Webinar: Addressing Self-Selection Bias
At the Sawtooth Software Turbo Choice Event held in February 2016, power users came together to share tools, tips and new research about choice experiments. One topic of interest included self-selection bias and what impact it may have on study results. When dealing with a large number of attribute levels, we often want to trim that list down into a manageable set. However, if we allow the respondent to self-select the items that are screened in, we can introduce some amount of self-selection bias into the estimation.
This is an encore presentation of Kenneth Fairchild's paper on self-selection bias where he looks at various approaches of addressing that bias and how they impact the estimation of the part-worths. This webinar was given exclusively to our suite license holders.
This webinar occured on Wednesday, June 8, 2016. A recording of the webinar is found below.