Anova

Running Head: ANOVA

ANOVA and Nonparametric Tests Simulation

RES/342

December 5, 2010

Terrance C. Feravich

ANOVA and Nonparametric Tests Simulation

The purpose of this simulation is to decide which type of test should be used for analysis, the Kruskal-Wallis Test or the ANOVA. The simulation involved analyzing variables such as productivity and customer satisfaction.

The three lessons I learned relative to ANOVA and Nonparametric Testing are that prior to using either ANOVA or the Kruskal-Wallis test, I must ensure that all the assumptions have been met. For example, there are three assumptions for ANOVA and they are a. The errors are random and independent of each other; b. Each population has a normal distribution; and c. All population has the same variable. Secondly, if there is no information about the distribution on the population sample, one should not assume that the population has a normal distribution; it is recommended to use Kruskal-Wallis test to compare the data for multiple populations. Thirdly, the interpretation of Kruskal-Wallis test is similar to that of the parametric one way ANOVA however; nonparametric tests are usually easier to calculate.

This simulation has taught me the tools and concepts of taking suggestions from others within an organization in order to perform the tasks at hand. Suggestions that were made, as a manager, he or she should be able to make decisions in the areas that require improvements. At my workplace, productivity level is low because the level of competency among the employees is low due to language barrier and also training in the position they hold. Proper training sessions were not implemented prior to their date of employment. They were just shown how to do the job but not explained why. I feel with proper communication and training, the level of productivity will increase and thus sales and delivery time will also improve.

ANOVA and Nonparametric Tests...