For every conceivable question there are several possible and theoretically acceptable forms it can take. Questions can be asked in either a closed or open format, and may employ the use of attitude scales.
Closed questions are those in which the respondents are simply asked to choose a reply from a number of predetermined options. These can be as simple as
‘Yes/No’ questions, or can be more lengthy and complex such as the standard classification of ethnicity which allows for nine different responses. Closed questions are:
• easily asked;
• easily understood;
• quick to answer;
• quick to code for analysis.
Some examples of closed questions are:
HOW MUCH INTEREST DO YOU NORMALLY HAVE IN LOCAL POLITICAL ISSUES?
A great deal [ ] 1 Quite a lot [ ] 2
Some [ ] 3
Not very much [ ] 4 None at all [ ] 5
DID YOU VOTE IN THE LOCAL ELECTION THAT WAS HELD EARLIER THIS YEAR?
Yes [ ] 1 No [ ] 2
However, closed questions are criticised for forcing respondents into a predetermined response rather than letting them answer in their own words. If closed questions are used, then it is important to ensure that the choices offered respondents are:
• Mutually exclusive – it must not be possible for an answer to fall into two categories. For instance, in a closed question which asked a respondent her or his age, the following must be avoided in which a 35 year old has the opportunity to assign her- or himself to both category 3 andcategory 4:
WHAT IS YOUR AGE?
Under 18 [ ] 1
18–24 [ ] 2
25–35 [ ] 3
35–44 [ ] 4
45–54 [ ] 5
55–64 [ ] 6 65 and over [ ] 7
• Exhaustive – you must ensure that allof the possible answers are catered for in the response options, and that you have not left something important out. For example, in a question asking people what issue would most influence their vote in a forthcoming election, the researcher needs to ensure that the list of options is not so short that it does not include the full range of issues likely to be important to respondents. Piloting the question before the full survey is conducted will provide the researcher with a clear idea of the range of answers likely to be given to the question, and which should therefore be included in the list of issues. At the very least, there must be an ‘Other’ option available to any respondents for whom the list of options does not include the issue of most importance to them.
An open question is one where the researcher asks the question and leaves a blank space for the respondents to record their response. In this sense the questions are good because they do not force the respondent into a predetermined category that can obscure nuances. They also allow the respondent greater freedom of expression, and open up the possibility for more qualitative-style data to be generated. An example of an open question used in a questionnaire is:
WHICH COMMUNITY, NATIONAL, OR INTERNATIONAL ISSUE ARE YOU MOST CONCERNED ABOUT? (PLEASE WRITE IN FULLY USING THE SPACE PROVIDED)
_________________________________________________________________
_________________________________________________________________
However, responses to open questions may be long and complex, and where written in by the respondent they may be hard to read. For these reasons, open
questions may be difficult to code and analyse. Open questions are used less often than closed questions because of their time-consuming (and therefore costly) nature.
Attitude scales consist of a number of statements that the respondent is asked to agree or disagree with to differing degrees. An example is given below:
HOW MUCH DO YOU AGREE OR DISAGREE WITH THE FOLLOWING STATEMENT?
I THINK THAT IT’S IMPORTANT TO VOTE IN LOCAL ELECTIONS:
Strongly agree [ ] 1
Agree [ ] 2
Neither agree nor disagree [ ] 3
Disagree [ ] 4
Strongly disagree [ ] 5
The use of scales is a way of measuring an individual’s position on an attitude continuum and thus generates more quasi-qualitative data. The most commonly used scale is the Likert scale. When used, a scaled question must be balanced ensuring that the responses should range from positive through neutral to negative in a unidimensional manner. When respondents are offered a scale that is skewed in some way, this is likely to induce a biased response.
A good questionnaire should include a mix of closed, open, and scaled questions.
Furthermore, it will take into account the type of data to be collected and the resources (time and money) that are available for the study. A good questionnaire is also one in which the questions that have been asked are well crafted and are the product of careful consideration. In designing a good questionnaire, researchers should do their utmost to avoid the many pitfalls that present themselves, including the use of:
• ambiguous language;
• language that is not appropriate for the targeted audience;
• prejudicial language;
• questions which lead respondents to answer a question in a way that could result in biased data.
Of equal importance is the need to structure the questionnaire so that it aids the progression of the interview. Asking the questions in a relatively logical order enables a sensible flow to develop. It should also be borne in mind that overly lengthy questionnaires are likely to put many people off from participating in a study. It is important, therefore, to be concise and economic with the subject areas in a survey as well as the language used in the actual questions.
The postal questionnaire used for the Youth and Politics project is given as an example in Extract 6.1. Notice the instructions given for respondents on how to complete the questionnaire, the general layout, the mix of question type, and the use of both open and closed questions.
EXTRACT 6.1 Questionnaire from the Youth and Politics project (Henn et al. 2002)
Please answer as many of the following questions as you are able. For all questions (except where asked otherwise), please indicate your answer by ticking the relevant box. So, if you voted in the recent European parliamentary election, you would tick the response box number one, as in the example here . . .
7. Did you vote in the recent European
parliamentary election on June 10th 1999? Yes 1 No 2 1. Are you (please tick all that come close to describing your situation):
In education (full-time) 1 In education (part-time) 2 In paid work (full-time) 3 In paid work (part-time) 4 On an apprenticeship 5 On a government training
scheme/New Deal 6
Unemployed 7 Self employed 8
Other (please write in)__________________________________ 9 2. How much interest do you normally have in local political issues?
A great Quite a Some Not very None at
deal 1 lot 2 3 much 4 all 5
3. How much interest do you normally have in national political issues?
A great Quite a Some Not very None at
deal 1 lot 2 3 much 4 all 5
4. Generally speaking, how often would you say you talk about political issues with your friends or family?
A great Quite a Some Not very None at
deal 1 lot 2 3 much 4 all 5
5. How much do you agree or disagree with the following statements?
a) I think that it’s important to vote in local elections:
Strongly Agree Neither agree Disagree Strongly agree 1 2 nor disagree 3 4 disagree 5 b) I think that it’s important to vote in national elections:
Strongly Agree Neither agree Disagree Strongly agree 1 2 nor disagree 3 4 disagree 5 c) I think that voting is a waste of time:
Strongly Agree Neither agree Disagree Strongly agree 1 2 nor disagree 3 4 disagree 5
6. Did you vote in the recent local election
on May 6th 1999? Yes 1 No 2
7. Did you vote in the recent European
parliamentary election on June 10th 1999? Yes 1 No 2 8a. Do you intend to vote in the next parliamentary
general election (for a Member of Parliament)? Yes 1 No 2 8b. If you do intend to vote in the next
parliamentary general election, do you
know which party you will vote for? Yes 1 No 2 9. Which community, national or international issue are you most concerned
about? (please write in fully using the space provided)
_____________________________________________________________
____________________________________________________________