Multivariate experimental designs have become more common across social sciences. One of these approaches is the factorial survey experiment, a variant of a conjoint design (for an overview see Wallander 2009). It is a survey experiment used to test intended behavior or attitudes, with a number of not so commonly applied features like sampling of experimental conditions and mixing within-and between-respondent designs.
Manual: Factorial Survey Experiments in Qualtrics
Below the link to a short manual I wrote with others on how to implement the factorial survey or in fact any vignettes part in Qualtrics. We show you several options: copy&paste your vignettes, a Python script to create the vignettes and how to use embedded data. Please note that this is not an introduction to Qualtrics as there is plenty of information on the Qualtrics help page.
Please note that other survey tools might be as apt to implement a factorial survey. Just make sure to check that you can include all the necessary randomization (and that you can also export this information in a meaningful way for analyses).
2018-12 Factorial Survey Experiment in Qualtrics
(Most recent version December 2018)
A workshop at your department?
In the past years, I have shared my knowledge and experience of the method at several one- or two-day workshops (see here). If you and others at your department might be interested, please get in touch so we can discuss the options.
Information sources on the method that I have found good and useful:
- The only (short) book I know of with all the information you need, includes an online appendix with some Stata files to make your life easier: Katrin Auspurg & Thomas Hinz (2015). Factorial Survey Experiments. SAGE Quantitative Applications in the Social Sciences Series. Link to publisher
- The GESIS Institute in Cologne/Germany offers courses on the factorial survey experiment (almost) yearly. Check their most recent programme.