Posing the right question

Be clear and specific and keep it simple

The respondent must directly understand what you mean with your question.

• Try to pose a question as shortly as possible.

  Example of a wrong question:: "If, during your assessment interview, you are confronted with problems with which you are or were involved, and if you then subsequently are asked to reflect on your own actions, would you then try to imagine what your conversation partner is trying to achieve with this question?" This question is far too long and far too complicated.

   

• Use conversational language and avoid abbreviations. If you can’t avoid using difficult words, make sure you explain them.

  Good example: "How would you assess the GUI of the online survey tool? GUI stands for Graphical User Interface;

   

• Avoid double negatives in the question.

  Example of a wrong question: "To what extent were the following reasons not influential in your not choosing our product?" So never use the word ‘not’ twice in one question.

   

• Do not pose suggestive questions (and do not suggest answers) and keep questions as neutrally as possible. In this way, you prevent the respondent from doubting your intentions and, more importantly, that you are steering the results.

  Example of a wrong question: "ave you ever made use of this attractive special offer?" or
  "Do you also think that NETQ GROW offers a good solution for easily holding online surveys?"

   

• Do not take any answers for granted and make no assumptions.

  Example of a wrong question: "Are you sometimes irritated by telephone sales people??"

  Good example: First ask the question "Are you ever phoned by a telephone sales person??", then ask "How often are you phoned by telephone sales people? " and finally ask "What is your opinion of being phoned by sales people?"

   

• Formulate questions as unambiguously as possible. In other words, make sure that a question can only be interpreted in one way.

  Example of a wrong question: Does the question "Do you like football?" mean 'does the respondent enjoy playing football', or 'enjoys watching football on television', or 'enjoys reading about football', or …

  Example of a wrong question: When you say "Have you in the last year…?" does that mean 'in the current year until now …', or 'in the whole of the previous year …', or 'in the past 12 months …'.

   

• Stay current. The memories of respondents often don’t go back that far. Therefore avoid questions about matters that are too far back in time or too far ahead. The rule of thumb is 1 month.

  Example of a wrong question: "A year ago, you received a letter about road works in your neighbourhood. Did you find the contents of that letter useful?"

   

• Do not ask questions about matters that respondents cannot easily estimate or know.

  Example of a wrong question: "What is the distance from your house to the nearest petrol station?" or
  "Is your gross annual salary above or below the average gross annual salary in your municipality?"

   

• Be as specific as possible. Avoid indeterminate words such as 'little', 'a lot', 'sometimes', 'often' etc.

  Example of a wrong question: "Do you often carry out online surveys?" The word 'often'can mean one thing for one respondent and something else for another. It is therefore better to ask "How many online surveys have you carried out yourself in the last month?"

   

• Make sure you don't ask two questions in one.

  Example of a wrong question: "Are you satisfied with the service and the price?" These are two questions; if the answer is you don’t know whether the respondent is only satisfied with the service, only satisfied with the price, or is satisfied with both.

Keep focus

Stay relevant and don’t unnecessarily burden respondents.

• Do not make your questionnaire too long (or too lengthy).

   

• Ask valid questions that can give measurable answers to the aimed research question(s).

  Example of a wrong question: If you ask the question "What do you enjoy drinking most?" and give the answer options 'coffee', 'tea', 'beer' or 'wine' the respondent cannot really give any answer, because as a rule these drinks are enjoyed at different times of the day. It is better to ask "What do your prefer to drink in the morning?", or "What do you prefer to drink with dinner?", or "What do you enjoy drinking most in the summer?" etc.

   

• 'Need to know' or only 'nice to know'? There are often questions included that may be ‘nice to know’, but have no added value for the research objective. In this way, respondents are taxed unnecessarily, which can mean that the reliability of the research declines. Therefore look at each question critically.

   

• Keep the respondent awake by making use of different types of questions. In this way, you can prevent the respondent becoming tired and completing the questionnaire on automatic pilot, so that the reliability of the research decreases.

   

• Avoid too many open questions to make sure you don't tax respondents too much. In addition, respondents generally do not give extended answers to open questions, unless they have a very forthright opinion. And also, open questions are much more difficult to analyse than closed questions, and it takes more time to analyse too.

   

• Do not ask for an explanation too frequently. An explanation can be very interesting and can supply qualitatively valuable information, which can give your research more depth. If, however, you ask this too frequently, the respondents have the tendency to reply positively and so avoid the explanation question. So ensure you only ask for an explanation if the respondent is (very) dissatisfied or has given a mark lower than a 7.

   

• Show a progress indicator so that a respondent can estimate how far he has progressed with the questionnaire.

Remain acceptable

• Avoid questions that could be sensitive for respondents. And if you nevertheless want to ask such a question, ensure that the respondent is not obligated to answer.

  Example: "Have you ever had a miscarriage when you were more than 3 months pregnant?"

   

• Avoid questions to which respondents will give socially desirable answers.

  Example: "Do you think alcohol prevention among children is important?" A respondent is not very likely to answer 'no' to such a question. In this way, a bias occurs in the answers in favour of the required behaviour.