Reason for using frequency and percentage to analyze categorical data

Abstract

Categorical data are commonly generated in social science, education, management, library and information science, and other fields of research. Such data classify observations into groups or categories, such as gender, educational qualification, occupation, age groups, responses to survey questions, and levels of agreement. Frequency and percentage are among the most appropriate descriptive statistical techniques for summarizing and interpreting categorical data. Frequency indicates the number of respondents who selected a particular category, while percentage expresses that frequency relative to the total number of valid observations. This article discusses the major reasons for using frequency and percentage in the analysis of categorical data. It also examines their usefulness in analyzing responses obtained from both non-Likert categorical questions and Likert-type scales. Furthermore, the article discusses the use of a 50% benchmark as a criterion for interpreting research findings. The article argues that frequency and percentage make research findings simple, transparent, comparable, and easy to interpret. For Likert-type data, frequency and percentage can show the distribution and intensity of respondents' opinions without necessarily imposing assumptions about equal intervals between response categories. However, the 50% benchmark should be clearly justified by the researcher and should not be presented as a universal statistical requirement.