The analysis of contingency tables is fundamental in applied statistics to evaluate relationships between categorical variables. However, conventional methods such as Chi-square and Fisher’s exact test have limitations in tables with small sample sizes or low values, affecting statistical validity. This study proposes a linear scaling method, transforming contingency table values to an expanded range of 0–100 while maintaining original proportions. This technique enhances the sensitivity of statistical tests, allowing the detection of significant patterns that might be overlooked in smaller scales. The linear scaling was applied to simulated contingency tables reflecting clinical contexts, such as relationships between health problems and medications administered in ICUs. The original and scaled tables were analyzed using Chi-square and Fisher’s exact test, showing that scaling amplifies relative differences between cells, improving statistical interpretation. Additionally, simple correspondence analysis confirmed that structural relationships remain unchanged after scaling. The proposed method addresses challenges such as uncertainty and limited precision in diffuse data. With applications in various fields, including public health and economics, the linear scaling approach enables more robust and adaptable analysis, optimizing the use of complex data without compromising analytical validity.