Discussion Question 2: Analyze Your Own Data This discussion question asks you to collect and analyze your own data for a chi-square test for association. (1) Step 1 Find or collect your data. You will need frequency counts for two categorical variables. The sample size should be large enough that the expected counts will be at least 5 in each cell, preferably at least it = 50 overall. Many survey instruments give information on categorical data so it should not be hard to nd data online, or you can collect your own data. Be sure that you have information about both categorical variables of interest for each case. (Note: Several suggestions are given below, or come up with your own idea.) ACTIVITIES AND ASSIGNMENTS FOR WEEK 8 (2) Step 2 Conduct the analysis. Using technology or by hand, nd the expected counts and conduct the chisquare test. (3) Step 3 Submit a post in the discussion forum addressing each of the following parts: (a) How did you nd or collect your data? (If you found the data, give a clear reference. If you collected the data, describe clearly the data collection process you used.) What are the cases? What are the variables? What population do you believe the sample might generalize to? Is the sample data from an experiment or an observational study? Include a two-way table of the data. (b) Analysis State the hypotheses of the test. Conduct the test, showing all details such as expected counts, contribution of each cell to the chisquare statistic, degrees of freedom used, and the p-value. (c) State a clear conclusion in context. If the results are signicant, which cells contribute the most to the chisquare statistic? For these cells, are the observed counts greater than or less than expected? Whether or not the results are signicant, describe the relationship as if you were writing this for your friends on campus who have not taken statistics. If the results are signicant, can we infer a causal relationship between the variables? Topic or Resource Suggestions. Use one of these or come up with your own idea or nd your own source. There are many sites reporting frequency counts from survey results. (i). (ii). (iii). (iv). (v). (vi). (vii). Frequency of smoking (never, occasionally, frequently) and gender for students. Academic division (science, arts, business, engineering, ...) and whether the student has a Mac, PC, or neither, for students. Whether a person plans to vote in the next election and political party afliation. Handedness (left or right) and position in baseball. Number of baskets made from the foul shot line out of 20, for you and two friends. Any frequency counts given in a twoway table in a report by the Pew Research Center (pewre search.org) Any frequency counts given in a twoway table in a report on American College Health Associa tionNational College Health Assessment survey (www.achancha.org)