Screening question (qualifying question)

Imagine this: you're running a study among electric vehicle owners. You post the survey link on social media, send it out to your customer base — and a week later you've collected 800 responses. A win? Not quite.

On closer inspection, it turns out that 60% of respondents have only ever seen an electric vehicle in pictures. Their opinion about charging infrastructure is, to put it mildly, unreliable — but separating it from the opinion of actual owners after the fact is impossible. All the feedback you collected is compromised. A single screening question at the start of the survey could have prevented this.

What is a screening question

A screening question (qualifying question) is a filtering question at the very start of a survey that determines whether a respondent matches the criteria of the study's target audience. Those who don't fit are politely "shown the door" — redirected to a final thank-you screen without ever seeing the main questions.

In essence, screening is the bouncer at the club door. It doesn't judge people as good or bad — it checks whether the guest matches the format of the party. If you're hosting a jazz concert and someone shows up for a techno rave, they're simply in the wrong place. Likewise: a survey about B2B sales has no use for a respondent who works in retail. Their answers aren't wrong — they're just about something else.

Why filtering at the entrance matters

It might seem that the more responses, the better. In practice it's the opposite: off-target responses aren't merely useless, they're actively harmful.

Data dilution. When 300 "random passers-by" hide among 500 target responses, the averages, percentage distributions, and correlations all shift. You make decisions based on a picture where a third of the pixels are noise. And that noise looks like signal: the numbers are there, the charts get drawn, the conclusions suggest themselves. They're just wrong.

Wasted budget. If you pay for each response (for example, when working with a respondent panel), every off-target participant is a direct loss. In commercial research that pays per completion, the cost of a single incorrect response can run from a few cents to several dollars. With several hundred such "misses," it adds up to a noticeable sum.

The illusion of a sufficient sample. You see that you've collected 1,000 responses and you stop collecting. But if 400 of them come from people outside your target group, the actual size of your sample is 600. That may not be enough for the statistical precision you need — and you won't even know it until you start the analysis.

Screening doesn't reduce the number of useful responses — it removes the junk ones. The remaining sample is smaller in volume but significantly higher in value.

The types of screening questions

The format depends on the criterion you need to filter by. Here are the most typical scenarios — with example wordings.

Filtering by experience

The most common case. You're interested only in those who have genuinely encountered the subject of the study.

Example. A company is researching satisfaction with delivery. Screening question: "Have you ordered delivery of goods from online stores in the past 3 months?" Options: Yes / No. If the answer is "No," the respondent is sent to the completion screen.

The time frame ("in the past 3 months") is essential here. Someone who ordered a delivery two years ago remembers their experience too vaguely — their answers will be more of a reconstruction than a description of reality.

Filtering by demographics

Age, gender, region of residence, marital status — standard criteria for segmented research.

Example. An HR department is running a survey among middle managers. Screening question: "What is your current position?" Options: Specialist / Senior specialist / Department head / Division director / Top management / Other. If "Specialist" or "Senior specialist" is selected, the survey ends.

Filtering by segment membership

When you need to reach a narrow group: users of a specific product, customers on a particular plan, residents of a particular city.

Example. A SaaS company wants to understand why free-plan users don't upgrade to a paid plan. Screening question: "Which pricing plan of our service do you use?" Options: Free / Standard / Pro / I don't use your service. The target group is "Free" only; everyone else completes the survey.

Filtering by competence

Sometimes what matters isn't what a person does but how well they understand the topic. This appears in expert surveys and professional research.

Example. A study of opinions on cybersecurity among IT specialists. Screening question: "How would you rate your level of knowledge in information security?" Options: No knowledge / Basic knowledge / Confident level / Expert. If the answer is "No knowledge," the survey ends.

How to write a good screening question: five principles

1. Put the screener first. The filtering question should be the very start of the survey — before any substantive questions. If a respondent gets halfway through the survey and only then learns that they "don't fit," they'll feel annoyed and you'll lose their goodwill. On top of that, an off-target participant has already spent time — both yours and theirs.

