统计方法与试验在建设工程质量管理中应用商榷

On Quality Management of Construction Projects: Statistical Methods and the Trial

产品质量控制图是工业产品生产中进行生产过程质量控制的重要工具,这一工具被众多建设工程质量管理文献所引用。质量控制图的理论模型是无限重复进行的同一个可重复随机试验,只有当产品生产过程符合或近似符合这一模型时,控制图的应用才是有效的。建设项目施工过程中大多数工作并不符合上述模型,对此必须强调和引起注意。分析了可重复随机试验两种类型的结果,即对同一个对象的重复测量值和对同一类不同个体的测量值,以及当后者独立于随机试验而存在时应用统计推断方法的若干实际问题。

The product quality control chart is one of the most frequently used tools in the quality control process. And it has been quoted heavily by the related literature of quality management of construction projects. This paper points out that the theoretical model for the quality control chart only works when the process can be perceived as an infinite repeatable random experiment （or roughly so）. However, the real process of undertaking any construction projects could hardly be seen as compatible to the theoretical one, and this should be addressed and be given due attention to. Two types of the results from repeatable random experiment are discussed in this paper, i.e., one concerning repeated measurements of an object and the other about measuring different items of the same category. The latter could be independent of any random experiment and special attention should be given to the statistical inferential method to deal with such a case.