Understanding and Utilizing Likert Scales in Research

In research world, intensity of human attitudes, perceptions and opinions are of utmost importance. Getting beyond the yes/no questions allows the researcher to glimpse the gray scale of human life. This is where the Likert scale comes in as a very useful instrument. Likert scale was created in 1932 by psychologist Rensis Likert, and its use has become fundamental to market research, as well as to questioning regarding customer satisfaction and employee feedback across the social sciences. What is it, then, and how should you use it to obtain quality, actionable data? This detailed explanation will get into the Likert scale meaning, all its different forms along with Likert scale examples, and best practices in developing a strong Likert scale scale questionnaire, analysing data and making sure your Likert scale in research is valid and reliable.

Primarily, a Likert scale definition starts by explaining that it is a psychometric scale to be utilised in the survey questions of a research study to find out the attitude, agreement or perception of respondents. It does so by stating a statement and asking the respondent to say how far ones agrees or disagrees, using an agree/disagree scale.

It is important to differentiate between Likert items and a Likert scale.

A Likert item is a statement, followed by the answer choices (e.g. I am satisfied with the customer service I received.) It is a web-based survey, where each respondent is asked to answer the following questions by choosing the most appropriate option (Strongly Disagree to Strongly Agree).

Likert scale is a combination of Likert items that measure one common idea (e.g. Customer Satisfaction, using the score of questions about service, product price etc.).

This multi-item method is what makes the data more accurate and gives the freedom to do Likert scale data analysis.

One of the mostly used tools in research is the Likert scale due to its flexibility. The greatest advantage is that it can be tailored in order to fit in your objectives in the study. The Likert scale can be used when all you need is a brief picture of the level of customer satisfaction, or when you need the full-fledged academic research where the questions will be formulated.

The initial choice in the construction of Likert scale is the number of points of responses you want to provide. The two most used scales are 7 point scale and 5-point scale. Both have their advantages and the deciding factor depends on the objective of the research and the target.

This is the most well known and conventional one. It gives a middle ground thus simplifying the task of the respondents who are only required to state their views quickly without any confusion.

Example: Strongly Disagree | Disagree | Neutral | Agree | Strongly Agree

Best for: telephone surveys, general population surveys or where simplicity and quick responses are required. Being easy to comprehend, even those with less experience in filling in the surveys would not think twice about it.

The 7-point scale goes a step further and gives enough room to allow the respondent express their views clearly. It also allows distinction of minor give-and-take in opinion that would be lost in a 5-point scale.

Example: Strongly Disagree | Disagree | Somewhat Disagree | Neutral | Somewhat Agree | Agree | Strongly Agree

Best for: you need academic research, a professional study or when you need respondents to be more critical about their position. A typical example is the discretion that can be applied to the difference between the two responses of Somewhat Agree and Agree, and it may yield more information about attitudes and behaviors.

In brief, a five-point scale is quicker and simpler, whereas a seven-point one has more sensitivity and statistical power in detecting slight differences in attitudes.

The better the questions you ask, the better will be the responses. Statements that are poorly written are likely to perplex participants and resulted in unreliable information. The following are best practices of writing clear Likert scale questions:

Do not use jargon and technical terms nor double-barreled questions. Every thing ought to be easy.

Poor: The product was innovative, and timely delivery of the product.

Better: The product was innovative in features. / The product came at the right time.

All of the statements are to be able to measure a single attitude or perception Such as, instead of saying: “I am happy with the quality of the product, as well as, the price, and the packaging”, a single stem is superior.

Give the same number of positive and negative choices so as not to encourage any kind of bias. As another example, do not give three positive alternatives and a single negative alternative.

By creating a neutral mid-point, you permit the participants to remain neutral when they do not have an opinion.

Now, let us have a look at how Likert scales are applied to real surveys in the most popular fields:

Example 1: Employee Engagement Survey (5-point) Statement: I feel that as an employee I am important to this company. Response Options: Strongly Disagree → Strongly Agree

Example 2: Customer Satisfaction Survey (7-point) Statement: How do you feel about the usability of our website? Response Options: Extremely Dissatisfied → Extremely Satisfied

Example 3: Academic Research on Climate Change (7-point) A researcher may ask the participants to comment on a number of statements concerning the same topic:

It is my belief that human activity is one of the primary reasons that contribute to climate change.”

I am concerned with the long-term effects of global warming.

I approve of the government measures that control carbon emission.

Responses are scored on a 7-point scale which is based on agreement. This can be used to aggregate the results into a single measure that reflects a generalised concern with climate issues.

Gathering answers of course is only the beginning The data that comes out of a Likert scale is precious when it comes to the analysis and interpretation.

It should be mentioned that Likert data is ordinal (ordered data), not interval. In principle, the gap in the scale between Strongly Disagree and Disagree can be different than the gap between Agree and Strangely Agree. In practice, however, investigators still tend to attach numerical scores (e.g., to 5 or to 7) and analyse them as interval data.

Descriptive Statistics – Find the mean, median and mode of each question. The mean gives a Fast Idea of the average sentiment, the mode of the number of people sharing the same sentiment or feeling.

Frequency Distribution –Establish bar charts indicating the number of individuals that were selected into each choice This assists in drawing visual images on patterns

Cross-Tabulation – The information can also be compared between or among groups (e.g. age, gender) to see whether responding is significantly different.

Inferential Statistics – Use tests like Chi-square (to test relationships between the variables) or ANOVA (to compare groups).

Example: The average score in satisfaction is high, you may think that respondents are happy. A frequency chart will perhaps show that most people would only be classified as Agree, very few as Strongly Agree. That implies that satisfaction is good, yet it can be developed.

In order to ensure credibility and confidence in your survey, do the following:

Test it with a small group so that solutions can be found on unclear wording or the technical problems.

The importance of your results will lie in respondents. When probability sampling techniques (such as random or stratified sampling) are applied fairness is guaranteed and the results can be generalised.

Ensure that only one type of font is used all through (5- or 7-point scale) to avoid confusing the respondents. Ensure that only one type of font is used all through (5- or 7-point scale) to avoid confusing the respondents.

In sensitive areas such as employee morale, be sure that the responses are anonymous to help ensure that people answer truthfully.

The strength in Likert is that, it is simple but at the same time flexible. By allowing you to compare customer approaches and attitudes in a loosely based 5-point format style research questionnaire; or to be more rigorous with academic research in a 7-point version, it has demonstrated the ability to measure attitudes and perceptions in a very systematised way.

Making the proper survey in terms of number of points, asking explicit questions, interpreting answers in proper way, and adhering to the best practices, you can transform raw survey data into meaningful information. Likert scale is still one of the most valid means of comprehending how individuals think and feel.

[post-views]