In the aftermath of the financial crisis of 2008, brokerage firms struggled to get individual investors back into the stock market. To gain competitive advantage, market analysts in nearly every industry group their customers into different segments by placing advertisements where they can have the greatest impact. The brokerage business is no different. Because different groups of people invest differently, customizing advertising to the needs of these groups leads to more efficient advertising placement and response. Brokerage firms get their information about the investment practices of individuals from a variety of sources, among which the U.S. Census Bureau figures prominently.

The U.S. Census Bureau, with the help of the Bureau of Labor Statistics (BLS) and the Internal Revenue Service (IRS), monitors the incomes and expenditures of Americans. A random sample of Americans are surveyed periodically about their investment practices. In the file Investment Strategy Segmentation you'll find a random sample of 1000 people from the 48,842 records found in the file Census Income Data set on the University of California at Irvine machine learning repository. This subset was sampled from those that showed some level of investment in the stock market as evidenced by claiming Capital Gains ($), Capital Losses ($), or Dividends ($). Included as well are the demographic variables for these people: Age (years), Sex (male/female), Union Member (Yes/No), Citizenship (several categories), College (No College/Some), Married (Married/Single), Filer Status (Joint/Single).

To support her segmentation efforts, a market analyst at an online brokerage firm wants to study differences in investment behaviors. If she can find meaningful differences in the types and amounts of investing that various groups engage in, she can use that information to inform the advertising and marketing departments in their strategies to attract new investors. Using the techniques of Part III, including confidence intervals and hypothesis tests, what differences in investment behaviors can you find among the various demographic groups?

Some specific questions to consider:

1. Do men and women invest similarly? Construct confidence intervals for the differences in mean Capital Gains, Capital Losses and Dividends.

Be sure to make a suitable display to check assumptions and conditions. If you find outlier(s), consider the analysis with and without the outlier(s).

2. Make a suitable display to compare investment results for the various levels of Citizenship.
3. Do those who file singly have the same investment results as those who file jointly? Select, perform, and interpret an appropriate test.
4. Compare differences in investment results for the other demographic variables, being careful to check assumptions and conditions.

Once again, make suitable displays. Are there outliers to be concerned with? Discuss.

Summarize your findings and conclusions about the investment practices of various groups. Write a short report in order to help the market analyst.

Answer rating: 100% (QA)

1 Compute the 95 confidence interval for the mean diffe... View the full answer