How and when to uses bipolar answer scales (e.g., agree to disagree) 🔗

If you are asking questions to which participants may either agree or disagree, or any other form of double-sided scales (i.e., any question that has the possibility of an answer that is either negative or positive answer, or the possibility of an answer that is either agreement or disagreement), we recommend using a seven-point answer scale.

We recommend this "agreement" scale (as long as it makes sense to use it in the context of your question):

How and when to uses bipolar answer scales e g agree to disagree (1)

For more about why we like this scale, see this .

Good bipolar scale use examples 🔗

Here are some good examples of how and when to use an agreement scale:

>> agreementScale = [["Totally agree", 3], ["Agree", 2], ["Somewhat agree", 1], ["Neither agree nor disagree", 0], ["Somewhat disagree", -1], ["Disagree", -2], ["Totally disagree", -3]]

*maintain: Please indicate the extent to which you agree or disagree with the following statement.

*question: I like to be the center of attention.
	*answers: agreementScale

*question: I am a happy person.
	*answers: agreementScale

*question: I feel free to choose what I want to do in my life.
	*answers: agreementScale

*clear

Note that in the example above, the columns associated with these questions (in the CSV data file) will be stored with numerical values 3, 2, 1, 0, -1, -2, -3, rather than with the text values "Totally agree," "Agree," etc. Having the answers stored as numerical values makes data analysis much easier!

Also note that for two-sided/bipolar scales (like the agree/disagree scale above), if the center of the scale is neutral, and the extremes of the scale represent opposing components of the same concept, then it's usually best to code the numerical values with both negative and positive values (-3, -2, -1, 0, 1, 2, 3). On the other hand, with one-sided/unipolar scales (such as a scale about a specific factor that someone might report more or less of, but that they wouldn't report a negative value for), you will usually want all numerical values to have the same sign, e.g., 0, 1, 2, 3, 4.

Bad bipolar scale use examples 🔗

Here's a bad example of an unclear answer scale with varying "distances" between answers:

*question: How much do you like cake?
	I love it!
	It's great!
	So so
	I don't like cake

The example above includes a weird scale where the semantic distances between answers are not uniform, and the reader has to do extra work to understand each answer option.

Here's a bad example of unnecessary repetition instead of coding for an answer scale:

*question: I like to be the center of attention.
	Totally agree
	Agree
	Somewhat agree
	Neither agree nor disagree
	Somewhat disagree
	Disagree
	Totally disagree

*question: I am a happy person.
	Totally agree
	Agree
	Somewhat agree
	Neither agree nor disagree
	Somewhat disagree
	Disagree
	Totally disagree

In the pair of questions above, note how the answer options are re-listed redundantly. This requires repeatedly typing or copying-and-pasting the scale, which is unnecessarily time-consuming and prone to error. Also, the answer options above lack coding variables.

To avoid the unnecessary repetition (as well as the errors that can come from this), and to add coding variables in a much less tedious manner, you can make an answer scale and use it with *answers: agreementScale , as was done in the good examples earlier.


Next: