Analysis of Incomplete Software Project Datasets Using Structural Equation Modeling

摘要

Analyzing Incomplete Software Project Datasets is often difficult because of their characteristics to have significant amounts of incomplete data. Seldom do researchers find satisfactory inferences from such data sets. Experimental analyses of such datasets have shown contradictory conclusions. In this paper we present Structural Equation Models (SEM) for dealing with incomplete data. We examine the following questions to the best of our knowledge: "What is the concept of structural equation model with incomplete data? ", "Can incomplete data models actually work? ", "When will such kinds of incomplete data models fail?", and "How can these models be useful in drawing conclusions?" To answer these questions we examine the results obtained for the same models with and without different subjects and variables. These sensitivity analyses display critical aspects of the available data, and illustrate vital information about fitting models with incomplete data.