(1) Each group should select a topic based around a single dependent variable. Each individual member of the group should locate at least one social science journal article that contains regression analysis about that topic. Individuals within the groups should compare the findings. How strong are the regression models? (You may need to explain that the correlation coefficient is often presented not as Pearson’s r but as r2. The r2 represents the amount of variation in the dependent variable that is explained by the set of independent variables.) How is significance established for each independent variable? Have the group discuss which independent variables have been missed by researchers.
(2) Select a data set (such as the Census or the GSS) that is easily accessible on the Internet or in a computer lab. Instruct student groups to devise two tables and two graphs derived from this data set. One table and one graph should be deliberately deceptive, while the other set should be accurate. Have each group present their tables and graphs to the class, if possible, making copies to distribute to classmates. The class should try to identify the deceptions and the ways in which each group has tried to deceive them. If any group succeeds in deceiving the class, they deserve special recognition.
(3) How good, or bad, can it get? Each group member should review newspapers and magazines for 1 or 2 weeks, looking for graphs that purport to convey information about social issues. The group should then review all graphs and pick the best and the worst. One group representative can then make a nominating speech for these selections to the entire class. After all nominates are made, the class can review the criteria for good (and bad) graphs and vote for the winner and the loser.
(4) If you conducted Group Exercise #3, #4, or #5 in Chapter 8, have each group (or the entire class) enter their data into a spreadsheet or statistical program. Instruct them to clean the data after it has been entered. Have each group (or select new groups) to come up with hypotheses based on the variables for which data were collected. Have them generate graphs and measures of central tendency for each variable in their hypothesis. Have them construct appropriate frequency tables. Each group should prepare a presentation that contains an overhead of their tables and graphs or a poster, so that other students may see their tables and critique them as necessary.
(5) Separate students into groups and have each group figure out the following and design rough graphs appropriate for each: the modal favorite color, the median estimated distance that students in the group live from campus (distance is zero if they live on campus) and the mean number of credit hours that they are taking this term. When they have completed this information, construct measures of central tendency for the class based on the measures of central tendency for each group. What is the modal favorite color, the median distance from campus, and the mean number of credit hours for the class? You can then poll individuals from the class as a whole to determine whether the measures of central tendency based on a series of samples is similar to the measures of central tendency for the class as a whole (This is also a good way to demonstrate the advantages of using multiple samples from the same population and/or to demonstrate the concept of a grand mean.)
Individual and Group Activities
(1) Professional Styles of Presentation
This project will help you increase your presentational repertoire.
(a) Select five issues of at least two social science journals that present the results of original quantitative research.
(b) Review each article in each issue. Focus your attention only on those articles that include tables presenting quantitative data.
(c) Create a list for each journal issue that shows the total number of articles, the number of articles using quantitative data, and the number of data tables in each of these articles. For each table, record the statistics that were used and describe the format of the presentation.
(d) Compare what you have found to the statistics and table presentation described in the text. What additional presentational options have you encountered? Which do you prefer? Why? What additional statistics do you need to learn about in order to read this journal literature?
(2) Displaying Census Data
This exercise will introduce you to the data available online from the US Census.
(a) Go to www.census.gov. Click on “State and Country Quick Facts.”
(b) Pick any state from the drop down list to get an idea of what variables are available.
(c) Select a variable that you think will vary by region within the United States.
(d) Group states into regions that make sense given this variable (e.g., Northeast, Southeast, Midwest, Southwest, Mountain States, and Pacific States). Make sure to specify which states are in each region. Select two regions that you will analyze (This assignment will be more difficult if you have only grouped states into two groups!).
(e) For each region, compute the mode, median, and mean for the variable you have selected. A calculator will do you a world of good here! Also, note the range on the variable for each region.
(f) Decide which measure of central tendency is most appropriate for the variable you selected.
(g) Construct a table or a graph that displays the comparison of the measure of central tendency between the two regions, as well as the national parameters (which can be found under “USA Quickfacts.”
Write a brief report in which you describe this entire procedure and present evidence of a regional difference or a lack of regional difference.