Have an idea?

Visit Sawtooth Software Feedback to share your ideas on how we can improve our products.

Conjoint bias in attribute familiarisation exercise beforehand


I am running a conjoint study, and before the respondent commences the CBC tasks they are briefed as to what attributes that they are going to see with a brief description of each so that they understand what they are being displayed.

Is there any chance this could lead to bias in the results? I have read a bit on Experimenter Demand Effects which explains that this could potentially bias results. However, given that we are showing all attributes that are to be tested to respondents will this avoid any potential bias? Any help would be gratefully appreciated.

asked Aug 1, 2019 by DaveT (245 points)

1 Answer

0 votes

If you give all the attributes equal treatment in your briefing, that should minimize the chances that you're adding a bias.  We usually think a briefing adds to the quality of the experiment.
answered Aug 1, 2019 by Keith Chrzan Platinum Sawtooth Software, Inc. (93,025 points)
Hi Keith,

Thanks for the quick response, very helpful. Do you know if there are any research papers that back this up at all?

Sorry, I do not know that anyone has researched this question specifically.  My advice was based only on personal experience.