A Likert scale is a one-dimensional scale that researchers use to collect respondents’ attitudes and opinions. Researchers often use this psychometric scale to know the opinions and perspectives towards a brand, a product or a target market.
There are different variations of the Likert scales that focus directly on measuring people’s opinions, such as the Guttman scale, the Bogardus scale, and the Thurstone scale. Psychologist Rensis Likert drew a distinction between a scale that materializes from a collection of responses to a group of items (perhaps 8 or more). Responses are measured in a range of values.
Types of Likert scales
The Likert scale has become a favorite of researchers to collect opinions about customer satisfaction or employee experience. This scale can be divided mainly into two main types:
Likert par scale
Researchers use peer Likert scales to collect extreme opinions without offering a neutral option.
4-point Likert scale for importance
This type of Likert scale allows researchers to include four extreme options without one neutral option. Here the different degrees of importance are represented on a 4-point Likert scale.
8-point Likert scale for recommendation
This is a variation of the 4-point Likert scale explained above, with the only difference being that this scale has eight options for collecting opinions on the likelihood of a recommendation.
Odd Likert scale
Researchers use the odd Likert scale to give respondents the option to respond neutrally.
5-point Likert scale
With five answer options, researchers use this odd-scale Likert question to gather information on a topic, including a neutral answer option for respondents to select if they don’t want to answer among the extreme options.
7-point Likert scale
The 7-point Likert scale adds two more answer options at the ends of a 5-point Likert scale question.
9-point Likert scale
The 9-point Likert scale is quite uncommon, but can be used by adding two more answer options to the 7-point Likert scale question.
Characteristics of the Likert scale
The Likert scale emerged in 1932 in the form of a 5-point scale, which is widely used today. These scales range from a group of general topics to the more specific ones that ask respondents to indicate their level of agreement, approval, or belief. Some significant features of the Likert scale are:
Questions should be easily related to sentence responses, regardless of whether the relationship between the item and the phrase is evident.
Items should always have two extreme positions and an intermediate response option that serves as a graduation between the extremes.
Number of response options
It is essential to mention that although the most common Likert scale is that of 5 items, the use of more items helps to generate greater accuracy in the results.
Increased reliability of scale
Researchers typically increase the ends of the scale to create a seven-point scale by adding “very” to the top and bottom of the five-point scales. The seven-point scale reaches the upper limits of the scale’s reliability.
Use wide scales
As a general rule, Likert and others recommend that it is better to use as wide a scale as possible. You can always collapse responses into concise groups, if appropriate, for analysis.
Lack of a neutral option
Considering these details, scales are sometimes narrowed down to an even number of categories (usually four) to eliminate the “neutral” possibility on a “forced choice” polling scale.
The primary Likert record clearly states that there may be an intrinsic variable whose value marks respondents’ reactions or attitudes, and this underlying variable is the interval level, at best.
Advantages of The Likert Scale
There are many advantages to using a Likert scale in a survey for market research. They are as follows:
Ease of application
This universally accepted scale can be easily understood and applied to various customer or employee satisfaction surveys.
Quantifiable response options
Quantify the Likert items with no apparent relation to the expression and perform a statistical analysis of the results received.
Analyze the range of opinions
There may be a sample with varied opinions on a particular topic. The Likert scale offers a ranking of the opinions of these people surveyed.
Simplicity of response
Respondents can understand the intent of this scale and respond quickly to the question.
Likert scale data and analysis
Researchers regularly use surveys to measure and analyze the quality of products or services. The Likert scale is a standard classification format for studies. Respondents give their opinion (data) on the quality of a product/service from high to low or from best to worst using two, four, five or seven levels.
Researchers and auditors usually group the collected data into a hierarchy of four fundamental measurement levels: nominal, ordinal, interval and ratio levels for further analysis:
Data in which responses classified into variables do not have to have a quantitative data or order are called nominal data.
Data in which it is possible to sort or classify responses, but it is impossible to measure distance are called ordinal data.
Aggregated data on which order and distance measurements can be made are called interval data.
Reason data is similar to interval data. The only difference is an equal and definitive relationship between each piece of data and the absolute “zero” that is treated as the point of origin.
Data analysis using nominal, interval and ratio data is generally transparent and simple. Ordinal data analyzes the data, especially as it relates to Likert or other scales in surveys. This problem is not new. The efficacy of the treatment of ordinal data as interval data remains debatable in the analysis of surveys of various applied fields. Some of the significant points to keep in mind are
Researchers sometimes treat ordinal data as interval data because they claim that parametric statistical tests are more powerful than nonparametric alternatives. In addition, inferences from parametric tests are easy to interpret and provide more information than nonparametric options.
