This thesis is a partial analysis of the Naval Space Command statistical orbit determination algorithms. Through a process called Differential Correction, data from space surveillance radar observation stations is synthesized with previously accumulated element sets to maintain accurate orbital object position information. Differential Correction is a nonlinear least squares process employing statistical techniques to minimize the residual measurement error thereby increasing relative position information accuracy. This study focuses specifically on the algorithmic methods of solution to the systems of normal equations generated by the Differential Correction process. A comparison and analysis of the present Naval Space Command method and the singular value decomposition method is presented. Algorithmic constructions are presented for both methods and problematic areas are highlighted. The principal focus herein is to demonstrate the benefit of singular value decomposition when attempting to solve systems of equations whose coefficient matrices are dense and nearly singular. These results generalize to commonly employed normal equation solution algorithms and are intended for further study and possible incorporation by Naval Space Command as part of future modernization plans.