Even if you are not yet familiar with the term Likert Scale, chances are you’ve come across scale question types. They are among the most comment types of questions asked, particularly in online surveys. While easy for respondents, it’s the variety in the units that can make Likert Scales difficult or intimidating for survey creators to write.
Here’s how the pros take the guesswork out of Likert Scales.
At their most essential, they ask for a single-select response with a limited set of options representing a range, scale or continuum. Scale questions are designed to capture the survey taker’s opinion or sentiment. The units in the scale vary.
In this post, we help you determine the right units to use for your Likert scales depending what factor you’re trying to measure.
Common Likert Scales and Their Units
You can measure almost anything using a Likert Scale, and you can use any number of points along that scale to capture the nuances to your audience’s reactions.
One of the most common use of a Likert scale is to measure a customer’s satisfaction with a product or service. Usually, scale goes from “Very Dissatisfied” to “Completely Satisfied.” See this example:
There are many other variations of this basic Likert scale. Try using it to measure one of the following:
- Acceptability (Not at All – Completely Acceptable)
- Agreement (Completely Disagree – Completely Agree)
- Awareness (Not Aware at All – Extremely Aware)
- Concern (Not at All Concerned – Extremely Concerned)
- Familiarity (Not Familiar – Very Familiar)
- Frequency (Never – Always)
- Importance (Not Important – Extremely Important)
- Likelihood (Not Likely – Extremely Likely)
- Quality (Poor – Excellent)
Remember: just because you can create as many points along the scale as you’d like, doesn’t mean you should. Our research shows that 41% of survey builders prefer to use a numbered scale with units from one to five.
Writing for the Middle: What to Label the Midpoints on a Likert Scale
For many Likert scales, the biggest challenge for survey creators is determining the best label for the scale’s midpoint.
Not every scale needs a midpoint. Some questions, like one that measures awareness or familiarity, may not lend themselves to having a neutral or ambivalent center. In these cases, it’s perfectly acceptable to include an even number of points.
For example, for a question that measures familiarity, you could use the following units: Not at All Familiar, Slightly Familiar, Somewhat Familiar, and Very Familiar.
While your familiarity scale could measure a center point, it may not be necessary for your data, particularly as the differences may be subtle or unclear to respondents.
For other questions, a center point is important, as it captures a balance between extremes or a neutrality, impartiality, or lack of opinion.
In the following example, the units are: Very Below Expectation, Did Not Meet Expectations, As Expected, Above Expectations, Greatly Exceeded Expectations.
Some other possible labels for Likert Scale’s midpoint include:
- As Expected
- Neutral
- No Opinion
- Neither Agree Nor Disagree
- Unsure
- Occasionally
- Sometimes
- Maybe
The best midpoint label will depend on context.
Creating Easy to Use, Easy to Analyze Likert Scales
The best response scale is one that is easy to understand, clearly discriminates between respondents’ perceptions with clearly written labels.
In addition to text labels, a numeric value, also called a reporting value, is associated with each point. This makes it easier for you to compare and analyze the responses you collect.
Best Practices for Multiple Likert Questions
When using multiple Likert questions within a single survey, it is best to set up the questions with the same scale and labels.
The most important reason to do this is to minimize survey fatigue and maintain response accuracy for your respondents. It’s simply easier for survey takers if the scale questions you ask are consistent across the survey, and they are less likely to make mistakes in their responses.
Using a consistent scale also tends to make reporting and analysis easier for your team, too. It will be easier to identify inconsistent or inaccurate data, which can then be cleaned.
If your survey is part of a series or longitudinal study, maintaining consistency in scales between surveys makes it easier to compare trends in data over time. This way, you’ll be comparing apples to apples rather than spending extra time trying to convert data collected on a 1-5 scale to that collected by a 0-10 scale.
Achieving Survey Success with Likert Scales
With the right units, clear labels, and consistent format across all of your questions and surveys, your Likert scales will help you collect more and better data on your audience’s perceptions and opinions.