ÂÜÀòÉç¹ÙÍø

CenSoC and CHERE Workshops

Introduction

Centre for the Study of Choice and the Centre for Health Economics Research and Evaluation at the University of Technology, Sydney will host four workshops, Sydney, Feb 2008

 : : : Posting

: : call

Announcing CenSoC and CHERE Workshops February 2008

The Centre for the Study of Choice (CenSoC) and the Centre for Health Economics Research and Evaluation (CHERE) at the University of Technology, Sydney (UTS) announce 4 workshops at UTS February 12 – 22, 2008. Workshops are conducted by one or more CenSoC/CHERE team members who have broad and deep expertise and experience in their respective areas. Workshops can be taken together or separately as modules. Class sizes are limited to 25 to maximize the learning experience.

Please note:

  1. A limited number of fee-waiver scholarships are available for all workshops described below. If you wish to be considered for a fee-waiver scholarship, please submit to Prof Jordan Louviere (jordan.louviere@uts.edu.au) a) a MAXIMUM two page, single-spaced letter outlining why you are applying for the scholarship, how the scholarship will benefit you academically and/or commercially, and how you intend to make use of what you learn; and b) a complete CV (curriculum vitae or resume) that is sufficiently detailed for us to understand who you are, what you do, your academic or commercial achievements and anything else that you think makes you stand out from the crowd. A committee of CenSoC/CHERE academics will evaluate all scholarship applications and make recommendations to the Executive Director. All decisions of the committee/Executive Director are final.
  2. Academics and PhD (research) students who would like to attend the workshops can receive a 1/3 fee reduction on a first-come, first-served basis. The number eligible for this fee discount is limited to the first ten people who apply for the discount in each workshop. To apply for the discount you must simultaneously register for the workshop and email Prof Louviere that you intend to apply for the discount (see above email address). Applicants for discounts will be considered in order of time-stamps on email applications.
  3. Except as noted below, morning and afternoon tea/coffee and a light lunch is provided as part of the workshop fee.

Workshop 1: Implementing Online Stated Preference Discrete Choice Experiments (DCEs)

Workshop Leader: David Hackman

What this workshop is about:

The workshop aims to provide a broad overview of how to design, implement and analyse stated preference DCEs online. This workshop allows participants to a) be aware of the steps involved in implementing a stated preference discrete choice experiment online; b) appreciate some of the practical tradeoffs needed to conduct a stated preference discrete choice experiment online; c) understand the implications of some of the decisions that must be made and tradeoffs involved in implementing stated preference DCEs online.

Who Should Attend:

Two potential groups of attendees will find this workshop attractive; separate workshops will be run for each group:

  1. Academics or practitioners who have completed a PhD in psychology, economics, finance or statistics and/or have substantial experience in applying behavioural decision theory or choice modelling (including substantial coursework and training).

    When: 11 and 12 February 2008 (9:30am-4:15pm)

  2. Individuals interested in understanding how to implement stated preference DCEs online, but who have not completed a PhD in psychology, economics, finance or statistics and/or have little expeience applying behavioral decision theory or choice modeling

    When: 21 and 22 February 2008 (9:30am-4:15pm)

Please note: the above two workshops will be treated as ONE workshop for the purposes of scholarships, academic discount places and class size.

We Assume That Participants Have The Following Knowledge:

  1. Basic statistics and data analysis (e.g., regression).
  2. Basic understanding of binary logit and multinomial logit models and the concept of Best/Worst scaling (see pre-readings).

Advanced Preparation Required of All Participants:

  1. Read Sections: 1.2 (pp. 3-5), 2.1, 2.2 & 2.3 (pp. 15-21) of Train, Kenneth (2003), Discrete Choice Methods with Simulation, Cambridge University Press, Cambridge,
  2. Flynn, T.N., J.J. Louviere, Peters, T.J. & J. Coast (2006) "Best-Worst Scaling: What It Can Do for Health Care Research and How to Do It" CenSoC Working Paper Series, No. 06-001,

Advanced Preparation Recommended for All Participants:

    Chapters 1 & 2 of Train (2003) and chapters 1, 2 & 3 of Louviere, J.J., Hensher, D.A. & J.D. Swait (2000) Stated Choice Methods: Analysis and Application, Cambridge University Press, Cambridge.

Registration:

    Complete the attached form and email to: workshops@censoc.com

    Workshop notes and all computing hardware/software are provided. However, participants should bring a memory key/stick to save and work on files. Only morning tea and afternoon tea are provided, no food is provided.

Fees: $1600 per workshop

Workshop 2: Designing Generic Discrete Choice Experiments

Workshop Leaders: Deborah Street and Leonie Burgess

What this workshop is about:

This 3-day workshop covers the theory of and provides practical experience in constructing generic discrete choice experiments (DCEs). We cover the basic principles of experimental design on the first day. We then focus on construction of DCEs, including comparing six widely used design construction strategies. We will show participants how to construct DCEs using the software that accompanies the workshop textbook The Construction of Optimal Stated Choice Experiments: Theory and Methods by Deborah Street and Leonie Burgess (published by John Wiley, 2007). This textbook is provided free to participants.

Who Should Attend:

The workshop is intended for practitioners in fields that use DCEs, including marketing, transport, applied economics (e.g., health, environmental, agricultural) and government.

We Assume That Participants Have The Following Knowledge:

We assume that participants have basic stats knowledge and some familiarity with DCEs.

