Applying ANOVA and Nonparametric Tests - Essay Example

ANOVA which means Analysis of Variance is a statistical tools used for analyzing the level of variations between two population groups. It is used totest hypotheses concerning means when we have several populations. It is used to test the hypothesis that the means of two or more groups are equal with the assumption that, the sampled populations are normally distributed. Other wise, it can be referred to as distribution based test (Doane, D & Seward, L 2007) Non-parametric test on the other hand is the opposite of parametric test.

Parametric test operate under the assumption that, the sample is normally distributed and it is used to analyze non-quantitative data; it deal with qualitative data. For non-parametric tests, the assumption of normality does not hold at all thus advantageous to parametric. However, it is not thought to be powerful like the parametric tests. Thus ANOVA and non-parametric tests are both different due the fact of normality. It is also called the distribution-free test. In the work place, tangible decisions by the management are based on the credibility of the data that is available and which can be analyzed producing results which are valid and reliable (Orris, J 2007).

Thus, to be able to progress in the work place (as an investor), I will use ANOVA to determine whether for example the work of certain group of rehabilitators is the same as work by another different group of rehabilitators. This will mostly be done to reduce expense since if a certain group can perform a certain task as another group can, then there will be no need of having more groups if the available group can do the same kind of work, this will reduce frictions in my work as I will only deal with a single group.

Further I can use the same technique in case some materials are lost to determine whether the available materials can be used in place of the lost ones. On the same note, I will use Chi-square to test whether the prices of the houses are independent or dependent of the gender, size, finish, of age, work experience, salary among many others. A way will be sought on what the prices of the houses should be hinged on; mostly on the finish and size. Further, I will use Chi-square to test whether there is any goodness of fit between the salaries of the tenants and the house one occupies or some people just want to stay in big homes where they will be spending the whole of their pay and no savings.

Spearman rank correlation will be used to determine if rehabilitators’ motivation results to increased finality in terms of the number of houses and the kind of finish given to the house. Mainly, it is expected that, when people are paid well, their morale is high translating to high performance. This can be tested by first getting their performance before the pay increase and the performance after the increase. A positive correlation coefficient will imply that the high pay has a positive relationship with the performance; par rise = high performance.

A negative correlation coefficient will imply that, pay rise = low performance; if the pay is added, the performance goes up-something not expected in most instances. Mann-Whitney test is the alternative of the t-test of independent variables. In the work place, it will be used to test whether the performance of women is the same of the performance of men in terms of buying homes or renting, or whether working day and night shifts results to better results than day shifts only and vice versa for the contractors.

This will be done by taking two samples from the population of contractors; equal number of both females and males. One will be used for the day and night shifts while the other will be used during the day shift. Their performance will be recorded and then compared together to see whether there is any difference at some level of significance (either 95% or 99%-the mostly used statistical levels). One males and females rehabilitators’ performance, the same will happen and an equal number of males and female will be required for the study.

Their performances will be compared to see if there is any difference. Wilcoxon-Signed rank test will be used as the Mann Whitney test above. However, in Wilcoxon, the difference in performance between the day and night shifts will be ranked. This will be based on whether the difference is positive or negative. Both positive and negative will be used to come up with a conclusion. If +ves are many, then the situation determines in which direction the population tilts to. I will use the runs test to see whether those that apply for a certain house are awarded that house in randomness since they all have money and have an equal chance of happening thus randomness will be applied.

In conclusion, I will use non-parametric test to determine the quality of the services to provide to the buyers and my tenants making sure that the smallest loss possible is made in the process. My recommendation will be that, for the decision made to me trustworthy, the data collected about each and every variable should be valid and reliable. Otherwise the decisions will not be believed by anybody. Reference Doane, D & Seward, L (2007). Applied Statistics in Business and Economics. Burr Ridge, Illinois: McGraw-hill Orris, J (2007).

Basic Statistics Using Excel and Mega Stat. Burr Ridge, Illinois: McGraw-Hill