Modeling count data: Understanding & modeling risk & rates

Date:
Location:
Course duration:
Software:
Level:
Delivered by:

25-26 October 2012
New York City, NY, USA
2-days
Stata
Intermediate
Prof. Joseph M. Hilbe, Arizona State University

 

 

Contents

 

Back to top 

Course Overview

Timberlake Consultants invite you to attend the following two day course to be held in New York. The course will be delivered interactively using Stata 12.

The course provides clear and understandable guidelines on modeling count data including the identification and handling of both under- and over dispersion. The course guides participants in how to select the most appropriate model for a given count data situation, then how to best construct, interpret, and comparatively evaluate each type of model as to its fit. Discussed are the standard Poisson and negative binomial models and their extensions; eg scaled, bootstrapped, zero-inflated, heterogeneous, truncated, censored, generalized, longitudinal/panel, nonparametric, latent class/finite mixture, Bayesian and related models.

 

Back to top 

Agenda

Day 1: Morning

  • Nature of risk, rate, and rate ratio
  • Methods of estimation: maximum likelihood; IRLS
  • Poisson regression: The Poisson algorithm; Model construction; Incidence rate ratios and offsets; Interpretation; Model fit; Marginal effects
  • Over dispersion: Identifying and methods of handling extra-dispersion; Tests for extra-dispersion
  • Negative Binomial Regression: Varieties of negative binomial; Derivation of the basicNB2 negative binomial model; Constructing and interpreting NB2 regression; Analysis of fit; Marginal effects
  • Relationship of count and binomial models

Day 1: Afternoon Workshop

  • Software considerations for modeling and evaluation of count models
  • Setting up a count model: from tables; from grouped data; from observation data
  • Exercises in modeling - selecting the appropriate model for specific types of data
  • Constructing a count model project

Day 2: Morning

  • Variety of count models
  • NB2 vs. NB1
  • NB-H -- Parameterizing the scale parameter
  • Generalized Poisson (GP) and Generalized negative binomial (NB-P; Waring; others)
  • Poisson-inverse Gaussian (PIG)
  • Modeling data with no zero counts
  • Modeling data with excessive zero counts: ZIP, ZINB, ZIGP, ZIPIG
  • Hurdle models
  • Censored and truncated models
  • Finite mixture models
  • Nonparametric count models
  • Overview of count models for longitudinal and clustered data
  • Bayesian count models

Day 2: Afternoon Workshop

  • Software considerations for modeling and evaluation of enhanced count models
  • Exercises in modeling - selecting the appropriate model for specific types of data
  • Constructing a count model project

 

Back to top 

Method of instruction

Lecture/discussion of subject material; PM: on-hands directed computer use applying methods learned in morning session to simulated and real data. Stata primarily used for examples.

 

Back to top 

Prerequisites

A basic knowledge of linear models; some background with generalized linear models desirable.

 

Back to top 

Prices

Cost (per participant):

  Early Registrations Late Registrations
Commerical / Government $1240.00 $1550.00
Academic / Non-profit Research $992.00 $1240.00
Student Registrations $680.00 $850.00

 

  • All costs exclude local taxes, where applicable
  • Late Registrations: Registrations made within 6-weeks before the start of the course
  • Student registrations: Attendees must provide proof of full time student status at the time of booking to qualify for student registration rate (valid student ID card or authorised letter of enrollment)
  • Additional discounts are available for multiple registrations
  • Cost includes course materials, lunch, refreshments and the use of computers (please advise us if you have any dietary requirements)

The number of delegates is restricted. Please register early to guarantee your place.

If you need assistance in locating hotel accommodation in the region, please notify us at at the time of booking.

 

Back to top 


Registration

We welcome delegates to find out more and register for the course by contacting our sales and training team either by email: info@timberlake-consultancy.com, phone: +1 908 686 1251 or by filling out an online registration form.

 

Back to top 


Terms & Conditions

Payment of course fees required prior to the course start date.

Registration closes 5-calendar days prior to the start of the course.

  • 100% fee returned for cancellations made over 28-calendar days prior to start of the course
  • 50% fee returned for cancellations made 14-calendar days prior to the start of the course
  • No fee returned for cancellations made less than 14-calendar days prior to the start of the course

 

Back to top 

Return to:  Training Calendar  |  Home













 
Newsletter Registration