Reason For Using Frequency And Percentage To Analyse Categorical Data

Abstract

Data analysis in social science and information management research depends heavily on matching statistical techniques with the correct level of variable measurement. Categorical data, comprising nominal and ordinal scales such as demographic profiles, information access levels, and Likert-scale responses, represent qualitative classifications that lack equal intervals or a true arithmetic zero. Consequently, applying continuous descriptive statistics, such as means and standard deviations, to categorical responses distorts empirical findings because these tools assume numeric order and equal spacing that categorical data do not possess. This paper examines the fundamental reasons for using frequency distributions and percentages as primary descriptive statistics for categorical data. Drawing upon core data processing principles such as editing, classification, and tabulation, the study demonstrates how frequencies summarise raw observation counts while percentages standardise these counts into comparable relative proportions across uneven groups. Furthermore, the paper provides a methodological rationale for applying baseline benchmarks, such as a 50% majority threshold, to evaluate responses gathered from both non-Likert and Likert-derived instruments. The article emphasises that while frequency and percentage are essential first-line tools that simplify complex datasets and improve transparency, they remain descriptive and do not automatically establish statistical significance. Ultimately, utilising frequency and percentage analyses preserves the mathematical integrity of categorical variables, offering a clear, valid, and accessible foundation for evidence-based decision-making in academic and administrative research.