Concentration on Likert scales
However, treating ordinal data as interval data without examining dataset values and analysis objectives can mislead and misrepresent the results of a survey. To analyze scalar data more appropriately, researchers prefer to consider ordinal data as interval data and concentrate on Likert scales.
Median or range to inspect data
A universal pattern suggests that mean and standard deviation are unfounded parameters for detailed statistics when data are on ordinal scales, just like any parametric analysis based on normal distribution. The nonparametric test is performed based on the appropriate median or range to inspect the data.
Practical Application of the Likert Scale
For example, to collect feedback on a product, the researcher uses a Likert scale question in the form of a dichotomous choice question. The question is formulated as “The product has been a good purchase”, with the options of agreement or disagreement. The other way to frame this question is, “Please indicate your level of satisfaction with the products,” and the options range from very dissatisfied to very satisfied.
When responding to an item on the Likert scale, the user responds explicitly based on their level of agreement or disagreement. These scales allow to determine the level of agreement or disagreement of the respondents. The Likert scale assumes that the strength and intensity of the experience are linear. Therefore, it goes from total agreement to total disagreement, assuming that attitudes can be measured.
Best practices for analyzing the results of Likert scales
Because the Likert element data is discrete, ordinal, and limited in scope, there has been a long dispute over the most logical way to analyze Likert data. The first choice is between parametric and non-parametric tests. The advantages and disadvantages of each type of analysis are generally described as follows:
Parametric tests involve a regular and uninterrupted division.
Non-parametric tests do not involve a regular and uninterrupted division. However, there is concern of a reduced ability to detect a difference when it exists.
What is the best option? This is a real decision that a researcher must make when he decides to analyze the information received from a survey that uses questions of the Likert scale.
Over the years, a number of studies have attempted to answer this question. However, they have been inclined to examine a limited number of potential distributions for the Likert data, which makes the generalization of the results suffer. Thanks to increased computing power, simulation studies can now thoroughly evaluate a wide range of distributions.
The researchers identified a diverse set of 14 distributions that are representative of actual Likert data. The computer program extracted pairs of self-sufficient samples to test all possible combinations of the 14 distributions.
In total, 10,000 random samples were generated for each of the 98 distribution combinations. The sample pairs were analyzed using the two-sample t-test and the Mann-Whitney test to compare the efficacy of each test. The study also evaluated different sample sizes.
What were the results?
The results show that the type I error rates (false positives) of all distribution pairs are very close to the target amounts. If an organization uses any of the tests and the results are statistically significant, it doesn’t have to worry too much about a false positive.
The results also show that, for most pairs of distributions, the difference between the power of the two tests is trivial. If there is a difference at the population level, either analysis is just as likely to detect it.
There are some pairs of specific distributions in which there is a power difference between the two tests. If an organization performs both tests with the same data and does not match (one is significant and the other is not), this power difference only affects a small minority of cases.
In general, the choice between the two analyses is a loop. If an organization needs to compare two five-point Likert data sets, the analysis method doesn’t usually matter.
Both parametric and non-parametric tests systematically provide the same security against false negatives and also offer the same protection against false positives. These patterns are valid for sample sizes of 10, 30, and 200 per group.
Our specialists wait for you to contact them through the quote form or direct chat. We also have confidential communication channels such as WhatsApp and Messenger. And if you want to be aware of our innovative services and the different advantages of hiring us, follow us on Facebook, Instagram or Twitter.
If this article was to your liking, do not forget to share it on your social networks.
You may also be interested in: Best Civil Engineering Thesis Topics: A List of Unique Ideas
Bowling, A. (1997). Research Methods in Health. Buckingham: Open University Press.
Burns, N., & Grove, S. K. (1997). The Practice of Nursing Research Conduct, Critique, & Utilization. Philadelphia: W.B. Saunders and Co.
Jamieson, S. (2004). Likert scales: how to (ab) use them. Medical Education, 38(12), 1217-1218.
Likert, R. (1932). A Technique for the Measurement of Attitudes. Archives of Psychology, 140, 1–55.
Paulhus, D. L. (1984). Two-component models of socially desirable responding. Journal of personality and social psychology, 46(3), 598.