Hello Ulf, A very belated 'thank you' for this very useful reply. I'm a member of the QTI working group and will pass on your thoughts to the group as it's always very helpful to get these insights from real world users. Many thanks, and best wishes, Rowin On 5/7/06 17:03, "Ulf-Dietrich Reips" <ureips@genpsy.unizh.ch> wrote:
Hi Rowin, all,
At 15:01 Uhr +0100 5.7.2006, Rowin Young wrote:
I would also be very interested in seeing the results of your research and am looking forward to them being announced.
I'll be happy to send them.
I also wondered how many (if any) of the tools you've looked at use the IMS Question and Test Interoperability specification (http://www.imsglobal.org/question/index.html), an open standard mainly being used in eassessment but which can also be used for surveys.
We are aware of the IMS standards (and have used an earlier version in our own e-learning development, see e.g. http://www.methpsy.unizh.ch/mesosworld/archive/standards/STD-QTI.html and http://www.methpsy.unizh.ch/mesosworld/archive/leitfaden/ulf/assessmentgen.h... l ), but the IMS has a number of impracticalities and inconsistencies. In particular, none of the research on Web survey and Web experiment methodology has found its way into the standards (e.g., certain types of errors happen more easily with some response types or rendering types than others, so it is better to offer the latter in a tool). Also, the psychology and reality of Web surveying is not appropriately reflected in the standards. For example, the frequency of how often a question type is needed in creating Web surveys ought to be reflected in a tool's usability and selection of question types, but the standards are indifferent in this respect. Furthermore, response-type and rendering may theoretically ideal to be handled independently (as aimed for in the standards), but this is not the psychological reality for responders to and creators of questionnaires. They experience and prefer to handle both as a conglomerate. We can discuss further details off-list, if you like. However, we will consider the IMS in terms of the total picture of question types, layout types and response options etc. available in tools.
Best --u