Screening questions

Screening questions let a survey creator target questions to a specific audience by filtering respondents. With screening questions, respondents first see your screening question and then those who select a threshold answer such as “Yes” or “I plan to” can answer the remaining questions in your survey.

We recommend that the first screening question in the survey have three or more answer choices. Validation tests have shown that incidence accuracy is better when you don't use only binary answers like Yes/No. This helps eliminate confusion for the respondents, provides a better user experience, and improves survey quality.

There is a maximum of four screening questions per survey. Additionally, surveys with screening questions require a minimum purchase of 100 responses.

For example, if you only want answers from dog or cat owners, you would start your survey with a screening question like "What type of pet do you own, if any?" You would flag answers such as "Dog" and "Cat" from respondents who you'd like to continue to take this survey. From there, you could ask them "How many times did you take your pet to the groomer last year?", and "How many times did you groom your pet at home last year?"
How do I set up a screening question?
  1. In Step 2 of the survey creator, select your question type, and enter your question and answers.
  2. Select the check box for Screen with this question.
  3. Select the check box checkbox next to each answer that qualifies a respondent for this audience.

Please note that not all question types can be used as screening questions.

Learn more about our question types

How do screening questions affect the price of my survey?

Since screening questions affect the incidence of your survey, including them in your survey requires an audience test. Surveys with screening questions that require an audience test have custom prices.

Learn more about our audience testing process

 

How do I set up my screening question(s) to succeed?
  1. Whenever applicable, the multiple-answers format is recommended for questions where respondents can identify with more than one answer choice and select a combination of answers. Multiple-answers format has a system-default opt-out answer "None of the above", which is automatically screened out when the question is set as a screening question. When looking for dog or cat owners, you might ask "Which of the following pets do you own, if any?" and provide these answer options::
    • Cat [screen-in]
    • Dog [screen-in]
    • Fish
    • Bird
    • None of the above [automatic screen-out]
  2. Questions about ethnicity or race, where some respondents may identify with more than one, are best written in the multiple-answers format, especially if you are screening in for one or a few. For example, for the question “Which of the following describes your ethnicity or race?”, add these answer choices:
    • White
    • Black or African American
    • Asian or Asian American
    • Hispanic or Latino
    • Other/Prefer not to say
    • None of the above
    Questions that ask about sensitive demographics, such as race or ethnicity, should provide answer choices in groups or ranges with an "I prefer not to say" or "Other" as the opt-out answer. While the multiple-answers format already includes a system-default “None of the above”, with single-answer format for sensitive demographics questions, you would have to add "I prefer not to say" or "Other".
  3. When you prefer to use the single-answer format, try to turn a binary question such as "Do you own a pet?" into "Do you own at least one of these pets?" We recommend that the question provides all possible answers instead of simply leading the respondent to choose between one answer or the other, such as “Yes” or “No” only. For example:
    • Cat
    • Dog
    • Fish
    • Bird
    • Other pet(s)
    • I don't own a pet
  4. While a survey can have up to 4 screening questions, it may not always be ideal to use this many. The incidence rate of the survey is based on the rate of the last screening question in the survey when there is more than one. For example, if your first 2 screening questions screened in ~5% of the population (which is already small), the last 2 screening questions may further reduce the population of respondents to a tiny fraction and result in an overall incidence rate below the acceptable minimum of 5%. In this case, you may need to consider using fewer screeners and broadening the audience of your survey.

Learn more about our incidence-rate threshold

Was this helpful?
How can we improve it?