> ## Documentation Index > Fetch the complete documentation index at: https://help.voxdash.com/llms.txt > Use this file to discover all available pages before exploring further. # Step 5 · Screening and Targets > Review the screener questions the AI applies at the start of each call, and set distribution targets to compare what you planned with what you achieved. **Step 5 · Screening and targets** is the fifth step of the AI-CATI study setup. It has two parts: * **Screening and eligibility** — a read-only view of the screener questions and screen-out rules taken from your questionnaire. * **Distribution targets** — optional targets per answer, used to compare planned and achieved completes during and after fieldwork. > [!NOTE] > > This step never blocks you. Once a questionnaire is bound in [Step 2](https://help.voxdash.com/AI-CATI/Setup-Questionnaire.md), Step 5 counts as complete even if you add no targets. ## Screening and Eligibility This section shows what the AI interviewer applies at the start of every call, exactly as it was written in the Questionnaire Builder. It is **read-only**. ![Screening and eligibility card](https://mintcdn.com/voxdash/OFPcyQtvMysHCjUR/images/AICATI/b-step5-screening.png?fit=max\&auto=format\&n=OFPcyQtvMysHCjUR\&q=85\&s=325ccd46c3d0b1ac7c8a47adbcaa7271) You will see: | Item | What it shows | | - | - | | **Screener questions** | Every question tagged as a screener, with its number, question type and answer options. | | **Screen-out rules** | The eligibility logic. A respondent is screened out as soon as one of the expressions is true. | | **Reason code** | The code stored on the sample record when a respondent is screened out, for reporting. | | **Screen-out script** | What the AI says to a respondent who does not qualify. | To change the wording, the answer options, the eligibility logic or the screen-out script, edit the questionnaire in the Questionnaire Builder, publish a new version for AI-CATI, and re-bind it in [Step 2](https://help.voxdash.com/AI-CATI/Setup-Questionnaire.md). If the questionnaire has no screener, the card shows **No screener found** with a link back to Step 2. To learn how to mark a question as a screener, see [Tags](https://help.voxdash.com/Questionnaire-Builder/Tags.md). ## Distribution Targets Distribution targets say how many completed interviews you want for each answer of a question — for example, 50% women and 50% men, or a set number of completes per region. ![Distribution targets card](https://mintcdn.com/voxdash/OFPcyQtvMysHCjUR/images/AICATI/b-step5-targets.png?fit=max\&auto=format\&n=OFPcyQtvMysHCjUR\&q=85\&s=56dfe7f4348bdeaedf28b5a507d4754c) > [!WARNING] > > Targets are stored for **target-vs-achieved comparison only**. They are **not enforced during calling** — the AI keeps dialing and completing interviews regardless of how full a target is. To control who is dialed, use [Sub-samples](https://help.voxdash.com/AI-CATI/Sub-samples.md) or the quota features of the questionnaire. ### Add a Target **1. Select Add target** A picker lists the questions in the bound questionnaire. Each question can have only one target. **2. Choose a question** The question's answer options become the rows of the target. **3. Set the comparison base (optional)** The number the percentages are measured against. It defaults to the study's target completes (set in [Step 4](https://help.voxdash.com/AI-CATI/Setup-Sample.md)). Lower it when the question is only asked to a subset of respondents. Leave it empty to use the study default. **4. Enter a target for each row** Type the number of completes you want per answer. The **Percent of base** column updates as you type. ### Target Rows | Column | Description | | - | - | | **Label** | The row name. Every row needs a label. | | **Answer values** | The answer codes that count toward this row. | | **Target** | The number of completes you want for this row. | | **Percent of base** | The target as a share of the comparison base. | | **Achieved** | Filled in during fieldwork. It stays empty during setup. | For questions that accept numbers, use **Add range** to create a row from a minimum and maximum value (for example, ages 18–24). You can remove rows, remove a whole target, and use **Expand all** / **Collapse all** to manage long lists. ### Badges | Badge | Meaning | | - | - | | **On base** | The row targets add up exactly to the comparison base. | | **Off base** | The row targets do not add up to the comparison base. Review the rows. | | **Question changed** | The question changed in the bound questionnaire after the targets were saved. Check the rows before saving. | | **Question missing** | The question no longer exists in the bound version. Remove the target or re-bind another version. | ### Rules | Rule | Details | | - | - | | Comparison base | Must be greater than 0 and no more than the callable records in the sample. | | Row target | Must be 0 or more, and no more than the base (or the callable records). | | Sum of rows | Rows must add up **exactly** to the comparison base — for example, "The targets add up to 90 but must equal the comparison base of 100." Multi-select questions are exempt, because one respondent can count in several rows. | | Rows per target | A target needs at least one row. | ## After Launch During fieldwork, the **Achieved** column fills in and the **Composition** card on [Live Monitoring](https://help.voxdash.com/AI-CATI/Live-Monitoring.md) shows achieved versus target. After fieldwork, [Results](https://help.voxdash.com/AI-CATI/Results-and-Export.md) highlights any row that deviates from its plan by 10% or more. ## Next Step Continue to [Step 6 · Calling Rules](https://help.voxdash.com/AI-CATI/Setup-Calling-Rules.md). This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.