Estimating Learning Curves from Aggregate Monthly Data
研究如何用包含固定和可变工时的月度汇总数据估计学习曲线，提出分布式滞后模型处理在制品波动，并用政府合同数据检验模型在分析生产中断影响时的效果。
In this paper the problems of using aggregate monthly data to estimate learning curves are investigated. Here, aggregate monthly data on labor hours are assumed to contain some of both fixed and variable labor hours. They are also assumed to be influenced by fluctuating quantities of work in process. A distributed lag model is developed to deal with these two characteristics of aggregate monthly data. The model is generalized to permit production rate to influence labor productivity. This generalized model is then estimated and compared to a cumulative average learning curve in analyzing the impact of a production break. A set of production data which arose from a government contract claim is used for this purpose.