Parametric distribution is based on a normal distribution. The sample to be analyzed should be taken from a population that meets the normality assumption. Non parametric tests are used when the assumption in the parametric test is not met or they are not right. Non parametric test always rank the variable from low to high. Then the ranks are analyzed using the right method. So the VP should use non parametric test to test data collected from the regions the employees are distributed, and also group the employees into certain groups. He does not need to analyze the mean of the data. There are several reasons why the VP should use the non parametric data. First, the outcome of the data collected from the employees is a rank, and the values are ranked from low to high.
Some of the values in the data collected might be off the scale. An hypothesis is statement that is tested in a research. The hypothesis developed in a research describes what one expects to happen in the study. The hypothesis in company W is whether the amount of product sold in all regions have increased as a result of sales software. The researcher has to develop a null hypotheses for the research question, and alternative hypotheses. The alternative hypotheses should be supported in the research, and the null hypotheses should be rejected. For example, the alternative hypotheses should support that the number of products sold in each region has increased. The alternative hypotheses are represented using HO, and the null hypotheses are represented using HA(Stephens, 2006).
The null hypothesis for the study is:
HO: As a result of the sales software there is increase in the number of product sold in each region or there is decrease in number of product sold in each region.
The alternative hypothesis is:
HA: As a result of sales software, there is increase in number of products sold in each region.A null hypotheses is used to test data collected to see how probable the data. If the data being tested appears improbable then the researcher concludes that the null hypothesis is false. If the data being tested looks reasonable under the null hypotheses, the researcher has to make the right conclusion. The null hypotheses in this case can be true or false. The data does not give enough information to make the conclusion. For example, in company W, the researcher can test the null hypotheses in the data collected, and reject it in favor of the alternative hypotheses.
The null hypothesis in this case is HO (Stephens, 2006).The alternative hypotheses is tested instead, and a null hypotheses for two tailed test is represented by µ1 ≠ µ2.The chi-square test is used to check if the data has a specific distribution.Chi square test is applied to all types of data as it is flexible. The data to be tested is grouped into different categories, and then the researcher checks the number of data points in each category. If the difference with the excepted number is too big, the researcher should reject the null hypotheses. Then the alternative hypothesis should be tested. In this case the alternative hypothesis should be tested instead of the null hypotheses, and the result will be the software has led to increase in the number of products sold (Donelly, 2004).
Donelly, R. (2004).The complete idiot's guide to statistics. Alpha Books
Stephens, L. (2006) .Schaum outline of theory and problems of beginning statistics .Edition2.McGraw Hill Professional