Determining your respondents’ actual attitudes when conducting your research is essential to improving the accuracy of your findings. By the end of this article, you’ll have a better understanding of what a Likert scale is and how you could use it to assess your respondents’ true sentiments.
Inhaltsverzeichnis
Likert Scale – In a Nutshell
- Likert scales are instrumental when there’s a need for detailed opinions and attitudes.
- The most common scale used in Likert-type rankings is a 1-5 scale (1 being strongly disagree, five being strongly agree).
- A Likert scale can quantify opinions, attitudes, and behaviors.
Definition: Likert scale
- A Likert scale is a rating scale that researchers use to establish a respondent’s real attitudes and opinions.
- A typical Likert scale in a research survey consists of a question or a statement followed by a string of answers.
- The respondent then picks a solution that best describes their feelings about the question.1
Creating a Likert scale
The questions are phrased as statements rather than questions, and each answer option is assigned a number.
Also, you should avoid double negatives when phrasing your questions
Example:
“How often do you eat lunch?”
- 1 for “every day”
- 2 for “3 times a week”
- 3 for “once per month”
- 4 for “never” and
- 5 for “sometimes”
Example:
X “Do you not eat cereal?”
✓ “Do you eat cereal?”
Questions or statements
- A Likert scale question is used to measure the opinions of your respondents, while statements are used to give them a chance to express their views.
- Using a Likert scale statement instead of a question is appropriate if you want to find out what your readers think about something without measuring their opinion with numbers.
- On the other hand, Likert scale questions should be used when you want to measure the opinions of your respondents on a particular topic, such as whether or not they think it is essential for children to be able to read before the age of 8.
Use positive formulations
A positive and negative formulated Likert scale is a scale in which a respondent can select one of two possible answers to an item on the scale, such as “strongly agree” or “strongly disagree.”
Keep it simple
As you create your survey, consider making it as clear as possible by asking only one question at a time.
Answer options in the Likert scale
Besides measuring degrees of agreement or disagreement, a 5-point scale can also be used to measure quality and probability.
Typical answer options
Choosing the correct answer option for your scale can be tricky. So, here are some of the most common alternatives to consider:
| Agree: | Quality: | Probability: | Experience: |
| • Strongly disagree • Disagree • slightly disagree • slightly agree • agree |
• Very low quality • low quality • moderate quality • good quality • very high quality |
• Never happen • low probability • happen sometimes • high probability • always happen |
• Negative experience • neutral experience • positive experience |
Unipolar and bipolar answers
A unipolar scale survey is a survey that asks respondents for their opinions on one topic. This could be a survey asking people about their favorite ice cream flavors.
This is different from a bipolar scale survey, which asks people to answer multiple questions about the same topic:
Examples:
- “Do you prefer ice cream that’s creamy or crunchy?”
- “Do you prefer ice cream in cones or sundaes?”
In most cases, surveys will be bipolar because the answers are nuanced and not just one-dimensional. 3
Analyzing Likert scale data
When analyzing data, it’s essential to know whether you have ordinal data or interval data.
Ordinal data vs. Interval data
Ordinal data is a type of data that can be ranked or ordered, but the differences between the data values are not necessarily equal.
By contrast, interval data has a specific value that can be measured.
Example:
If you want to measure how easy it is for people to do something, ask them to rate their ability from 1-5 (1 being easiest and five being hardest).
Example:
If you wanted to know how tall someone is in inches or centimeters, you could measure their height in inches or centimeters.
The Likert scale is an example of ordinal data because it asks respondents to rank their agreement or disagreement with statements.
Testing statistics
- Testing statistics determine whether or not there is a significant relationship between two variables.
- The most common testing statistic is the correlation coefficient, which can be used for interval data and is expressed as a number between -1 and 1.
- A higher value means a more substantial relationship exists between the two variables. In comparison, a lower value means less of a relationship between them.
Example:
Let’s say you were interested in the relationship between how much your students like their school and how much they learn in class.
You could take two classes of students, give them a survey on how much they like their school, and then provide them with a test on what they learned in class. Then, you could calculate the correlation coefficient between these two variables.4
Descriptive statistics
Likert scales gather information about how people feel about different topics.
Descriptive statistics are used to summarize the collected data.
To analyze Likert scale data you could use descriptive statistics like:
- mean (average)
- median (middle value)
- mode (most popular value)
- standard deviation (how spread out the values are)
Pros and cons of the Likert scale
Understanding each option’s pros and cons is essential before deciding what scale to use.
|
✓ Pros: |
X Cons: |
| • Scales that use Likert-type questions are straightforward. • The Likert scale can measure people's opinions, feelings, and attitudes. • Likert-type scales are easy to construct, administer, and score. • It's a great way to understand what people think about a topic. • You can have multiple questions on one page that you can use to see how people respond differently. |
• You must be sure how respondents interpret the questions in your Likert scale. • Likert-type scales are not well suited for survey questions requiring various responses. • Knowing how much people agree or disagree with you can be challenging. • You can only get a complete picture if you ask enough questions and are clear about what your audience is looking for. • There's a risk of bias in favor of one side of an issue. |
FAQs
A Likert scale question uses a five or 7-point scale to help researchers better understand their respondents’ beliefs.
Likert scales are used to measure attitudes and opinions.
The response options for most Likert scales are 1 (Strongly Agree), 2 (Agree), 3 (Neutral), 4 (Disagree), and 5 (Strongly Disagree).
Sources
1 Typeform. “What is a likert scale survey, and what are they good for?” Accessed January 9, 2023. https://www.typeform.com/surveys/likert-scale-questionnaires/.
2 McLeod, Saul. “Likert Scale Definition, Examples and Analysis.” Simply Psychology. August 3, 2019. www.simplypsychology.org/likert-scale.html.
3 SurveyMonkey. “What is a Likert scale?” Accessed January 9, 2023. https://www.surveymonkey.com/mp/likert-scale/.
4 AFIT Data Science Lab R Programming Guide. “Assessing Correlations.” Accessed January 9, 2023. https://afit-r.github.io/correlations.