Statistics and Modeling for Direct Marketers Seminar

Basic Statistical Review and Applications

• How “chance” affects the results from DM list and package tests
• How to adjust for these “chance” effects (normal distributions etc.)
• How to build forecasts for the results from your next promotion
• How to determine your needed sample size
• How to test if two list or package results are “statistically different”

Predictive Modeling and Segmentation Modeling

• What the difference is between the two types of modeling procedures, how they relate to each other, and when to use each

Correlation Analysis

• How to read scatter diagrams and plots of residuals
• How to read a correlation matrix, and what it tells you to avoid

Multiple Regression Analysis

• When to use a multiple regression program
• How to analyze data prior to regression analysis and why it’s imperative to perform this step
• How to model the non-linear response patterns so common in direct marketing response analyses
• How to use “dummy variables” in regression analysis
• How to perform simple regression analyses using Excel™

AID/CHAID or "Tree" Analysis

• How to interpret the output of an AID/CHAID program
• Why an AID/CHAID analysis is not used for scoring
• How AID/CHAID can be used to choose variables for a regression model

Networking Opportunity:
Lunch will be served each day to facilitate networking with your peers.

DAY 2
9:00 A.M. – 4:00 P.M.

Factor Analysis

• When to use factor analysis prior to using regression
• How to interpret the output of a factor analysis
• How to use factor analysis to analyze your marketing research data

Cluster Analysis

• How to use a cluster analysis to create your own customer target market groups
• How to decide on the “right” number of clusters to include in a cluster analysis

Discriminant Analysis

• The similarities and the differences between discriminant analysis and regression analysis. When to use each.

Logistic Regression

• What it is
• Why it is the recommended method to be used when performing response analysis
• How it is different from ordinary regression and/or discriminant analysis

How to Get Started

• What hardware and software you need
• How to determine if you should use consultants, service bureau, or do it in-house
• What you can expect to gain in the first year