"What did you eat for dinner last night?"

Example of an opinion question:

"What is your favourite food?"

A questionnaire often consist of both fact and opinion questions.

You want to collect data from a large number of people in many different locations.
What you want to know is fairly straightforward, i.e. relatively brief and uncontroversial information.
You have time to wait for the research data. Questionnaires can take a lot of time to plan, write, send out and retrieve.
The respondents can be expected to be able to read and write. Take into consideration the respondents' age, intellect, culture and eyesight etc.
You need standardized, easy to compare, answers form identical questions.

The questionnaires can in general be divided into two types, the postal type and the internet type. It is often hard to add questions to a survey once it has been sent out so it is important plan the questionnaire thoroughly in advance. In order to make it as easy as possible for the respondents to answer, every survey should be as short as possible and focus on only the crucial issues of the research topic. In brief the survey should be as user friendly as possible. Here are some guidelines one can follow to ensure this:

Only ask questions vital for the research. Try to be as concise as possible.
Sort out duplicate questions. For instance if you in the same survey both ask for the respondents date of birth and in another question ask for his age one of these questions can go.
Try out the survey in advance and consider the length of the questionnaire as well as the questions asked.
Consider the graphical layout of the questionnaire to minimize the risk of confusion.
Make sure that the questions are not irritating, annoying or offending. In order to get truthful answers you have to gain the respondents trust and respect.
Avoid the use of leading questions.
Avoid vague questions.
Make sure that the questions only can be interpreted in one way.
Use a minimum amount of technical language so that anyone can understand.
Keep the questions as short and straightforward as possible. This will avoid unnecessary misinterpretations and make the process of answering the questions as easy as possible.
Avoid words or phrases that may cause offence. Always strive for using the political correct term.

The order of the questions is also something to consider when you are composing a questionnaire. Start with the most straightforward and least sensitive questions. Further on when you have gained the respondents trust you can ask the more personal and sensitive questions. Also make sure that the order of the questions are not steering the respondents to answer the questions in a certain way. That is to say the answers to later questions should not be affected by previously asked questions.

There are a number of ways that one can ask questions in a survey but they are categorised as either open questions or closed questions.

The open questions are those that leave the respondent to express their views, feelings and opinions in their own words. In general an open question is a question with a number of empty lines for the respondents to fill in the answer. The advantage of an open question is that it gives room for the respondent to express his own opinion and the response is more likely to reflect the complexity of the respondents' feelings and views. The disadvantage of using open questions on the other hand is that it is more of an effort for the respondent to fill in a survey full of open questions. The second disadvantage is that it is rather time consuming to process and analyse the raw data from open questions.

A closed question is questions where the respondent answers by choosing among a number of answers defined in advance by the researcher. The answers can have as little as two answers to chose from (Yes or No, Male or Female) or can have a more complex set of answers to chose from. The advantage of using closed answers is that the research data takes minimal effort to analyse and it is easy to compare the answers to each other. The disadvantage on the other hand is that you will only get the answers that you have defined and there's little room for the respondent to express exact facts or true feeling. More complex questions will therefore be answered with less subtlety as if an open question was asked.

Email/webpage
The advantage of internet based questionnaires is the ability to reach a lot of people at a low cost. The main challenges are to bypass email spam filters and to get access to relevant email address lists. Some researchers email a link to a webpage where the user can click trough a questionnaire that is then sent straight in to the researcher's database. There are readily available scripts for this at no or little cost.

Paper/broschyr
This is the most common type of questionnaire where the questions are printed on a sheet of paper or in a brochure or similar. This can be passed out to a group of people or sent out by mail. Common methods of collecting the answers may be to include a return envelope.

Assisted/interview
This version borders to being a rigidly structured interview. The main difference from ordinary questionnaires is that the researcher asks the questions and fills in the answers. This may be done on the street, over the telephone or in a meeting. The advantage is that no question is skipped and a sampling process can be built in to the process. Example: standing on a street corner in a crowded place and approach people that seem to be in the target group.