What are PERT Estimates?
本文质疑PERT网络参数估计的本质，结合帕金森定律、目标管理及历史基础，提出一种反映实际行为的底层分布及其实用近似，可用于风险分析。
This paper questions what network parameter-estimates really are. This follows recent papers on Parkinson's law, on the relationship with Management by Objectives, and on the historical basis on which these estimates are made. An underlying distribution is proposed that follows from these points—both a general shape and a pragmatic approximation—which can be shown to reflect behaviour found in practice, and has indeed been used in practical risk analysis.