Hypothesis Reflection
Learning Team A:
Kala Chavez, Jessica Gonani, Stacie Knauss, Michael Orcutt, and Troy Stodard
QNT 351/Statistics for Business and Economics
June 17, 2014
Don Silva

Hypothesis Reflection
This week Learning Team A is tasked with discussing three parts to hypothesis research. These objectives included evaluating the steps to test a research hypothesis, compare the means of two or more groups, and calculating the correlation between the two variables. By discussing these three objectives, team A will gain a better understanding of how the objectives are associated with statistical analysis.
Research hypothesis
Testing a theory or hypothesis sounds like an easy task but it is not. There are many steps to testing a research hypothesis. The five steps of hypothesis testing are (The University of North Carolina at Chapel Hill, 2012): 1. Stating the research question or questions 2. Specifying the null and alternative hypothesis involved 3. Calculating the test statistics 4. Computing the probability of test statistics or the rejection region 5. Stating the conclusions
The first step, stating the research question, will clearly define and identify the research question. The population of interest, hypothesized values, and the parameters of the variables are clearly stated. Next is to state the hypothesis of at least two possible outcomes. These are the null and alternative hypothesis. The null hypothesis states that nothing will change or nothing will occur because of the process. The alternative hypotheses are the opposite of the null hypothesis. These state the experiment did have an effect over all. The next step is to calculate the test statistic. The test statistic tests the hypothesis to determine if the null hypothesis is right. Computing probability of the test statistic or rejection region is the second to the last step. The last step is to state the conclusions of the all the tests. Stating the conclusion in a clear definition will result in the confirmation of either a null or an alternative hypothesis.
Comparing the means of two or more groups: Stacie and Mike There are three main methods that are utilized to compare the means of two or more groups. These methods are called Turkey, Bonferroni, and Scheffé. Each method is designed for a specific type of pairwise comparison. The Turkey and Bonferroni methods can be used when the pairwise comparisons are similar (McClave, Benson, & Sincich, 2011). The difference between the methods is that the Bonferroni method does not necessarily need equivalent sample sizes, but the Turkey method does (McClave, Benson, & Sincich, 2011). The Bonferroni method compares pairwise and general contrasts, but the Turkey method just links pairwise comparisons (McClave, Benson, & Sincich, 2011). The Scheffe method is named after the statistician Henry Scheffe. This method is used if there are large levels within a linear regression analysis. By using this method it will account for multiple comparisons at the same time. As a single step method this method applies to all possible contrasts among the factor level means. This is unlike the Turkey method that only considers pairwise differences when calculating an equation. This method is the preferred method over the Turkey-Kramer method because it provides narrower confidence limits.

Calculating the correlation between two variables: Jessica and Troy

References
McClave, J. T., Benson, P. G., & Sincich, T. (2011). Basic statistics for business and economics (11 ed.). Upper Sadle River, NJ: Pearson Hall. doi:9780321640116