Conjoint and Discrete Choice: A Cookbook for Practitioners

By Paul Richard McCullough

Download This Article

Dear fellow market researcher:

I’ve just published a paper in the most recent edition of Marketing Research magazine titled “A User’s Guide to Conjoint Analysis” (Marketing Research, Summer 2002). In it I give a comprehensive list of the practical do’s and don’ts you should be aware of when designing and implementing a conjoint or discrete choice study.

Here are just a few of the many issues contained in the article:

• Choose the right approach: The first step in doing Conjoint right is to pick the most appropriate method for your particular objectives and circumstances. There are many variants of the conjoint method. For example, there are ratings-based conjoint, choice based conjoint, constant sum conjoint, hybrid approaches, full-profile or partial-profile, Hierarchical Bayes, latent class or traditional estimation techniques. Each of these techniques or variants has its strengths and weaknesses. Picking the right technique for your situation is a trade-off.

• Be aware of the pitfalls: Ever heard of the IIA bias? Do you know which technique minimizes the Number of Levels Effect? Do you know how attribute range impacts attribute importances? Do you know how many conjoint tasks you should have or how large a sample? Conjoint analysis is a powerful technique when done properly. But to do it right, you need to be aware of the ways you can do it wrong.

• Keep it simple: Conjoint studies can quickly get complicated. Design the simplest possible study that still answers your key questions. Simpler studies will take less time to complete, which will, in turn, increase your response rates. Simpler studies are more likely to yield accurate data because they are less likely to confuse or fatigue your respondents. Simpler studies are generally cheaper to design and execute. And almost always easier to explain to management.

You can review the entire article by either clicking on the title highlighted in the opening paragraph or in the left hand margin. The article is listed on our website under its original title, “How To Do Conjoint Analysis Right.”

If you’d like any assistance with your next conjoint project, using either online or more traditional data collection, please give us a call. MACRO would be happy to help in any way that we can

Conjoint analysis has long been recognized as a superior tool for:

• Developing pricing strategy
• Designing new products
• Identifying key product features and benefits

MACRO has been conducting conjoint studies since I founded the firm in 1987. I have published extensively in the field and regularly present papers at internationally recognized conferences on conjoint analysis. We have surveyed tens of thousands of respondents online, both b-to-b and b-to-c, since we began conducting online studies in 1995. All of our web surveys are programmed internally by full-time MACRO staff, trained to use the latest Sawtooth Software web survey and conjoint tools. I personally oversee all of MACRO’s projects

Our file server has a large pipeline. No survey is too large for our system.

If you need assistance either programming and hosting a web survey or designing and analyzing a conjoint survey, please give us a call. If you need fast turn-around or have an especially complicated study design, please give us a try.

MACRO CONSULTING
Intelligent Decisions