Improve Phase

Data stand at the forefront of any statistical decision making. Improving the quality of data is as good as improving overall quality of the process. This is ascertained through a clearly stated operating procedure of the process and knowledge of key performance indicators that work for the process. The techniques of making data error free, systematizing the information flow and in the process, reducing wastes, etc., will be the primary objective of data quality. Many valuable tools for improving the data quality are presented in this chapter. The practical importance of designed experiments and robust designs are thoroughly examined here. The three principles of experimentation, namely replication, randomization, blocking, or local control, help the experimenter to decide the appropriate data that need to be collected, and the design to be used for finding the optimum level of the process. The necessity of complete, partial fractional designs, and robust design are also stressed here. It is expected that for all inferential studies, the data should follow a normal distribution. In the absence of this, the inferences drawn from the data may not be reliable. This problem is explained through the use of normalization, standardization, and stabilization techniques for a processed data. Another common problem seen in statistical research and management is to suggest a best model for a given data. Among the choice of many competing models, how to decide the best is even more crucial for researchers. This is addressed through three methods in this chapter, and they include a parametric, nonparametric, and the simulation techniques.