5 tips for good survey design

Survey image

These days, there are a number of free survey software platforms available to enable anyone to design and run a survey free of charge.  If you are taking advantage of this and designing your own survey without the help of an expert, here are some tips to help you avoid some common mistakes.

Tip #1:  Give respondents one option only

It is important to pay great attention to detail when it comes to the response options you provide your respondents.  A common mistake is to ask respondents to choose one response, yet their response could legitimately be placed in more than one option.

INCORRECT:  What age group do you belong to?  [Under 16, 16-25, 25-35, 35-45, 45-55, 55 or over].

Here, respondents who are aged 35, 35, 45 or 55 years of age could all legitimately fit into more than one category. You must organise your response categories to ensure respondents can only legitimately fit into one.

CORRECT:  What is your age? [Under 16, 16-25, 26-35, 36-45, 46-55, Over 55]

Tip #2:  Avoid asking two questions in one

Two questions are often asked in one because the survey designer may conflate two ideas they may not see as mutually exclusive. But if there are two ideas you are measuring, it is crucial that they are asked as separate questions.

INCORRECT:  How would you rate our service in terms of speed and efficiency?

Here, the survey designer does not see the need to distinguish between speed and efficiency, yet, it is possible that something may have been delivered quickly but not necessarily efficiently, if mistakes were made in the process.

CORRECT:  How satisfied were you with the speed of our service?  [1- very satisfied to 5- very unsatisfied].  How satisfied were you with the efficiency of our service? [1- very satisfied to 5- very unsatisfied].

Tip #3. Avoid leading questions

Based on their structure, questions can ‘lead’ respondents to a particular response. This is often unintentional, and is a common mistake when you are too close to the subject matter of your survey.

INCORRECT: Our organisation was recently recognised for its excellence in communication with its members.  As a member, how satisfied are you with our communication with you?  [1- very dissatisfied to 5- very satisfied]

The issues here is that the respondent is being subtly manipulated into feeling a positive disposition towards the organisation before the question has been asked. It is important to remain neutral and avoid using persuasive language.

CORRECT:  As a member of our organisation, how satisfied are you with the way we communicate with you?  [1- very dissatisfied to 5- very satisfied]

Tip #4.  Be consistent with grammar

If you ask a question using a particular adjective, it is important to use the same adjective in the responses.

INCORRECT:  How happy were you with the support you received from our help desk?  [1- very unsatisfied to 5- very satisfied]

Here, happiness is the key measure, yet respondents are given options on a scale of satisfaction. Happiness and satisfaction may be subtly different, and by using both, the question is confusing and inconsistent.

CORRECT:  How happy were you with the support you received from our help desk?  [1- very  unhappy to 5- very happy]

Tip #5.  Don’t ask too much of respondents

If your survey is too long, too complex or confusing, respondents are likely to get frustrated, skip questions and not finish the survey.  It is not possible to ask respondents everything you have ever wanted to know.  It is important that you prioritise and respect the time and effort respondents are making to answer your survey.  Try to be ruthless with your questioning and ask only what is important and what you really need to know.


You only get one chance to ask your respondents what you need to know.  If your survey does not flow, has errors and ambiguities, or is too long and drawn out, it is not worth running it, as your data will be compromised.  I recommend spending a considerable amount of time up front ensuring your survey works, is engaging and meaningful.  Even better, seek the help of an expert who can help you.  You will be grateful in the end.  As a parting exercise, please consider how helpful these survey design tips were from the options below.

1-  Extremely helpful

2- Helpful

3- Neither helpful nor unhelpful

4- Unhelpful

5- Extremely unhelpful

Back to All Posts