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

# Long text

> The Long text question type: an open answer in the respondent's own words. Covers the one-time probe and probe lines, summary read-back, don't-know and refusal handling, recommended timing, and what is recorded.

## What it's for

A **Long text** question asks an open question and records the answer in the respondent's own words — *"What's the most important issue facing your local area at the moment?"* or *"Is there anything we could have done better?"*

The answer is captured verbatim. Nothing is matched against a list, and there's no format check or screen-out.

***

## Settings

Alongside the settings every question shares ([Question settings](/surveys/question-settings)):

| Setting                                        | What it does                                                                                                                                                                                                                                        | Default               |
| ---------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | --------------------- |
| **Probe once for more after the first answer** | After the first answer, the interviewer asks one neutral follow-up to draw out more.                                                                                                                                                                | Off                   |
| **Probe lines**                                | Your own follow-ups, one per line (up to 10, each up to 300 characters). *"One follow-up is chosen from these. Empty uses the template's own pool, which opens “Is there anything else you'd like to add about that?”"* Shown when the probe is on. | Empty (built-in pool) |
| **Read every answer back**                     | Reads back a one-line summary of the answer for a yes before recording it.                                                                                                                                                                          | **On**                |

A Long text question has no answer options, **Format** or screen-out settings.

<Tip>Probe lines are only spoken when **Probe once for more after the first answer** is on. Lines left in place while the probe is off are kept but never used.</Tip>

***

## What the respondent hears

**The ask.** The voice prompt. The respondent answers in their own words.

**The probe** (when on). Exactly once, after the first answer, the interviewer asks a neutral follow-up — *"Is there anything else you'd like to add about that?"* or one of your **Probe lines**. Whatever they add is joined to the first answer into one verbatim record. *"No, that's it"* leaves the first answer as it was.

Write probe lines that are neutral and open — *"Is there anything else about that you'd like to mention?"*, *"Anything else at all?"* — not ones that steer the answer.

**Confirmation** (on by default). The interviewer reads back a short summary covering every point made — *"So water infrastructure, and climate change — have I got that right?"*

* A "yes" records the answer.
* Any kind of "no" asks the respondent to give their answer again; they don't correct the summary piece by piece.
* What's recorded is always the respondent's own words, never the summary.

Turn **Read every answer back** off for a quicker interview when a summary adds little — for example, a short "any other comments?" at the end. See [Answer confirmation](/surveys/answer-confirmation).

**"I don't know"** gets one nudge — *"No worries — even a general sense of it would be useful."* — then the question is skipped. A hedged answer such as *"I don't really know, but I suppose the parking"* counts as an answer.

**A refusal** gets one reassuring nudge — *"That's completely fine — it's only ever reported alongside everyone else's. Would you be happy to say anything at all?"* — then the question is skipped.

Silence, "say that again", FAQs, busy and stop work as on any question — see [Question settings](/surveys/question-settings#what-happens-when-the-interviewer-asks).

### Timing

People pause while they think through an open answer. Set **Eagerness** to **low** on Long text questions so a mid-thought pause isn't taken as the end of the answer. See [Attempts and step levels](/surveys/attempts-and-step-levels).

***

## What is recorded

| Variable                                 | Type | Example                                                          | Notes                                                                                     |
| ---------------------------------------- | ---- | ---------------------------------------------------------------- | ----------------------------------------------------------------------------------------- |
| `{{survey.<key>.result}}`                | text | `The state of the roads, and not enough parking near the shops.` | The respondent's own words, including anything added after the probe. Empty when skipped. |
| `{{survey.<key>.status}}`                | text | `captured`                                                       | `captured`, `skipped`, `invalid`, `abandoned`, or empty if never reached.                 |
| `{{survey.<key>.termination_requested}}` | text | `false`                                                          | Always `false` — a Long text question never screens.                                      |

In **Results → Questions** the most common answers (up to 12) are listed with counts; in the CSV export the text is in the `answer` column, ready for coding. See [Responses and export](/surveys/responses-and-export).

***

## Next Steps

* [Question settings](/surveys/question-settings) — the shared fields.
* [Answer confirmation](/surveys/answer-confirmation) — summary read-back compared with other types.
* [Attempts and step levels](/surveys/attempts-and-step-levels) — pacing open answers.