The Likert Scale Explained
You’ve probably already come across this common type of rating scale. Maybe it was an email survey from a department store asking how satisfied you were with a recent visit. Or it might have been that postcard handed to you at the last sales conference you attended.
If you’re trying to measure someone’s attitudes or behaviours, a Likert scale is one of the most popular (and reliable) ways to do so. A Likert scale measures attitudes and behaviours using answer choices that range from one extreme to another (for example, not at all likely to extremely likely). Unlike a simple “yes/no” question, a Likert scale allows you to uncover degrees of opinion. This can be particularly helpful for sensitive or challenging topics or subject matter. Having a range of responses will also help you to more easily identify areas of improvement – whether you’re sending out a questionnaire to understand the levels of effectiveness of the course you’re teaching or gathering clients’ opinions regarding the quality of service at your beauty shop.
Here are some best practices that you’ll find handy to read before you embark on your questionnaire or survey.
Likert Scale Best Practices and Tips
- Keep it labelled. Numbered scales, or scales that are marked from 1 to 5, for instance, have also proven problematic for survey respondents, since they might not immediately know which end of the spectrum is positive and which is negative. Instead, use words to label your scales, i.e. ranging from ‘poor’ to ‘excellent’.
- Keep it unipolar. Whenever possible, try to use a “unipolar” scale that ranges from “extremely” to “not at all", rather than extremely one thing to extremely another. For example, it’s better to use a scale that ranges from “extremely brave” to “not at all brave”, rather than a scale that ranges from “extremely brave” to “extremely shy”. Why? Unipolar scales are just easier for people to think about and you can be sure that one end is the exact opposite of the other, which makes it methodologically more sound too.
- Keep it odd. Scales with an odd number of values (e.g. 1-to-7, 1-to-9, 0-to-4) will have a midpoint. How many options should you give people? Studies have shown that respondents have difficulty defining their point of view on a scale greater than seven. This means that if you provide more than seven response choices, people are likely to start picking an answer randomly, which can make your data meaningless. What’s the magic number to use? Our methodologists recommend five scale points for a unipolar scale and seven scale points for a bipolar scale.
- Keep it continuous. Response options in a scale should be equally spaced from each other. (The distance between scale points should be the same throughout the scale, which makes the scale clear and less ambiguous.) This can be tricky when using word labels instead of numbers. Check out our methodology blog for some helpful tips.
- Keep it inclusive. Scales should span the entire continuum of responses. For example, if a question asks how hot your coffee was and the answers range from “extremely hot” to “moderately hot”, then respondents who think the coffee wasn’t hot at all won’t know which answer to choose.
- Keep it logical. Add question logic to save your survey takers some time! For example, imagine that you want to ask how much your customer enjoyed your restaurant and then only obtain further details if they were unhappy with something. To do this, just add question logic so that only those who are unhappy skip to a question asking for improvement suggestions.
- Keep it interrogative. Ask questions whenever possible instead of using agreement with statements. Avoid scales of agreement/disagreement whenever possible as they are subject to an acquiescence bias. This means that survey takers are more likely to agree with statements, regardless of what they say, and less likely to read the question carefully.
Whether you’re measuring satisfaction, effectiveness, likelihood or frequency, consider using a Likert scale for your next survey or questionnaire! Here at SurveyMonkey, we recommend using a Multiple Choice (Single Answer) Question Type to create a Likert scale.