# Aug 2: Statistics in Practice

In part 1 of this week’s brief, we looked at political analytics; in Part 2 we extend that look to commercial domains. Our course spotlight is Persuasion Analytics, taught by Ken Strasma, who pioneered the use of statistical modeling to microtarget voters in the 2004 U.S. presidential campaign.

See you in class!

Chief Academic Officer, Author, Instructor, and Founder
The Institute for Statistics Education at Statistics.com

# Method Spotlight

## Uplift Modeling for Churn

In Part 1 we looked at uplift modeling to optimize messaging in a political campaign.  In Part 2, we look at commercial models, which were the first applications of uplift modeling – testing promotions of new products, cross-selling , and reducing churn.  The latter – reducing churn – is a good example […]

# Student Spotlight

After several years working as a researcher and analyst for the government, Dave Plumstead decided he needed a deeper statistical skill set – to go beyond tabulating and describing, and be able to explain why, and predict.  He came to Statistics.com and enrolled in […]

# Problem of the Week

You work for an oil company, reviewing an oil field available for lease.  Company geologists estimate a 30% probability that the peak production will be 200,000 barrels per day, a 30% probability that the production will be 50,000 barrels per day, and a 40% probability that the flow will be uneconomically low and the field will be shut down.  What is the expected value of the field, in terms of barrels per day? […]

# Course Spotlight

## Aug 23 – Sep 20: Persuasion Analytics

Your instructor is Ken Strasma, who pioneered the use of statistical methods for microtargeting voters as targeting director with the 2004 Kerry campaign, and again with the 2008 Obama campaign. The methods covered in the campaign apply to any undertaking to persuade large numbers of people (e.g. advertising), but are presented in the context of political campaigns.

You will learn how to:

• Find voter targets that are appropriate to a campaign phase
• Assess the effectiveness of a voter contact
• Designing a survey instrument for use in a predictive model
• Implement predictive models with voter data
• Add uplift modeling to predict whether a voter responds better with a “treatment”