August 2014 release: Study of Ozlab and other Wizard-of-Oz systems
2014-08-31Now available for downloading:
Perspectives on Ozlab in the cloud: A literature review of tools supporting Wizard-of-Oz experimentation, including an historical overview of 1971-2013 and notes on methodological issues and supporting generic tools
urn:nbn:se:kau:diva-33617
Students who would like to have a printed copy can collect this in the lab. (Other interested parties are of course also welcome to request a hard copy; just send Malin a mail.)
The Wizard-of-Oz method has been around for decades, allowing researchers and practitioners to conduct prototyping without programming. The extensive literature review in the field reported here, however, revealed that the re-usable tools supporting the method do not seem to last more than a few years. Generic systems started to appear around the turn of the millennium, but very few are still in use. New systems are designed nevertheless. The systems and issues presented here should be of interest to people in the field of prototyping interaction design.
This review was inspired by the authors’ ongoing re-development of their own Wizard-of-Oz tool, the Ozlab, into a system based on web technology. The report takes stock of some key features of Ozlab as well as reviews and contrasts other re-usable Wizard-of-Oz tools with the ambition to list every generic tool.
The introductory chapter compares and contrasts prototyping in general with Wizard-of-Oz prototyping and provides an historical overview of Wizard of Oz in the development of digital interactive systems, spanning the years 1971-2013. Chapter 2 briefly describes the operation of Ozlab, and Chapter 3 presents the literature review of generic WOz tools. Chapter 4 discusses how interaction is supported by WOz tools and Chapter 5 how platform dependency affects the longevity of generic tools, while Chapter 6 points to the limitations in the Wizard-of-Oz method itself from several perspectives. Chapter 7, finally, presents concluding remarks including a list of points for future methodological analysis and development.