Things to keep in mind when creating surveys

This document introduces common mistakes we see on Google Surveys. Having good survey design for your research helps to enhance data quality and a better user experience.

Spelling and grammar issues (all)

The most most mistakes involve spelling and grammar, which can be crucial for your data quality.


Using “u” instead of “you” or “Cofee” instead of “Coffee” could be misunderstood by respondents and could affect the quality of your responses.

Best Practice: 
Use a spelling/grammar checker and preview your content before submission.

Using a suitable language (all)

Questions and answers need to be written in a language widely accepted and understood in the target country.



The question and answer choices are written in Japanese while its targeting option is set to the U.S.

Best Practice:

Rewrite the question/answer choices in English, keeping the targeting option set to the U.S.


Change the targeting option to Japan, keeping the question/answer choices in Japanese.

Unnecessary capitalization (all)

Question/answer text may not contain unnecessary capitalization of words or phrases.


Unnecessary emphasis on individual words or phrases.

Best Practice:
Making a word or phrase bold by putting asterisks around it (*Apple*), or italic by using underscores (_Apple_), also has the same highlight effect.


Asking for demographic information (all)

If you plan to ask respondents to specify their age range, gender, race, ethnicity, religion, or immigration status, make sure the question meets the following requirements:

  • The answer choices are in groups or ranges.
  • There is an “I prefer not to say” answer choice.
  • For questions asking for age, the answer choices shouldn’t directly or indirectly contain any age under 18.

The answer choice “Under 18” is not allowed, and it does not include “I prefer not to say” option.

Best Practice:
Remove “Under 18” and add an “I prefer not to say”.

Inappropriate question format (all)

If you want respondents to select only one choice, the Multiple-answers question format isn’t appropriate because it allows them to choose more than one answer.


The respondents can choose more than one answer.

Best Practice:
Use the Single-answer format or Rating Scale question type instead.

Missing an opt-out answer choice (Single Answer Q)

An opt-out answer choice such as “None of the above” or “Others” can help keep the response quality high when the answer choices don’t apply to everybody.


Those who don’t drink alcohol at all won’t be able to answer the question.

Best Practice:
Add an answer option “None of the above” or “Other”.


Conflicting answer choices (Multiple-Answers Q)

More than one answer and there is an automatic “None of the above” option to the list. Make sure there are no conflicting answer choices in the list.


If a respondent picks “Germany” and “I have never been abroad before”, then it won’t make sense and result in poor data quality.

Best Practice:
Remove the option, “I have never been abroad before”.


Rating scales not clear (Rating)

The “Rating Scale” format lets people choose from a range of options. The clarity of the range is crucial in order for respondents to provide accurate responses.


The question asks “how important”, but the scale ranges are from “no” to “yes”.

Best Practice:
Edit the ranges from "Not important” to “Very important".

Character limit (Open-ended)



The question asks for as many names as possible, but respondents can enter only a maximum number of characters (e.g., 44 characters in English).

Best Practice:

Change the second sentence to "Please enter 2 to 3 names that you can recall".


Was this article helpful?
How can we improve it?