Abstract
This study introduces a new practical and adequate statistical model for fitting COVID-19 data in Saudi Arabia. The alpha power Kumaraswamy-Burr III distribution (APKIII) is developed based on combining two well-known techniques, namely, the Alpha power transformation and Kumaraswamy generated family. As a consequence of the combined effect of these two techniques and the beneficial impact of the Kumaraswamy and Burr III distributions, this approach will typically offer a flexible statistical model for various applications. The APKIII is of particular interest as its density takes various symmetric and asymmetric possible shapes. In addition, the APKIII hazard function is attractive as it exhibits a wide range of asymmetrical forms, including decreasing, increasing, approximate-symmetrical, and reversed J-shaped. The theoretical features of APKIII are derived, focusing on those that may be of general relevance in probability and statistics. The performance of the APKIII in modeling data is evaluated through its application to two real-world data of current interest: the daily COVID-19 infected and recovered cases of Saudi Arabia. As a result of this novel modeling technique, APKIII delivers a better fit compared to competing distributions.