Software Estimation & Measurement Issues

本文的研究假设是：在高技术产品的制造过程中，软件开发过程与传统的工程生产过程具有一些共同的特点。例如，在不同的产出范围内，绩效可能会有所不同：它可能在某些范围内显示规模经济，也可能在其他产出范围内显示规模不经济。为了在特定制造设计的一定范围内克服这种规模不经济，生产工程师设计了另一种具有更大能力的生产工艺，以处理更大量的产出。为了对提出的理论和相关的软件评估假设进行实证调查，需要访问大量的软件项目，这些项目不仅必须具有特定的变量，而且还必须具有每个组织特定的项目规模分类的信息。统计研究将使用来自两个大型软件项目存储库的数据进行。**一旦获得关于规模经济的存在与否的研究见解，我们将重新审视跨规模范围、项目范围和跨组织的因果因素和相关的生产力影响。将对项目前、项目中和项目后的情况进行研究评估。**

The research hypothesis is the following: the software development process shares some characteristics with traditional engineering production processes implemented in the manufacturing of high technology products. For example, performance can vary across ranges of outputs: it may display economies of scale within some range, and it may display diseconomies of scale across other ranges of outputs. To overcome such diseconomies of scale within some range of a specific manufacturing design, production engineers design another production process with a larger capability to handle larger volumes of outputs. To investigate empirically the proposed theory and related hypotheses for software estimation purposes, requires an access to a large set of software projects for which not only specific variables must be available, but as well information about the projects size classification specific to each organization. Statistical studies will be carried on with data from two large repositories of software projects.**Once research insights obtained on the presence-absence of economies-diseconomies of scale, we will revisit causal factors and related productivity impacts across size ranges, project scope and across organizations. Research will be conducted for estimation for pre-project, in-project and post-project contexts.**