Applying ANOVA and Nonparametric Tests Simulation
Dear savchuk,

Neither you nor I am allowed to share personal information. So I'll not be able to login as you. That's against the TOS for this site. I can, however, answer the first two parts of this without a problem.

1.     What are three lessons you learned relative ANOVA and Nonparametric tests?

ANOVA is used for testing the differences between the means when there are multiple groups. It is possible in this case to simply use a t-test, but as the number of groups increase several problems arise. First of all, using multiple t-tests with more than a few groups requires many pairwise comparisons. Secondly, doing many tests increases the probability of getting at least one of them wrong. ANOVA allows us to test to see if there is a difference between the means of multiple groups while only using 1 test. ANOVA relies on the ratio of 2 separate estimates of the variance to determine if there is a difference in the means. The ratio of these two estimates gives the test statistic. The numerator of this ratio becomes large in the event that the means are not equal. Hence, the test.

One of the assumptions of ANOVA is that the residuals are normally distributed. That is, ANOVA is a parametric test. In the case that we doubt that the residuals follow a normal distribution, ANOVA is not appropriate. Instead, we can use a nonparametric alternative. These sorts of tests use rank to preform similar tests without necessitating parametric type assumptions. The non-parametric analog to ANOVA is the Kurskal Wallis test. It uses the Rank and average Rank to preform the same test.

2.     As a result of using this simulation, what concepts and analytic tools will you be able to use in your workplace (real estate agent)…ie: how do you expect to apply what you learned?

These procedures can be used in pricing tests for real estate. For example, if I were attempting to sell properties in different neighborhoods, I might be interested in coming the average selling price of houses in those neighborhoods. If I had several neighborhoods to compare, I could use ANOVA to test for the equality of their means (or Kruskal Wallis if the normality assumption is in question).

3.     Based on your experience, what additional information would you recommend to the key decision maker in the simulation to solve the challenge given?

If you could tell me the challenge and the results of the test, then I'd be glad to answer the question.


---

Let me know if you have questions or need any points clarified!

Dear savchuk,

for #2, I was thinking that you would compare the average prices of houses for sale so that you could better know how to price a house that you're trying to sell in that area. That is, if a neighborhood has higher selling prices, you'd price the house that you have for sale higher in that neighborhood than you would in another.

for #3, Could you give me a clue about what the challenge in the simulation is? I haven't seen it.