Non-parametric variables measured using the Likert scale (ordinal scale) have a slightly different analysis approach compared to parametric variables. This Likert scale approach is often used in social and management research. In such studies, we are often confronted with variables that cannot be directly quantified.
For example, when measuring employee motivation, we cannot directly obtain numerical values for analysis. Therefore, we need an instrument to measure it, one of which is the Likert scale.
Generally, these variables do not follow a normal distribution and are classified as non-parametric variables. In this article, Kanda Data will discuss in detail how to analyze non-parametric variables, specifically those measured using the Likert scale.
Definition of Non-Parametric Variables
Before diving deep into the use of the Likert scale in research, we first need to understand what non-parametric variables are. Non-parametric variables are those that do not require certain distribution assumptions and are usually measured using nominal or ordinal scales.
Non-parametric variables are often found in social research, where the variables are measured using nominal or ordinal scales. An example of a non-parametric variable is customer satisfaction levels (very satisfied, satisfied, neutral, dissatisfied, very dissatisfied), which are typically expressed on an ordinal scale.
As mentioned earlier in the introduction, non-parametric variables do not directly provide numerical values that can be processed. For instance, in a satisfaction survey, the response “very satisfied” does not represent a specific numerical value, nor does the next response, “satisfied.”
Therefore, in data processing, ordinal variables measured using the Likert scale require an instrument to convert respondent answers like “very satisfied” and “satisfied” into scores (numerical values), allowing for further analysis. The converted non-parametric variable scores can be analyzed using non-parametric methods such as the chi-square test, Kruskal-Wallis test, or Mann-Whitney U test, depending on the research objectives.
Case Study Example of Likert Scale Variables
To help readers better understand how to analyze variables using the Likert scale, I will illustrate a case study. A researcher is conducting a study with companies in the industrial region of Province XYZ as the unit of analysis.
The researcher collects a sample of 48 companies from the industrial area. The researcher’s goal is to investigate the relationship between employee motivation and employee performance in companies located in Province XYZ.
Based on the researcher’s initial review, both employee motivation and employee performance are subjective variables and are difficult to measure directly. Therefore, the researcher designs an instrument to measure employee motivation and performance using a more structured approach. Drawing from previous literature, the researcher decides to use the Likert scale.
To measure variables such as motivation and performance, the researcher uses the Likert scale as an instrument to systematically assess individual perceptions or responses. Respondents are asked to rate a series of statements, ranging from “Strongly Agree” to “Strongly Disagree.”
Basic Theory of the Likert Scale
The Likert scale was developed by Rensis Likert in 1932 as a method for measuring attitudes. This Likert scale approach has developed rapidly and is widely used in social and management research today. Essentially, measuring variables using the Likert scale is similar to measuring variables using the ordinal scale.
In a Likert scale instrument, respondents are asked to indicate how much they agree or disagree with a series of statements related to a topic. For example, in the case study, the variables are employee motivation and performance.
The Likert scale typically has five response options, with a score of 1 for “Strongly Disagree,” 2 for “Disagree,” 3 for “Neutral,” 4 for “Agree,” and 5 for “Strongly Agree.”
Creating a Likert Scale Instrument
After understanding the basic theory behind using the Likert scale, we need to thoroughly understand how to create a Likert scale instrument. Based on the case study example, we will create a Likert scale instrument for employee motivation and performance variables.
The first step in measuring these two variables is determining the indicators that represent employee motivation and performance using relevant theories and previous research.
From these indicators, construct statement items for each indicator. The number of statements is not limited, but it should correspond to the existing indicators. Each item must undergo validity and reliability testing before being used as an instrument for data collection in the field.
Here are examples of Likert scale instruments that can be used to measure employee motivation: (1) I feel motivated to complete tasks on time (Strongly Disagree, Disagree, Neutral, Agree, Strongly Agree). (2) I have a strong drive to achieve the set targets (Strongly Disagree, Disagree, Neutral, Agree, Strongly Agree). (3) Additional statements representing the employee motivation variable can be created.
Next, examples of Likert scale instruments that can be used to measure employee performance: (1) I often complete tasks before the deadline (Strongly Disagree, Disagree, Neutral, Agree, Strongly Agree). (2) I always achieve the performance targets set by the company (Strongly Disagree, Disagree, Neutral, Agree, Strongly Agree). (3) Additional statements representing the employee performance variable can be created.
As previously explained, to construct a Likert scale, we must first determine the relevant indicators for the variable being measured. These indicators should cover key aspects of the variable. For example, when measuring employee motivation, indicators could include “internal motivation,” “personal goals,” and “drive for achievement.” Each indicator is then translated into statements that can be answered using the Likert scale.
Difference Between Favourable and Unfavourable Statements in Likert Scale
When creating Likert scale statements, we must understand the difference between favourable and unfavourable statements. In the Likert scale, statements can be either favourable (positive) or unfavourable (negative). Favourable statements, when responded to positively, reflect a high value on the measured variable, while unfavourable statements reflect the opposite.
Let’s look at an example to better understand. A favourable statement might be: “I am satisfied with my job,” while an unfavourable statement might be: “I often feel unmotivated to work.”
Hopefully, up to this point, we have gained a better understanding of how to create a Likert scale instrument. Next, I will discuss Likert scale scoring and tabulation techniques. The analysis of validity and reliability for Likert scale instruments will be covered in a future article.
Likert Scale Scoring and Tabulation Techniques
Once the research is assumed to be completed, the next step is scoring and tabulating the data. Essentially, the scoring technique is done by assigning scores to each response provided by respondents.
Each response option is given a numerical value (for example, 1 for “Strongly Disagree,” 2 for “Disagree,” 3 for “Neutral,” 4 for “Agree,” and 5 for “Strongly Agree”). These values are then tabulated to generate a total score. For further analysis, the total score or average score of the employee motivation and performance variables can be used to observe general trends in respondent responses.
Likert Scale Variable Analysis Options
After scoring and tabulating, the next step is data analysis. In this article, I won’t delve into each test in detail, but I will provide a general overview of the tests that can be used to analyze variables measured using the Likert scale.
If we want to examine the correlation between employee motivation and employee performance measured using the Likert scale, we can use the Spearman rank correlation analysis. If we want to test the effect of employee motivation on employee performance, we might consider using ordinal logistic regression. These tests will be discussed in greater detail in future articles.
Conclusion
Non-parametric variables measured using the Likert scale are widely used in social and management research. The use of the Likert scale provides flexibility in measuring variables that are difficult to measure directly and allows researchers to identify trends in respondents’ attitudes or perceptions.
This is the article that Kanda Data can present at this time. I hope it is useful and provides new insights for everyone. Stay tuned for updates in future articles and thank you.
