A Structural Equation Model for Understanding the Use of Cloud by Software Development Teams

Abstract

The primary research of this thesis is a technology adoption study which particularly focuses on cloud technologies in software development activities. Structural Equation Modelling (SEM) is used to test the conceptual model which is designed with inclusion of novel suggestions to theories in the existing literature. Before this main primary research, a detailed systematic literature review (SLR) is presented on cloud computing studies that use SEM as the statistical analysis method. This SLR summarises the current state of literature by analysing previous studies and methodically categorises them. In the scope of SLR, 96 cloud computing studies from 2009 to June 2018 that employed SEM obtained from four online databases are selected and relevant data are extracted to answer the research questions. A trend of increasing SEM usage over years in cloud studies is observed, where technology adoption studies are found to be more common than the use studies. Articles appear under four main domains; namely business, personal use, education and healthcare. Technology Acceptance Model (TAM) is found to be the most commonly used theory. Adoption, intention to use, and actual usage are the most commonly selected dependent variables in SEM models whereas security & privacy concerns, costs, ease of use, risks, and usefulness are the most common selections for causal factors. Technology adoption studies are, in essence, social behaviour studies that aim to understand effects on behavioural intention to adopt or use innovations. In this thesis, a hybrid model is designed and tested in order to understand the changes in software developers’ intention to adopt cloud for their software development activities. This hybrid model consists of factors taken from TAM and Technology-Organisation-Environment (TOE) theories. In addition to these, novel suggestions are included in the model under the name of Personal-Organisation-Project (POP) structure. The reason for the novel suggestions is that the characteristics of projects are predicted to affect the project team members’ intention to use cloud in their work even in the same organisation. Upon completing the initial statistical analyses, 15 of the 21 initial hypotheses in the model are accepted with high significance levels. Then the rejected hypotheses are removed and the model is modified with exploratory SEM analyses and a revised final model is reached. 16 new regression relationships between variables are discovered. Discovered effects are discussed and ensured to be logically consistent and reasonable in real life systems. In the end, the hybrid model itself is validated as a technology adoption model and specific conclusions are drawn for developers’ intention to use cloud in software development organisations in Turkey.