1. After finding the omnibus result in an ANOVA, using the Bonferroni test can help you to find out more specific information without causing added error. Please describe what a Bonferroni test can do, and what it protects against. 2. Once the ANOVA is calculated or run in SPSS, what test comes next in order for the researcher to understand the complete picture of his/her results? 3. How do nonparametric statistics differ from parametric statistics? Non parametric tests state if data don't or cant (e.g. categorical) form norm distribution then can't infer from sample to population. Used these when data don't have mean (e.g. ranked data) or meaningful & equal intervals between values (e.g. Likert scale).Usually equivalent non-parametric tests to parametric tests (e.g. Mann-Whitney U test = t-test). Non-parametric tests let us infer something about sample, but not population as whole. In the Parametric tests if data do form a norm distribution, can infer from sample to population (because data also norm distributed).Most common stats test, are based around mean & estimates of population mean. 3 assumptions: data nor distributed, variance (or spread of data) should be approx. equal in samples, no extreme scores. Computational & Table analysis (worth 20 points each) 4. Dillards has hired marketing researchers to help them determine which handbag designers they will stock their sales floor with this fall. The researchers surveyed 100 customers to determine if there was a preference for one particular designer. Below are the results. a. Complete the table and use a chi-square analysis to report the findings from the following scores. Be sure to correctly report the chi-square value, degrees of freedom and the p value. Also include the critical value. b. Was there a significant difference found between customer preferences and the four different designers? Expected value Observed value Michael Kors X 17 Coach X 33 Kate Spade X 38 Burberry X 12 A researcher wanted to see if people's happiness could be increased over the course of a week. In one group, she had participants engage in random acts of kindness, the second group began practicing gratitude, the third group engaged in both random acts of kindness and a gratitude practice and in the fourth (control group) she had them write about 3 things that happened to them each day. Below are the results obtained from the study. Be sure to answer ALL questions below. ANOVA Sum of Squares Between Groups df Mean Square 15.335 3 5.112 Within Groups 2817.729 105 26.836 Total 2833.064 108 F Sig. .190 .903 A. How many groups were used in this study? B. How many participants were used in the study? C. Report the findings the way we practiced in class, including the F ratio, degrees of freedom, the obtained statistic and your p value. D. Were the results from this study significant? How did you arrive at this conclusion? Computational & Reporting (30 points) 5. Researchers wanted to test out a new drug that was designed to reduce anxiety among those who suffer from social anxiety. They gave the first group a placebo, the second group a low dose of the medication and the third group a high dose of the drug. Below are the results from the study. Placebo Low dose High dose 3 5 7 2 2 4 1 4 5 1 2 3 4 3 6 E. State the null and research hypotheses. F. Set the level of risk (or the level of significance) associated with the null hypothesis. G. Select the appropriate test statistic. - What test statistic would you use to find the F value? H. Compute the test statistic value. a. Compute by hand as well as in SPSS. Be sure to provide your work as well as the SPSS output results. b. If necessary, run a post-hoc analysis in SPSS (Bonferoni) I. Determine the value needed for rejection of the null hypothesis using the appropriate table of critical values for the particular statistic. J. Compare the obtained value and the critical value. K. Make a decision. With this information would you conclude that the results are significant? Why or why not? Report your findings the way we practiced in class