ESTIMATION PROBLEMS IN RATE-AUGMENTED LEARNING-CURVES

In situations where production rate is variable, the learning curve ca n provide unreliable results. One proposed solution to the ''productio n rate'' problem is to multiplicatively augment the learning curve wit h a production rate explanatory variable, While widely used by cost an alysts, the rate-augmented learning curve has not proved reliable. It has been assumed, but not demonstrated, that data and measurement prob lems lead to unreliable parameter estimates, In this paper we demonstr ate that the parameter estimates are a function of the units in which the cost and delivery data are measured; hence, the estimates are alwa ys arbitrary, We propose a procedure for obtaining meaningful estimate s, but it requires assigning weights to the cost and delivery time dat a, We use a simple dynamic program to demonstrate that proper estimate s of sign and magnitude can only be obtained when the learning and rat e equations are estimated simultaneously and the sum-of-squares for th e rate equation receives a higher weighting in the estimation, These r esults have major implications for cost analysts, Since accurate param eter estimation requires weighting, estimation by unweighted ordinary least squares is a futile effort.