Case studies

In this section, we present summary case studies in areas such as database marketing, customer or employee satisfaction assessment, medical diagnostics and others. Some of these case studies use disguised client data, while others are hypothetical examples demonstrating various analytical techniques and possibilities. (Hypothetical case studies are edited versions of cases originally created by SPSS, Inc., and are used with permission. In some cases, we have altered the original raw data prior to the analysis, in order to make or emphasize specific points.)

Some of these case studies (e.g., the time-series/forecasting analyses) contain a few technical references to provide additional information for those readers who have some statistical training. But for those readers who have no technical background, these references can simply be skipped over, and the reader can still understand the important points and conclusions of the case study.

Check back periodically, because we will be adding case studies to this section of our web site.

	Securities Brokerage Case Study This case study documents a predictive market segmentation model designed to identify and profile high-value brokerage customer segments as targets for special marketing communications efforts. The dependent variable for this ordinal CHAID model is brokerage account commission dollars during the past 12 months. Predictors include proprietary client data (various account status and trading behavior variables) as well as syndicated demographic and lifestyle variables.
	Bank Loan Credit-risk Analysis * Here we apply binary logistic regression to the analysis of loan risk. We build a predictive risk model using a sub-sample of bank customers, then validate the model on a holdout sample of customers, and finally we apply the model to score a group of loan prospects.
	Telecommunications Churn Study * This example uses Cox Regression to generate a “survival analysis” for modeling "time to churn." Various demographic and product-usage characteristics are used to identify customers who are quick to switch to another service.
	Forecasting Catalog Sales of Men’s Clothing * Using ARIMA time-series analysis, we develop a sales forecast using predictors such as the number of catalogs mailed, the number of pages in the catalog, the number of phone lines open for ordering, and the amount spent on related print advertising.
	Understanding Fluctuations in Market Share * Here we use time-series analysis to better understand local market conditions influencing changes in a supermarket’s market share.
	Breast Cancer Survival Analysis * This is a case study of the influence of various patient characteristics on survival rates for breast cancer. The survival analysis technique employed is Cox Regression. This technique is useful in situations where we have censured observations--that is, where some of the patients do not die during the observation period. Techniques like this can help us understand many different types of "survival" situation, ranging from marketing applications such as customer loyalty analysis to engineering applications such as the influence of maintenance schedules and operational factors on a component's time-to-failure.

*Hypothetical case study originally created by SPSS, Inc., and used with permission. In some cases, we have altered the original raw data prior to the analysis, in order to make or emphasize specific points.