Adding Screening Questions to Your Study

To add screening questions to a Koji study, open the study's brief, go to the Questions tab and click Add in the Screening Questions section at the top. Write a single choice or multiple choice question, mark each answer as one that qualifies or disqualifies, and publish. From then on, everyone who opens your interview link answers the screening questions before the interview starts, and anyone who doesn't qualify is politely turned away before a single interview question is asked. Koji won't let you publish until every screening question is complete, and screening questions are available on every plan for studies created from 5 September 2026.

This guide covers how screening works in Koji, how to set up qualifying rules for each answer format, what disqualified participants see, and how to write screeners that keep the right people in.

Why screen before the interview

An interview with the wrong person costs you twice: once in credits and time, and again when their answers quietly skew your findings. Qualitative studies are small by design, so a handful of off-target participants can change what looks like the main theme. That is why experienced researchers screen hard and recruit more people than they need.

Traditional survey tools handle this with skip logic and "terminate" branches, which you build by hand and test yourself. In an AI-moderated interview the stakes are higher, because each conversation is longer and richer than a survey response. Koji puts screening in front of the conversation, checks your rules before you publish, and keeps screened-out people out of your results and reports.

How screening works in Koji

When a participant opens your interview link, they move through the study in this order:

1. The welcome screen
2. Screening questions, one at a time
3. Any pre-interview form you set up
4. The choice of a text or voice interview
5. The interview itself

Because screening comes before the text or voice choice, it applies to both modes. The AI interviewer never starts a conversation with someone who didn't qualify.

Participants must pass every screening question. The first answer that disqualifies them ends the screener. Before they pass, they can step back to an earlier screening question to change an answer. Once they've passed, the screener is behind them. Someone who returns after being screened out isn't asked again.

Step by step: adding a screening question

Step 1: Open the Questions tab

Open your study and go to the brief. The brief has three tabs: Context, Questions and Branding & Settings. On the Questions tab, Screening Questions is the first section, above Opening Approach and Interview Questions.

If the study has no screeners yet, you'll see "No screening questions. Everyone who opens the link can take part."

Step 2: Click Add and write the question

Click Add. A new screening question starts as Single Choice. Screening questions are numbered S1, S2 and so on, separately from your interview questions (Q1, Q2…), so it's always clear which is which.

Screening questions can be Single Choice or Multiple Choice. Koji's other structured question types (Open Ended, Scale, Ranking and Yes / No) can't be screeners, because only choice questions can decide who qualifies without a person reading the answer. If you need a yes or no screener, write it as a single choice question with the options Yes and No.

Each question needs at least two answer options and can have up to eight.

Step 3: Mark what each answer does

Every option has a Set… dropdown. An option without a setting is outlined in amber so you can spot it.

For a single choice question, mark each answer as:

• Qualify: choosing this answer lets the participant continue
• Disqualify: choosing this answer ends the screener

For a multiple choice question, mark each answer as:

• May select: fine to pick, doesn't decide anything on its own
• Must select: the participant has to pick this answer to qualify
• Disqualify: picking this answer screens them out

On a multiple choice question, a participant is screened out if they pick any Disqualify answer or leave out any Must select answer.

Step 4: Publish

When you publish, Koji checks every screening question. If one is incomplete, publishing is blocked with a message such as "A screening question is not finished," and the question shows what's wrong. The checks catch:

• A question with no text
• A question type other than single or multiple choice
• Fewer than two answer options
• Answers with no setting
• A question where no answer qualifies, which would screen out every respondent
• A single choice question where every answer qualifies, which screens nobody out
• A multiple choice question where every answer is May select, which also screens nobody out

You can save a draft with unfinished screeners; only publishing is blocked.

Ask the study assistant instead

You don't have to build screeners by hand. Ask the study assistant something like "Add a screening question so only people who manage a team of five or more take part," and it writes the question and marks the qualifying answers. When the assistant rewrites your interview questions, your existing screening questions are kept.

What disqualified participants see

A participant who doesn't qualify sees a short, polite message thanking them for answering and explaining that the study is looking for a specific group. No interview takes place.

Screened-out visits don't use credits, because no conversation is recorded: there's no transcript, no duration and nothing to bill. Screening questions also don't count toward your interview's length estimate and never appear in report charts.

Screened-out participants in your results

On your study's results, people who didn't qualify have the status Screened out. A Screened out filter appears as soon as there's at least one, so you can see how many people your screener turned away and what they answered, up to the question that ruled them out.

You can Archive screened-out rows to tidy the list, and Restore them later.

A high screen-out rate is a signal worth reading. It might mean your link is reaching the wrong audience, or that a rule is stricter than you intended.

Which studies can use screening questions

Screening questions are available on all plans, Free included, for studies created on or after 5 September 2026. Studies created before that date don't show the section and aren't screened. To screen participants for an older study, start a new study for your next round.

Screening applies to participants who arrive through your study's interview link. If you collect responses through an embedded widget, check your setup before relying on the screener for those participants.

Writing screeners that work

Koji enforces the mechanics; the judgement is yours. A few rules of thumb:

• Screen on behaviour, not identity. "How often did you use a project management tool in the last month?" is more reliable than "Are you a project manager?"
• Don't reveal the right answer. If your question reads "Do you use Slack daily?", people who want the incentive will say yes. Offer a range of tools or frequencies and qualify only the ones you need.
• Include an "None of these" option on multiple choice screeners and mark it Disqualify.
• Keep it short. Two or three screening questions are usually enough. Every extra question adds drop-off before the interview begins.
• Don't screen too tightly. Over-strict criteria can produce a sample that is pure but unrepresentative. See Tightening Your Screener Makes the Sample Purer and the Findings Worse.

For a full methodology guide, read Screener Questions for User Research.

Example: a churn study screener

S1 (Single Choice). When did you last use the product?

• In the last week: Disqualify
• 1 to 3 months ago: Qualify
• More than 3 months ago: Qualify
• I've never used it: Disqualify

S2 (Multiple Choice). Which of these describe how you used it? Select all that apply.

• I paid for a subscription: Must select
• I used it at work: May select
• I used it for personal projects: May select
• None of these: Disqualify

Only lapsed paying customers reach the interview, where Koji's AI interviewer asks open questions about why they left and follows up on each answer.

Related Resources

• Structured Questions in AI Interviews
• Screener Questions for User Research: A Complete Guide
• Screening Research Participants Effectively
• Understanding the Research Brief
• Publishing Your Study
• Single Choice Questions in AI Interviews