Optimum Software Process Improvement Paradigm for Quality Practices in Software Industry

Abstract

Overall behaviour of local software industry towards quality if simply phrased is, “No Quality Culture”. It is the cause of lower Information Technology (IT) exports due to non competitive nature of local software product development practices which are laden with delays, non- conformances and inconsistency. Local quality culture lacks quality awareness and is immature in following good quality practices and implementing quality improvement standards. As a step further in this direction the objective of this study is to map the actual environment and true culture of Small and Medium Software Houses (SMSH) towards quality improvement and process improvement by implementing Total Quality Management (TQM) philosophy. It was an exploratory research effort in the domain of Total Quality Management (TQM) and Software Process Improvement (SPI). The research begins with literature review of major quality standards implemented in the local industry. The behaviour of international quality standards was deliberated towards SMSH. A survey was conducted to evaluate the current quality practices and develop a process improvement model within the local SMSH. For this purpose software houses that were members of statutory and professional organizations such as Pakistan Software Export Board (PSEB), Pakistan Software House Association (PASHA) were selected. Listing of commercially available directory of RozeePak was also referred. For this survey the quality constructs and data collection instrument were designed based on literature review about small and medium enterprises culture and leading software quality models such as CMM, CMMI, ISO, SPICE and PSP/TSP. The results of the survey were analyzed and reported to high light quality problems being faced by SMSH to implement quality. Study included descriptive as well as empirical analysis. Descriptive analysis was based on comments via survey and personal interaction while conducting the survey. The empirical analysis included correlation and regression analysis of quality constructs. Structural Equation Modelling (SEM) technique was used to develop an optimized Lean Quality Improvement Model (LQIM) for standard quality practices in the local software industry. Eight quality constructs were developed to ascertain the level of current quality practices in the SMSH and evolve a LQIM. In correlation analysis all seven independent constructs were found significant towards the dependent variable Quality Improvement. Regression analysis revealed that only four of these independent quality constructs contributed significantly towards the dependant variable Quality Improvement. Through Structural Equation Modelling (SEM) the LQIM was evolved. This model presented four quality constructs and ten of their respective quality practices as significant. LQIM evolved as a tailored and economized paradigm according to the needs and perceptions of the local IT practitioners. Also LQIM evolved as an indigenous model which when improvised in accordance to the SMSH cultural and quality improvement recommendations is proven to be a fit model for SMSH. The LQIM has already been ratified according to generally accepted good fit indices in SEM analysis. In order to implement LQIM by SMSH implementation of Indigenous LQIM was proposed using the Deming’s philosophy of Plan, Do, Check, Act (PDCA) Cycle for continuous process improvement. The set of recommendations for SMSH software process improvement and proposed LQIM paradigm will give the innovative and flexible directions for SMSH to change their culture and improve their processes and software quality.