REASON FOR USING FREQUENCY AND PERCENTAGE TO ANALYSE CATEGORICAL DATA

Abstract

Categorical data are widely generated in educational, social, management, health and behavioural research, particularly through questionnaires and surveys. Such data classify respondents into distinct groups or response categories, including sex, marital status, educational qualification, yes/no responses and levels of agreement on Likert scales. Frequency and percentage are among the most appropriate descriptive statistical techniques for summarising categorical data. Frequency shows the number of observations in each category, while percentage expresses the frequency as a proportion of the total valid observations. This article examines the reasons for using frequency and percentage to analyse categorical data regardless of the degree or educational level of respondents. It also discusses the use of a 50% benchmark for interpreting data collected through both non-Likert and Likert-scale instruments