In the "Applying ANOVA and Nonparametric Tests" simulation, the Praxidike Systems Corporation is a software firm that develops and provides enterprise-wide software solutions to a wide range of clientele. Praxidike Systems has given me the opportunity as the Quality Assurance Manager to monitor on-time delivery of its projects and client satisfaction over a four month time period. In the simulation I had to apply ANOVA in addition to the nonparametric test, Kruskal-Wallis, to determine the causes of delay in project delivery and what can be done to eliminate defects in the actual product.
As I completed the simulation the lessons I learned relative to ANOVA and nonparametric tests include:In order to make the correct decision I first had to decide which type of test to use. The Kruskal-Wallis test is used to compare data for more than two populations and is used to analyze data which does not have a normal distribution, and/or is ordinal and not quantitative.
To use ANOVA there are three assumptions: Errors are random and independent of each other. Each population has a normal distribution and all of the populations have the same variance (University of Phoenix, 2009).
I also learned that when considering an ANOVA test the one-way analysis is just that, it only looks at one variable such as employee competency. A two-way ANOVA analysis is what compares two variables such as client satisfaction and employee competency. Accurately identifying whether to use a one-way analysis or a two-way analysis is a critical step in performing the tests.
Lastly, I learned to consider the correct variables, I made a couple of poor decisions in the simulation because I failed the correct variable.
By completing this simulation, I believe that I would be able to use the two-way ANOVA to determine the effectiveness of marketing materials...