Reason for Using Frequency and Percentage to Analyse Categorical Data

Researchers frequently encounter categorical data drawn from demographic variables and questionnaire items, and the choice of an appropriate statistical technique to analyse such data is central to the credibility of any descriptive study. This article examines the rationale for using frequency and percentage as the preferred descriptive statistics for analysing categorical data, irrespective of the academic degree or level at which the research is conducted and further explains the justification for adopting a 50% benchmark when interpreting data obtained from both Likert and non-Likert scale instruments. Drawing on established literature in research methodology and biostatistics, the article argues that frequency and percentage are suited to categorical data because such data cannot be meaningfully summarised using measures that assume numerical distance or order, such as the mean or standard deviation. It further argues that a 50% benchmark provides a logical, population-independent reference point for judging whether a response pattern is high or low, since it represents the midpoint of total possible response distribution. The article concludes with practical recommendations for students and researchers undertaking descriptive survey research at undergraduate, postgraduate, and doctoral levels.