Multiple-response questions have more than one answer per respondent. These answers need to be coded before they can be meaningfully handled in a crosstab. There are several variants of such coding, but the most common are dichotomous and categorical.
Dichotomous multiple-response set (MR-set)
Usage context: respondents were asked to choose one or more responses from a set of possible options and these options are coded as either 1 (if selected) or 0 (if not selected). A similar example of this type of coding is Yes / No answers.
Example: Name up to 5 streaming services you know
As a result, we get a range of ‘checked’ and ‘unchecked’ options for each streaming service that was mentioned. For further calculations, we need to know how many times a particular service was selected. That is when we use the dichotomous MR-set.
How to create a dichotomous MR-set
Open Meta-Editor.
Select all variables which are the answers to one multi-response question. In the example below we selected all the drinks that belonged to the UA brand.
Click on the ‘Declare multi-response set’ icon in the top menu.
In the pop-up window enter a descriptive title to label the new MR-set, one that you can recognize later. In this example, we used ‘UA-soft drinks’.
For Type tick ‘Dichotomous’ and choose what option will be counted (can be more than one). In our example, we chose ‘checked’.
Press ‘CREATE’.
Note! If the chosen options' titles contain similar parts the system reads this as a pattern and suggests a title for a new MR-set.
A dichotomous MR-set is a newly created variable summarising options such as platform services/brands/categories, and showing the number of times that the platform services/brands/categories were chosen by respondents in a survey. Below is the example for the UA-soft drinks. So you can see that Wellspring is the top soft drink brand, chosen by 1,295 respondents. All 12 soft drink brands are listed in order of preference.
You can:
- Create as many MR-sets based on the same type of variables that you need.
- Similarly use MR-sets for other variable types in the crosstab data analysis.
- Create dichotomous MR-sets with multiple options at once (see below).
Dichotomous MR-set with multiple options (more than two answers)
In DataTile a dichotomous MR-set can be created not only with dichotomous options (2 answers), but also with multivariate options (more than two answers, eg. Likert scales or Satisfaction scores). In this case, we must define what set of options will be counted as ‘checked’ for correct counting.
Example: How would you rate your overall satisfaction with {brand_name}? 1-2-3-4-5?
As a result, we get several ratings: one respondent - one answer - 5 options for each brand. Now we can create a dichotomous MR-set, choosing only options 4 and 5 - the two most favorable satisfaction ratings. They will be considered as ‘checked’ for the dichotomous MR-set.
DataTile provides the possibility to choose any number of options (in any order) to create a dichotomous MR-set based on them.
Further operations are the same as for dichotomous MR-set, where the dichotomy option is imitated by several options together.
Categorical MR-set
Usage context: respondents were asked to choose one or more responses from a set of possible options and these options are coded with unique names/numbers. Usually, these are open-ended questions where respondents get no hints and may answer whatever they want.
Example: Name up to 5 streaming services you know.
As a result, we get a range of terms or numbers. One term/number per similar answers: Netflix, HBO, Amazon, etc. For further calculations, we need to know how many times a particular service (its code) appears in the range. This is when we use a Categorical MR-set.
How to create a Categorical MR-set
Open Meta-Editor;
Select all variables which are the answers for one multi-response question;
Click on the ‘Declare multi-response set’ icon in the top menu;
In the pop-up window enter the title of the new MR-set
Tick ‘Categorical’ and press ‘CREATE’
The categorical MR-set is a new variable where the options are codes of services/brands/categories etc.
