Analytical Validation of NovoNumeric: An Indigenous, High-Performance Statistical Software for Indian Medical Research

Abstract

Statistical analysis is a fundamental pillar of medical research, yet many powerful software options come with a steep learning curve or prohibitive costs. NovoNumeric is a novel statistical software for macOS designed to address this by balancing accessibility with high performance. This study provides the first formal analytical validation of NovoNumeric by comparing its analytical output against established gold-standard packages. A comparative validation study was performed using an anonymized clinical dataset (n=174) from a prospective observational study on diabetic foot infections, conducted at Sree Uthradom Thirunal Academy of Medical Sciences, Thiruvananthapuram (Ethical Approval No. 22/IEC/SUTAMS/2023). Parallel analyses—including descriptive statistics, independent t-tests, chi-square tests, Pearson's correlation, and multiple linear regression—were conducted using NovoNumeric (v4.0), R (v4.3.1), and SPSS (v28). The primary outcome was the equivalency of numerical results. For every statistical test performed, NovoNumeric produced identical outputs to R and SPSS, with maximum absolute differences < 1 x 10-8. These findings confirm that NovoNumeric is a valid, reliable, and high-performance tool for core statistical analyses in medical research, demonstrating its suitability as a cost-effective solution for Indian academic and clinical research institutions.