James V Monaco

Predicting Schedule Risk: A Regression Approach James V Monaco

Past research shows that schedule growth within the acquisition community is difficult to predict and often adversely affects the cost and performance characteristics within programs. This study describes the use of a two-step procedure for assessing schedule growth of major defense acquisition programs using historical data. Through the operations of both logistic and multiple regression, we seek to predict if a program will experience schedule growth and, if applicable, determine the amount of growth. Past research on cost growth within the EMD phase of acquisition shows favorable results using this two-step process. We extend the use of logistic and multiple regression to the area of schedule growth in acquisition programs. We compile programmatic data from the Selected Acquisition Reports (SARs) between the timeframe of 1990 and 2003. With this methodology, we develop a statistically significant logistic regression model that accurately predicts schedule growth for approximately 89% of the data. For programs with schedule growth, we create a multiple regression model (adjusted R2 of 0.7426) to predict the expected amount of schedule growth.