Anova

ANOVA

Monica Jones

RES/342

Marsha Smith

November 16, 2011

ANOVA

Praxidike Systems hired me as their manager over Quality Assurance, my position is to ensure clients are greatly satisfied and that delivery only takes one trip. First order of business was deciding which test I would be using. Three important assumptions had to be made in order for ANOVA to be used.

1. Errors independent and random for one another.

2. Distribution of the populations is normal.

3. Variances are the same in all populations.

For a good fit the chi-square test needs to be implemented, to check if the population has a normal distribution or not.

In this case the hypotheses are

1. H0: Populations distribution is normal

2. HA: Population distribution is not normal

The results were that the statistics of the test lie outside the area of acceptance, and that the null hypothesis should be rejected. As concluded, it can be presumed that the distribution is normal in the population. The nonparametric Kruskal-Wallis test should be used

The technique of blocking is the next objective, and is not always needed. Blocking gives you the ability to view the effects of a treatment with a sample that is smaller. We have to make sure that time was not wasted in creating the block, and remember that the block is not easy to set up.

To may need to match two or more factors with the variable, when the block is being set up. This option may not be the case every time. The relative efficiency can be calculated to find if the block is required or not. The block design worked best in this simulations case.

In conclusion, I gathered that an ANOVA two-way with scheduled variances as (A) factor and the number of (B) factor defects. I noticed that in test statistic (F) for the column, row, and greater interaction than that of (F) critical. This means that the null hypotheses in the number of defects, scheduled variance and the effects of...