Workshop Outline:

  1. Introduction to the design of experiments
    • Factorial designs, fractional factorial designs, contrasts, D-optimality and efficiency.
    • Sources of designs and how to obtain more designs from available designs.
  2. Designing generic DCEs
    • The MNL model, likelihood, D-optimality and efficiency for DCEs.
    • Construction of optimal and smaller near-optimal designs to estimate main effects only, or main effects plus some or all 2-factor interactions.
    • Comparison of 6 commonly used DCE construction techniques.
  3. Advanced DCE design techniques
    • Optimal choice set size.
    • DCEs where each choice set has a common base, a none option or both.
    • Blocking choice sets into different versions.
    • Strategies to avoid dominated/dominating options.
    • Using other structures to construct DCEs like Latin Squares or BIBDs.

When: 14, 15 February

Registration:

Complete the attached form and email to: workshops@censoc.com. Workshop notes and all computing hardware/software are provided. However, participants should bring a memory key/stick to save and work on files.

Fees: $1600

Workshop 3: Introduction to Choice Models and Choice Experiments

Workshop Leaders: Jordan Louviere and Paul Wang

What This Workshop Is ÂÜÀòÉç¹ÙÍøt:

This workshop provides an introduction to and conceptual overview of random utility theory based choice models. It covers the basic concepts needed to understand what a choice model is and the theory that underlies choice models. It also illustrates how to apply choice models to various practical problems.

Who Should Attend:

Those who want to understand what choice modeling is all about and what they and/or their organizations can do with it; those who want to learn how to distinguish those who claim to "do" choice modeling but have no formal education in the field from those who have formal education and experience to know how to "do" choice model properly; those who want to learn enough about choice modeling to know if it potentially would be useful in their research.

When: 18 & 19 February

We Assume That Participants Have The Following Knowledge:

Participants should understand basic statistical concepts and be familiar with statistical models like multiple linear regression and analysis of variance.

Workshop Outline:

Session 1: The simple multinomial logit model (MNL)

  • Introduction to Choices and Random Utility Theory
  • The conceptual basis of McFadden’s MNL (also called a conditional logit model)
  • Properties of the MNL and model estimation
  • Interpreting estimation results and statistics associated with MNL
  • MNL examples

Session 2: More complex MNL specifications

  • Generic versus alternative-specific indirect utility specifications
  • Introducing covariates into utility specifications
  • The role of the variance-scale ratio in model comparisons
  • Testing for and capturing violations of IIA
  • More complex modeling examples

Session 3: Designing choice experiments for with alternative-specific models

  • The general problem
  • Simple pricing designs and associated models
  • LMA designs and associated models
  • Availability designs and associated models
  • More general models that allow for error variance differences

Registration: Complete the attached form and email to: workshops@censoc.com. Workshop notes and all computing hardware/software are provided. However, participants should bring a memory key/stick to save and work on files. Please note that food and drinks are not provided.

Fees: $1600

Workshop 4: Advanced Choice Models

Workshop Leaders: Denzil Fiebig, Jordan Louviere & Nada Wasi

Session 1: Modeling the choices of single individuals – Jordan Louviere

What This Workshop Is ÂÜÀòÉç¹ÙÍøt:

This workshop provides an introduction to the design and analysis of DCEs that can be used to model single people ("bottom-up" choice models). There are many academic and practical reasons why one would want to estimate models for single people, including avoiding assumptions in "top-down" choice models about distributions of preferences in a population, capturing and/or testing components of variance, and using closed-form models to approximate the choices of populations of people. We cover theory and methods developed by CenSoC that allow researchers to model single people.

Who Should Attend:

Those who wish to understand how to model single people and/or those who would like to have deeper understanding and appreciation of a variety of issues and problems with many commonly applied choice models.

We Assume That Participants Have The Following Knowledge: Participants should be familiar with basic choice models and choice experiments.

Workshop Outline:

  • Outlining the general problem
  • Combining advances in statistical design with more preference questions
  • Expanding preference responses
  • Different ways of estimating models/pros and cons
  • Implementing the approach and developing forecasting systems
  • Comparisons with traditional first-choice only data models
  • Potential academic and commercial applications

When: Afternoon of 19 February

Registration: Complete the attached form and email to: workshops@censoc.com. Workshop notes and all computing hardware/software are provided. However, participants should bring a memory key/stick to save and work on files.

Fees: $400

Session 2: Advanced Discrete Choice Model Estimation & Analysis – Denzil Fiebig & Nada Wasi

What This Workshop Is ÂÜÀòÉç¹ÙÍøt:

This workshop covers several advanced choice models that relax error assumptions underlying simple choice models, including models currently used in practice as well as emerging models developed by CenSoC research teams. Participants will come away with an overview and appreciation of types of models, underlying assumptions and potential applications. We also will use STATA to demonstrate and illustrate several of these models using real data.

Who Should Attend:

Those who would like to understand the limitations of simple choice models and gain an overview of several key past and current models that relax the restrictive assumptions of simple choice models. The workshop also would benefit those who wish to learn about the pros and cons of various models, including current "fads".

We Assume That Participants Have The Following Knowledge:

Participants should be familiar with basic choice models and choice experiments.

Workshop Outline:

  1. Problems of the MNL model
    • IIA assumption
    • Examples of unrealistic substitution pattern
  2. Relaxing the IIA assumption
    • Overview of possible correlations
    • Correlation across choices
    • Heterogeneous alternative specific constants (loyal consumers)
    • Heterogeneous tastes
    • Working examples with models available in STATA (e.g., random effects logit/probit)
    • Demonstrating mixed logit/available resources (e.g., Ken Train’s code)
  3. Very brief overview of other models used in practice
    • Nested logit (accommodating correlation across choices with similar unobserved attributes)
    • Latent class (accommodating heterogeneous tastes)
    • G-MNL

When: 20 February

Registration:

Complete the attached form and email to: workshops@censoc.com. Workshop notes and all computing hardware/software are provided. However, participants should bring a memory key/stick to save and work on files.

Fees: $800