As a front runner in the field of choice modeling and preferred partner of Sawtooth Software, global leader in conjoint software, we are proud to host 2 hands-on tutorials and presenting 3 inspiring papers on conjoint innovations, data fusion and on data quality at the Sawtooth Software 2019 Conference in San Diego, CA.:

Pre-conference Tutorials

Running a Conjoint Project from Start to Finish – Intermediate

Monday Sep 23 | 8:00 AM – 12:00 PM
Remco Don, Manager
Jeroen Hardon, VP and Location Director Hoboken

While theory of conjoint analysis is very important and much discussed at the Sawtooth Conference, we want to take the practical approach with this workshop and run you through an entire CBC study. We will go over each of the steps including data re-coding, counts, Logit, Latent Class and HB. During each of these steps we will show tips and tricks to get the most out of your data. How do you model the data, interpret the result and what are the key things to look out for? All of this will be discussed based on a real data set using Lighthouse Software.

Experimental Conjoint Solutions – Advanced

Monday Sep 23 | 1:00 PM – 5:00 PM
Jeroen Hardon, VP and Location Director Hoboken
Kevin Lattery, VP Methodology and Innovation

Sometimes standards are not enough to answer research questions at hand and one has to deviate from the roads most traveled. The trainers would like to get you on board for this 4-hour tutorial on advanced applications in CBC and MaxDiff. A number of challenging and interesting extensions of choice modeling will be discussed, including:

  • Custom design techniques combining Lighthouse and Excel and discuss the implications of these customizations
  • Showing different coding methods for faster processing
  • Combining multiple MaxDiff studies in one Latent Class estimation
  • Combining MaxDiff and CBC: different ways of data augmentation
  • Duct-tape solutions for the red-bus-blue-bus problem

Conference sessions

Can We Use RLH to Assess Respondent Quality?

Wednesday Sep 25 | 2:15 PM – 3:00 PM
Marco Hoogerbrugge, Research Director

In CBC studies, the Root Likelihood (RLH) is often a used as a sufficient indicator of the quality of the conjoint data. In the presentation we will discuss why this is not good enough, we will present an improved procedure for assessing the quality of the conjoint data (conceptually) and we will compare different variants of the new procedure.

Combining CBC and Dynamic Choice Models for More Accurate Forecasting

Thursday Sep 26 | 2:30 PM – 3:00 PM
Faina Shmulyian, Methodology Manager

In categories like Tech and Innovation, Travel, or Luxury Goods consumers often spend a significant amount of time considering products or offerings before making a purchasing decision. During this time, consumers’ preferences can be affected by advertisement, social media, socio-economic processes, or family and friends. We will outline and discuss an approach combining a standard Choice-Based Conjoint and Dynamic Choice Modelling allowing adjusting for the external effects and resulting in more accurate forecasting.

Data Fusion: A Flexible HB Template for Modeling Structures across Multiple Data Sets

Thursday Sep 26 | 3:30 PM – 4:15 PM
Kevin Lattery, VP Methodology and Innovation

Data Fusion analyzes multiple sets of data in relation to each other. We describe three general analytical approaches to data fusion: multi-stage, data stacking/augmentation, and structured modeling. These approaches can be used independently or jointly across diverse data fusion contexts. These data fusion contexts include common applications like “Anchored MaxDiff”, “Dual Response None”, and ACBC. We describe the data stacking in these common applications and provide an alternative structured model using a flexible hierarchical Bayes template.