How to create an Internet survey

By APSIS

2011-08-23

The answer to a question always depends on how the question is posed. This is a simple truth that will guide you as you create Internet surveys. A good survey with well thought-out questions will give you considerably more answers — answers that tell you the truth and that are possible to use for further analysis.

In this article, we take a look at how you should choose between different types of questions and what to thing about when writing the texts. At the end, we provide some information on what you should not do – sometimes mistakes can be revealing!

Questions with only one possible answer
Certain questions should only have one possible answer. Choose a layout with radio buttons or a drop-down menu.

Questions with multiple answers
Questions where it is possible to answer one or several different things (such as the question ”What or which newspapers do you read?”) should have clickable boxes.

Matrix questions
If you would like to have several questions on the same topic, use a matrix question. This is an excellent option if you want to know how your product is perceived on different levels (such as usability, pricing, sustainability) on a scale from 1-5.

Open questions
If you would like to give the respondent a chance to use his or her own words, use an open question. It is possible to have an open question directly following a related closed question if you want longer answers.

A few things to avoid
It is important that your questions are correctly asked in order for you to get usable answers. Let us take a closer look at the mistakes you should avoid when writing questions and answer options.

- Double questions
A common mistake is to ask several questions at the same time. Take this question, for instance: ”Do you think our service and our accessibility fulfils your expectations?”. If the answer is yes, how would you know which alternative the respondent meant? Is is the service or the accessibility?
Ask a separate question for each alternative if it is important for you to know both things.

- Technical terms and abbreviations
It is easy to become used to certain expressions. Do not take for granted that everyone will understand the words you use when talking to your business colleagues. Use simple words that everyone understands – it will increase the number of completed surveys significantly.

- Avoid negations
Avoid negative questions. Instead of writing ”Don’t you think we should get a new CRM system?”, ask ”Do you think we should get a new CRM system?” Be direct.

- Leading questions
Avoid questions that highlight one of the answer options, so called leading questions. Make sure that all options are equally appealing. ”What is your experience of using the razor?” is a better question than ”Do you find the razor difficult to use?”.

- Words with value
How much is enough? What is the characteristics of something good? How often is sometimes? The words enough, good and sometimes are three out of many words with value that you should try to avoid. Try to be precise in your questions. Use numbers, size, or amount as exactly as possible to give the respondent an idea of what you mean.

- Questions in the right order
The order of the questions have a significant impact on the number of respondents that complete your survey. A good tip is to start off with a couple of easy questions, and save the difficult and more demanding questions for later. They will not feel that hard this way. At the end of the survey, you return to a couple of simple, positive questions again. It is also useful to give the respondents the possibility to comment a couple of open questions. Close the survey with questions about gender, age and income. These are sensitive questions, but when placed at the end they usually become answered.

- Overlapping options
It is easy to get a bit confused when creating answers. Consider the question ”How old are you?”. There are overlapping options at the top example and correct options below. Which box would you tick if you were 35 years old?

0-15 years
15-25 years
25-35 years
35-45 years
45 or older

0-15 years
16-25 years
26-35 years
36-45 years
46 or older

- Missing options
Your survey has to have sensible answer options to present a realistic image of the world. If an option is excluded, consciously or subconsciously, the survey might show a totally faulty trend. You have to carefully consider which answer options to include in order to get a fair result. If you examine shopping habits and exclude the option ”Bulk purchases”, the survey will show that the respondents mostly shop in small quantities, which may not be the actual case.

- Neutral options
There are different opinions on how to use neutral options such as ”Don’t know”, ”Either/or” or ”No experience”. Many surveys exclude neutral options to get a clearer picture, something that may be effective in certain situations, such as attitude measurements. Be careful with this, though – you run the risk of creating a leading question where certain answers are highlighted. You should not expect everyone to be willing to take a stand in any given question, but rather give the respondents a neutral option on as a general rule. Your survey will be more accurate, since the respondents will give honest answers when it feels right to do so.

Some usable tips
• Make it possible to give anonymous answers and present your anonymity policy in the survey. More people will dare to give honest answers.
• Most participants want to know the results of the survey. Offer them the possibility!
• Only ask about things you really need to know. Make the survey quite short – you will get more answers.

© APSIS International AB. All rights reserved.