How to use the Likert scale in statistical analysis

The Likert scale is commonly used in survey research. It is often used to measure the attitudes of respondents by asking them to what extent they agree or disagree with a particular question or statement. A typical scale could be "strongly agree, agree, not sure / undecided, disagree, strongly disagree". Data from a survey using the Likert scale may seem easy to analyze, but there are important issues to consider by a data analyst .

Steps to follow:

one

Obtain the list data for analysis by coding the responses. For example, let's say you have a survey that asks respondents if they agree or disagree with a set of positions on the platform of a political party. Each position is a question of the survey, and the scale uses the following answers: totally agree, agree, neutral, disagree, totally disagree. In this example, we will encode the answers accordingly: strongly disagree = 1, disagree = 2, neutral = 3, agree = 4, strongly agree = 5.

two

Remember to differentiate between ordinal and interval data, because the two types require different analytical approaches. If the data is ordinal, we can say that one score is higher than another. We can not say how much higher, as we can with the interval data, which will tell you the distance between two points. Here is the trap with the Likert scale: many researchers will treat it as an interval scale. This assumes that the differences between each answer are equal in the distance. The truth is that the Likert scale does not tell us that . In our example, it only tells us that people with the greatest number of responses are more in agreement with the positions of the party than those with the least number of answers.

3

Begin to analyze the Likert scale data with descriptive statistics. Although it may be tempting, resist the urge to take numerical answers and calculate a mean. Adding a "strongly agree" answer (5) to two of the "disagree" answers (2) would give us an average of 4, but what is the meaning of that number? Fortunately, there are other measures of central tendency that you can use, in addition to the average. With Likert scale data, the best measure to use is the most frequent mode or response. This makes the results of the survey much easier for the analyst to interpret (not to mention the audience for his presentation or a report). You can also view the distribution of responses (percentages that agree, disagree, etc.) on a graph, such as a bar graph, with a bar for each response category.

4

Proceed next to the inference techniques that test the hypothesis put forward by the researchers. There are many methods available, and the best depends on the nature of your study and the questions you are trying to answer. A popular method is to analyze the responses using analysis of variance techniques, such as the Mann Whitney test or the Kruskal Wallis test . Suppose that in our example we wanted to analyze the answers to the questions about foreign policy positions with ethnicity as an independent variable. Let's suppose that our data includes the answers of the Anglos, African-Americans and Hispanics surveyed, so we could analyze the answers among the three groups of respondents with the Kruskal Wallis test of variance.

5

Simplify your survey data by combining the four response categories (for example, strongly agree, agree, disagree, strongly disagree) into two nominal categories, such as agreement / disagreement, accept or reject, etc. ). This offers other possibilities for analysis. The chi-square test is an approach to analyzing data in this way.

Tips
  • Remember that there are many approaches to analysis. Consider your research questions to determine the best method of analysis for your study.
  • Likert scales vary in the number of points on the scale. The five-point scale used here is the most common, but some Likert scales have 4-point response scales, where the unsafe scale is eliminated (undecided category). Some even have 7-point response scales.