition is met at that point. This records an early completion rather than a screen-out.

Screening questions use **Screening rule** instead. A screen-out identifies a participant as outside the required audience and can send them to the configured screen-out redirect.

## Insert an earlier answer into text

Type **@** in question text, help text, or an option label. Select an earlier question or variable, then [test the substituted text](/building-your-study/answer-piping#verify-each-route).

## Reuse earlier selections as options

Use **Options source → Pipe from question** on the target question. See [Reuse earlier answers](/building-your-study/answer-piping#reuse-earlier-selections-as-options) for selected, unselected, ranked, and numeric sources.

## Randomize options, rows, blocks, or sections

| Level            | Controls and result                                                                                   |
| ---------------- | ----------------------------------------------------------------------------------------------------- |
| Question options | **Randomize Order** changes their order per participant. A display count can limit the drawn options. |
| Matrix rows      | **Randomize rows** changes row order and can draw a subset. Grouped rows remain together.             |
| Question block   | **Randomize** changes question order; **Show only N** draws a subset.                                 |
| Sections         | A section group can randomize or draw sections as a unit.                                             |

Test rules that enter or leave a randomized area. Do not use a jump destination inside a randomized block.

## Create a derived variable

Open **Variables → Add variable** in the builder. Follow [Create a derived variable](/building-your-study/derived-variables) to define cases or a calculation, test the result, and verify stored values.

## Use a formula in a rule

Select **ƒx** or **Use a formula instead** on a condition row. Insert numeric references with **@** or `{`, then build the comparison. A formula can only use numeric questions or variables available before the point where it runs.

## Test the full logic

Open **Test** from the builder view menu. Run **Flow coverage** to identify unreachable content, then use a test link to complete each important route as a participant.

Test a branch where an earlier source is never shown, as well as one where a participant explicitly skips it. Rules read never-shown sources as blank (zero for numeric evaluation); this is different from an answer still waiting to be collected. Check that validation and screening finish with the intended outcome.

## Check the result

Complete both routes through a test link and inspect their final statuses. A saved rule must name its source, operator, and value; an empty condition is not a completed example.

## Related

* [Choose a question type](/building-your-study/question-types)
* [Screen participants](/recruiting/screening-questions)
* [Fix survey problems](/building-your-study/survey-validation)
* [Test and publish a study](/recruiting/testing-publishing)