Workload estimation in the absence of historical data

Abstract

This project began with a kick-off meeting and regular meetings thereafter. We developed a project plan that included a business case; a risk assessment; a list of high-level activities with predecessors and ending dates; a communications plan; a file layout; a roles, responsibilities, and matrix with contact information; a professional responsibility section that listed training dates and training material; and high-level flow charts that document the new processes. Estimating the workload imposed by the US Department of Treasury mandate presented a challenge. The US Department of Treasury (Treasury) mandates that agencies attach the component Treasury Account Symbol (TAS) code to all intra-governmental financial transactions. The US Government Printing Office (GPO) has obligating documents for all their trading partners' contracts. Prior to Fiscal Year 2013, none of the obligating documents requests the component TAS code. Nor do we have the component TAS code stored and retrieved in the system of record. Therefore, estimating the workload imposed by Treasury presents a challenge. This paper will present a Bernoulli random variable as the basis for estimation. To find the additional keying workload per form, we can either use the normal approximation for the binomial distribution or calculate the expected keying workload directly from the binomial. Both approaches will be presented each with a numeric example. For this project, both the binomial distribution and the normal approximation estimators give the same estimates for the workload. Therefore, to save time and simplify the calculations, it is recommended to use the normal approximation for the binomial distribution.