Anova Parametric Tests

ANOVA and Nonparametric Tests

Praxidike Systems develops software applications for their clients.  ,but are having problems with quality. The ability of the company to deliver software on time, and to the quality demanded but customers is slowly disappearing.  This executive summary will discuss potential solutions and tools that Praxidike could use to determine what is causing the problems Praxidike is experiencing, and to potentially reverse the negative trends.

In the first part of the simulation, a Kruskal-Wallis test was used to check whether different levels of training improved competency of engineers. The null hypothesis was that it didn’t, but the test caused the null hypothesis to be rejected, proving a different.  The data showed that software engineers that fell in the medium competency level were producing at the highest level of productivity and that software engineers at both the highest and lowest competency levels were less productive.  The data suggests that while providing training to the lower level engineers would yield benefit in raising them into a more productive bracket, the same could not be said for high competency engineers. Another challenge Praxidike Systems examined was through the analysis of the types of projects and how efficient engineers were on each type.  The test showed that the scope and complexity of the project were both factors in how well engineers of each level performed. In order to maximize productivity, it was suggested that not only should engineers be assigned by skill, but management should do their best to account for scope changes as soon as possible.

The last challenge for Praxidike Systems was their customer perceptions. An ANOVA test shows that the number of defects has a strong correlation to the customer’s perception of Praxidike systems, along with schedule variance and interaction. The natural suggestions then are to implement a system to track defects, along with plans and effort estimations and...