Likert scale is a type of survey scale invented by American scientist, Rensis Likert. You must have come across a survey scale when answering marketing questionnaires, or a Google form that your researcher friend had sent online for you to fill for their thesis. It is one of the most reliable methods of collecting information- it provides a range of options (numeric or verbal) on a response scale of 5 or 7 to measure the intensity of your response and if there is no opinion- to register the neutral response in the middle of the scale.
Types of Likert scale questions
Now that you know what a Likert scale is, there are some pointers that you must keep in mind for writing a strong Likert scale closed-ended question. Your goal should be to fix the respondent’s attention on one topic and make them search for their true feelings about it.
-
- Agreement questions: To check public opinion. The scale would be Strongly Agree, Agree, Neither Agree Nor Disagree, Disagree or Strongly Disagree.
- Likelihood questions: To see the probability of making a decision (like buying a product or service). The scale would be Very Likely, Likely, Neutral, Unlikely or Very Unlikely.
- Satisfaction questions: To see customer satisfaction. The scale would be Very Satisfied, Satisfied, Neutral, Dissatisfied or Very Dissatisfied.
- Importance questions: Inquire people about social subjects, policy or a certain product. The scale would be Extremely Important, Very Important, Neutral, Low Importance or Not At All Important.
Structure of a Likert scale
Before you phrase your Likert scale responses, you should know what kind of data is in the Likert scale and keep these pointers in mind:
- Decide whether you want a 5-point scale or 7-point scale. Odd number scales are more user friendly as there is a ‘neutral’ or ‘don’t know’ option besides the extreme anchors on either side. There are even measurement scales too, centred around 4 or 8 extreme options with no neutrality.
5 point example- How often do you get medical checkups? Options: Never, Occasionally, Sometimes, Often, Always
7 point example- How often do you get medical check ups? Options: Never, Rarely, Occasionally, Sometimes, Often, Very often, Always
- Depending on the statement and the kind of response you want, you can phrase it accordingly
Agreement: Strongly Agree, Agree, Neither Agree nor Disagree, Disagree, Strongly Disagree
Quality: Very Poor, Poor, Fair, Good, Excellent
Likelihood: Extremely Unlikely, Somewhat Unlikely, Likely, Somewhat Likely, Extremely Likely
Experience: Very Negative, Somewhat Negative, Neutral, Somewhat Positive, Very Positive
Remember that using two responses with the same meaning is not okay. For example, using ‘indifferent’ and ‘neither yes nor no’, mean the same thing so they should not be part of one scale
- Bipolar and unipolar scale: With a bipolar scale, you can measure both aspects of the response (for example, satisfaction or dissatisfaction) on the scale-
Are you satisfied with the results of this night cream?
Options: Extremely dissatisfied, Dissatisfied, Neither dissatisfied nor satisfied, Satisfied, Extremely satisfied
Using Likert scale in quantitative research
Due to the different point scales, it is possible to gather quantifiable data that can later be analysed through statistical and computational methods.
Interpreting Likert scale data
Before data analysis, it is important to know the distance between the responses. Are likert scales interval or ordinal- which means, are they evenly distanced or the distance between is not equal? If it is not equal, the data is ordinal. Interval-level data is equally distanced.
Analysis of ordinal data: Since it is ordinal, you can use descriptive statistics– mode, median or a bar chart would be appropriate to show the frequency of the responses.
Analysis of interval-level data: If the responses are in ranking order, you can use analysis of variance (ANOVA) or Pearson’s correlation for the inferential statistical analysis. In addition, mean and standard deviation are also used for each participant’s score. Analysing the likert scale with statistical software: It can be done through something as simple as Excel or depending on the data, with RStudio, SPSS, MATLAB, etc.
Advantages and limitations of using the Likert scale
One big advantage of using the likert scale is that it converts unobservable data into quantitative research through a user-friendly manner. Through the varied responses, you can get an accurate insight about your research topic.
Due to the respondent’s bias or laziness, they might choose an answer that is not reflective of their true feelings. Due to the restricted options, it also fails to capture the nuance of the respondent’s opinion.
Alternatives to the Likert scale
If this does not apply to your study, try using one of the following for your survey
- Slider scale
- Semantic Differential Scale
- Side-by-Side Matrix
The last word
Maybe your research does not need a Likert scale in its survey. Let us take care of your dissertation if you are still confused about which survey type to use and how to phrase your queries.
