Statistical Modeling Week 2007

October 29 - November 1, 2007, Boston

Bridging the Gap Between Theory and Practice - Our 31st Year!

Statistical Modeling Week is an annual conference sponsored by Statistical Innovations Inc., featuring applications-oriented seminars focusing on the latest trends in statistical analysis. Led by a group of experts in the field, the seminars teach professionals how to apply and understand today's major breakthroughs in statistical methodology. Statistical Modeling Week 2007 will take place in Boston on October 29 - November 1, 2007. Whether you work in marketing, advertising, health, econometrics or evaluation, you will learn to apply these important breakthroughs to your everyday work. And you'll learn more than the techniques - you'll learn the theories behind them. Numerous leading organizations have taken advantage of our prior seminars - here is a sample list. Attendance is limited. Register for Statistical Modeling Week 2007 today! Below are the course descriptions for Statistical Modeling Week 2007. You can also download our conference brochure. This year's conference will feature the following courses:

Courses

October 29, 2007. Introduction to Statistical Modeling Applications, Tony Babinec and David Rindskopf October 30 - 31, 2007. Latent Class and Finite Mixture Modeling, Jay Magidson and Tony Babinec November 1, 2007. Discrete Choice and Other Conjoint Models, Tom Eagle and Jay Magidson

Tuition and Enrollment Information

Tuition

October 29, 2007. Introduction to Statistical Modeling Applications	$895
October 30 - 31, 2007. Latent Class and Finite Mixture Modeling	$1,490
November 1, 2007. Discrete Choice and Other Conjoint Models	$895
Note: Discounts are available for multiple course registrations. After enrolling in the first course for an undiscounted tuition, you deduct $50 for each additional course you sign up for (the maximum discount would be $100 if all 3 courses are taken). Discounts are also available for additional people who sign up from your organization. Their tuition is $795 for each 1-day seminar and $1,390 for each 2-day seminar.

Enrollment

Register for the conference. For more information, e-mail or call us at (617) 489-4490.

Course Descriptions

Introduction to Statistical Modeling Applications

Instructors: Tony Babinec & David Rindskopf Description This important introductory seminar provides an overview and typology for the major kinds of regression and classification models and their applications. We distinguish between supervised and unsupervised techniques for exploratory and confirmatory applications. We also distinguish between the different types of variables (nominal, ordinal, and continuous), show which are appropriate for which models, and coding these variables properly for use in the models. We emphasize the importance of model choice and interpretation of results, and the relationship to data-mining applications. The Program Types of Models Exploratory vs.Confirmatory Descriptive vs. Predictive Supervised vs. Unsupervised Variables and Coding Issues Variable scale types Handling missing data Coding categorical variables Supervised techniques Dependent variable scale types Linear vs. Logistic Regression/ Discriminant Analysis Proper use of the CHAID tree-based technique Unsupervised techniques Cluster analysis - K-Means, Hierarchical vs. Latent Class approaches Data mining and Criterion-based Segmentation examples Use of the CHAID technique to segment based on nominal dependent variable (e.g., response) ordinal dependent variable (frequency of use, profitability) Scoring models, Gains Charts, and Validation issues Regression Models and optimal scoring Relaxing the Likert scale equidistant category assumption What happens when assumptions are violated What you will learn How to choose the appropriate model for different applications About useful SPSS, SAS and other programs Why proper coding of categorical variables is important When K-means clustering is likely to perform poorly How to use CHAID with an ordinal dependent variable The proper interpretation of interaction effects How the Gains chart is used to compare results from tree based and regression approaches Key Driver regression with rating scale data Back to Courses Register

Latent Class and Finite Mixture Modeling

Instructors: Jay Magidson & Tony Babinec Description Interest in Latent Class (LC) models is increasing rapidly. This is due to the growing evidence that they provide better solutions than the more traditional approaches to cluster, factor and regression analysis when the population is not homogenous. In this 2-day course we introduce LC as a probability model and describe various applications using the Latent GOLD program. On day 1, we focus on model fitting strategies and the interpretation of output. On day 2 we consider several advanced topics including identification issues, problems caused by data sparseness, use of bivariate residuals and repeated observations. The Program Day 1 Basic ideas of latent class analysis The concept of local independence The general probability model Handling nominal, ordinal, continuous and count variables LC measurement models LC factor vs. traditional factor vs. correspondence analysis LC regression and segmentation models Inclusion of covariates in LC models Model-based clustering / latent discriminant analysis Comparing models and assessing fit Uni-plot, bi-plot, tri-plot and profile plot Case studies and computer demos Day 2 Identification problems and boundary solutions Use of Bayes constants to eliminate boundary solutions The problem of local solutions Use of randomized start values Use of the bootstrap p-value Bivariate residuals to diagnose local dependencies Repeated measures / conjoint marketing LC Growth models Use of continuous factors as a random intercept What you will learn How to specify LC cluster, factor and regression/segmentation probability models. Interpreting graphical displays of results. What to look for when interpreting output. Strategies for assessing model fit with sparse and non-sparse data. How to isolate the scale effects in ratings data. How to include nominal, ordinal and continuous variables in the models. Why LC models improve over K-means clustering. Back to Courses Register

Discrete Choice and Other Conjoint Models

Instructors: Thomas Eagle & Jay Magidson Description Discrete choice modeling is becoming the preferred method for understanding and quantifying the behavior of decision makers. It is commonly used to forecast market share, design optimal products and services, configure bundles of products and services that will increase sales, estimate the size of markets and determine pricing strategies in competitive marketplaces. This course begins by introducing the theory and practical applications of conjoint and discrete choice modeling, and the random utility model. We then describe the latent class (LC) extension to this model and illustrate its application in traditional rating, choice, ranking and constant sum conjoint experiments. The Program How discrete choice modeling compares to traditional conjoint analysis The behavioral utility theory underlying choice models The simple (conditional) multinomial logit model Development of Excel-based simulators Stated preference vs. revealed preference data Independence of Irrelevant Alternatives (IIA) Existence of unobserved heterogeneity and violation of IIA Accounting for segment differences Assessing fit and comparing models Different types of LC choice models Simple competitive brand pricing experiment Max-diff (best-worst) and other partial/ full ranking models Constant sum and other weighted models Comparison of LC and Hierarchical Bayes (HB) models What you will learn Advantages and limitations of discrete choice models The behavioral utility theory underlying discrete choice models Which choice models to use in a variety of behavioral situations Review of software used to set up choice experiments How to specify, estimate, and interpret simple choice models Why real-world data violate IIA and how LC modeling resolves this problem About advanced models that simultaneously identify market segments and describe their behavioral patterns. Differences between the LC and HB approaches Back to Courses Register

Other Presentations

Check out information on other presentations and seminars conducted by Jay Magidson.