Introduction
Questions may consist of one variable (e.g., age) or multiple variables (e.g., aided awareness of multiple brands). Q is structured around questions and it automatically takes into account the different types of questions. Q automates analyses by understanding what the data means and this is done using the question type assigned to variables. Many common analyses are achieved by changing question type.
Method
Question Types
Q recognizes 13 different types of questions, such as Pick One, Pick One - Multi, Pick Any, Numeric and Text among others. See Question Types for a complete list and description of the different question types.
Setting and changing question types
Question Types are automatically set when data is imported into Q. However, Question Type can be modified by changing the Question Type drop-down in the Variables and Questions tab (this can be done by modifying the Question Type value for any one of the variables in a question).
Multiple variables can be grouped together as a question by selecting the variables in the Variables and Questions tab, right-clicking and selecting Set Question.
Once Question Type has been set or changed, there are some additional settings that change how the data is processed:
- If changing to a Pick Any or Pick Any - Grid question, the Count This Value settings in the Value Attributes will determine how percentages are computed.
- If changing to a Number, Number - Multi or Number - Grid question, the Value column determines the values used in computing any averages (you can recode the data by changing the value in the Values column).
- If changing to a Ranking question, the Value column determines the order of the ranking (higher values are treated as being ranked higher).
Split a question
If Q incorrectly combines variables together, or for some reason you want to split one of the variables off from the others, you can separate them. Tobreak the variable set into its constituent variables
- Select the Variables and Questions tab
- Right-click and select Split Variables From Question. This will split the question out into individual variables.
"Fake" questions
Q's Question Types are all examples of the types of questions that appear in questionnaires. In addition to ensuring that a question's Question Type has been set correctly, it is generally useful to also have “fake” questions. That is, it is useful to setup up the data as if the questions asked in the survey are different to those actually asked. For example, the true Question Type of the following question is Pick One - Multi:
Please rate your satisfaction with the following aspects of your bank...
Very Dissatisfied | Somewhat Dissatisfied | Neither Satisfied nor Dissatisfied | Somewhat Satisfied | Very Satisfied | |
---|---|---|---|---|---|
Fees and charges |
⃝ |
⃝ |
⃝ |
⃝ |
⃝ |
Customer service |
⃝ |
⃝ |
⃝ |
⃝ |
⃝ |
Online banking |
⃝ |
⃝ |
⃝ |
⃝ |
⃝ |
It is often useful to restructure the data from such a question as if a Pick Any question had been asked (i.e., to do a top 2 box analysis), such as:
Which of these are you satisfied with? Your bank’s...
⃝ Fees and charge
⃝ Customer service
⃝ Online banking
Alternatively, if interested in presenting average satisfaction scores, the data can be restructured as if the following question had been asked:
Please rate your satisfaction with the following aspects of your bank on a scale of 1 to 5..
- ___ Fees and charges
- ___ Customer service
- ___ Online banking
This is done by changing the Question Type to Number - Multi.
Usually, it is best to copy the variables from a question and turn these copied variables into the “fake” question.
Undoing a question
Sometimes a question will have been set up in such a way that it is not clear what the question means. You can view the original variable labels (the source labels read from the raw data file) in the Variables and Question tab by pressing the source labels icon in the toolbar, which causes a new column, Source Label, to appear between Name and Label.
To change variable labels back to their original labels, right-click on the variables of interest and select Revert to Source Label (this option will only appear if the Source Label and the Label are different). To undo the setup of a question in it's entirety, select all variables in the question, right-click and select Revert to Source.
See Also
How to Create a Second Version of a Variable
How to Combine Variables into a Single Question
How to Modify Variables and Questions
Comments
0 comments
Article is closed for comments.