Multiple response question
A multiple response question is characterised by the fact that none, one or more than one answer can be selected from a list of preselect answers.
Example of a multiple response question
An example of a multiple response question is this one:
In which West-European countries did you ever spend your holidays?
[ ] The Netherlands
[ ] Belgium
[ ] Luxembourg
[ ] France
[ ] Spain
[ ] Portugal
The use of multiple response question
Multiple response questions are very useful in surveys. The respondent reads a list of possible answers and ticks the answers that are applicable to him. In interviews usually an open question is used. If so, the interviewer might have a list of possible answers to tick the answers the respondent mentioned.
The advantage of a multiple response question for the respondent is the speed of answering such questions. It is always faster two tick an answer than writing down an answer or a list of answers.
The disadvantage of a multiple response question is that only the answers are listed that seems to be of interest to the researcher. A respondent, who wants to add Andorra in the above list, might think that this country is of no interest for this research. From his point of view that is correct. If the scientist also wants to collect such data, he might add an alternative in which the respondent can write his added answer in a box or on a line.
The advantage of multiple choice questions is firstly the speed of data collection and secondly, the respondents make fewer mistakes, so that time can be saved to check the data. Therefore the research can be done more quickly. Therefore the research can be done more quickly.
What kind of data do multiple response questions deliver?
A multiple response question is actually a list of dual questions. If an alternative has been ticked this means yes, and if it isn’t ticked it means no. Yes is usually coded as 1 and no as 0. This provides a number of variables that is equal to the number of alternatives in the list. Most providers deliver data with only 1 as being ticked and no is left blank. Now it is hard to make distinctions between a real no and unanswered question (because the respondent was allowed to skip these questions or did not finish the questionnaire). This might give a wrong impression of the percentage respondents that ticked the answer.
A multiple response question provides a list of variables measured at a nominal level.
One final remark. In front of the answer brackets are used. This distinguishes multiple response questions from multiple choice question where circles are used.
SPSS has a special command for analysing multiple response questions. We explain this in our SPSS-tutorials.