2. Don't hint at the "right" answer. This is the key mistake that nullifies a screener. If the question is preceded by "Survey for cat owners," a respondent who has no cat but is bored or wants a reward will easily pick the needed option. The wording must be neutral, with no hints about which answer "lets you through."

3. Use specific, verifiable criteria. The question "Are you interested in healthy eating?" is a poor screener. Almost everyone will say "yes," because no one wants to look like a person who's indifferent to their health. Better: "How many times in the past week did you cook a meal at home from fresh ingredients?" — that's a concrete action that's harder to embellish.

4. Offer several options rather than a binary choice. A question with "Yes / No" options is easy to "guess." A list of five or six options, of which only one or two are the target, significantly lowers the chance of a random match. The respondent doesn't know which option "opens the door" — and answers more honestly.

5. Provide a polite ending. A person who doesn't pass the screener shouldn't see a blank screen or a "You don't qualify" message. The correct ending is a thank-you screen: "Thank you for your time! Unfortunately, this survey is intended for a different audience. We appreciate your willingness to help." This preserves trust — perhaps this person will be a fit for the next study.

Common pitfalls

Too strict a filter. If you put three screening questions in a row and each one screens out 50% of the audience, only 12.5% of those who arrived will reach the main survey. Formally that's an ideal sample — but in practice you may spend months collecting the number of responses you need. The balance between sample purity and collection speed is one of the main challenges in designing a screener.

A screener that gives itself away. If a question sounds like "Are you our customer? (Survey for customers only)" — that's not a filter, it's an invitation to lie. The survey's title and description must not contain the selection criteria. A neutral description like "A study of consumer habits" works significantly better.

Filtering by self-assessment. Questions like "Do you consider yourself an advanced user?" give unreliable results: people systematically overstate their competence. Where possible, replace self-assessment with objective indicators: tenure, frequency of use, specific actions.

The absence of an "Other" or "None of these categories" option. If a respondent can't find a suitable option, they pick the nearest one — and contaminate the data. Always leave an "exit," even if it leads to ending the survey.

Screening and logic jumps

A screening question is a special case of a broader mechanic: logic jumps (branching logic, skip logic). Logic jumps let you route a respondent along different paths within the survey depending on their answers. Screening is a logic jump in which one of the paths leads straight to the final screen.

But the possibilities of logic jumps are broader. You can do more than just filter out off-target people — you can split target respondents into different branches. For example: send everyone who selected "I use it daily" to a block of questions about intensive usage, and those who selected "I use it once a month" to a block about barriers to more frequent use. In both cases the respondent stays in the survey but sees only the questions relevant to them.

For more on designing high-quality surveys with branching, see the guide to creating surveys.

How many screening questions to use

There's no universal number, but there's a benchmark: one or two, in exceptional cases three. Each additional screener reduces conversion — the share of people who reach the main part. If there are too many filters, the Response Rate drops, and the cost of a single complete response rises.

When there are more than three selection criteria, it's worth asking: are they all really necessary? Often one well-formulated question replaces two or three superficial ones. Instead of the chain "Do you have a car?" + "Is it a passenger car?" + "Did you buy it in the past 2 years?" you can ask: "Did you buy a new passenger car in the past 2 years?" — one question, three criteria.

Screening questions in SurveyNinja

In the SurveyNinja builder, screening is implemented through the logic jumps mechanism. Setting it up takes a couple of minutes: you create the first question of the survey, set a condition (for example, "If the option \"No\" is selected — go to the final screen"), and from there the platform does everything automatically.

A few features that are useful specifically for screening:

• Flexible transition conditions. You can configure routes not only by "yes/no" but also by specific answer options, by combinations of several conditions, by numeric values. This lets you build multi-stage filters without unnecessary questions.
• A final screen with custom text. Screened-out respondents see not a standard placeholder but a message you wrote yourself. You can thank them for participating, explain the reason, and even provide a link to another survey that this person is a fit for.
• Hidden variables. Through hidden variables in the URL you can pass information about the traffic source — and analyze which channel brings the most off-target responses. This helps optimize the distribution of the survey.

A screening question isn't a bureaucratic barrier but a tool for protecting data quality. Two minutes to set it up saves hours of cleaning and rechecking results.