Data security in digital accounting: A logistic regression analysis of risk factors

This study examines the impact of cybersecurity measures on preventing data breaches in Jordanian organizations using digital accounting systems. A logistic regression model analyzes survey data from 231 organizations to assess the effects of employee training, firewall protection, two-factor authentication, and system update frequency on data breach occurrence. The Receiver Operating Characteristic (ROC) curve evaluates the predictive accuracy of the model. The findings indicate that system update frequency is the most effective factor in reducing breaches, while employee training, firewall protection, and two-factor authentication exhibit weaker, statistically non-significant effects. The ROC curve analysis shows poor predictive accuracy (AUC = 0.44), highlighting the need for additional variables to improve the model’s performance. The study concludes that frequent system updates play a crucial role in enhancing data security, whereas other measures alone provide limited protection. A holistic approach integrating multiple security practices is essential for mitigating data breach risks. Organizations should prioritize regular system updates while incorporating employee training, firewalls, and two-factor authentication into a multi-layered security strategy. Additionally, policymakers must strengthen cybersecurity frameworks tailored to the specific challenges faced by Jordanian organizations.