[QTI] Re: Hi and some questions

[Steve Lay]

Michael Piotrowski wrote:
> Examples are helpful, but they are no replacement for a formal
> specification.  The QTI 2.0 specification describes too many
> requirements, assumptions, etc. only in informal language, and does not
> formalize them in the XML schema.  This makes validation very hard.

QTI is, unfortunately, not so easliy represented using XML as might be 
hoped.  I have no doubt that it could be done better than it is.  If 
there are specific improvements that you can see to improve clarify or 
correct mistakes then please do report them to the IMS as they do 
eventually get through to the editors for discussion/incorporation into 
the specification itself.

On your more general point, there are other technologies that could be 
employed to increase the power of the schema definition (such as OCL) 
but these approaches increase the demands on the community implementing 
the specification.

It is perhaps better to think of QTI more like a programming language 
than a simple text-markup language like HTML.  The bit inside itemBody 
is much like HTML but the surrounding bits are more like a combination 
of HTML with a scripting language.

Once you think of this as a programming language then you'll understand 
that although the schema may help you write syntactically valid QTI it 
is limited in its ability to catch compile-time errors like response 
variable names that are undeclared.  The analogy can be stretched 
further too: it is easy to write useless computer programmes which can 
still be compiled and run!

However, the language of the specification is not supposed to be 
informal any more than the language of the XML specification itself is 
supposed to be "informal".  Remember that the symbols in your XML schema 
definitions are really just a short-hand notation for their natural 
language meanings.

Steve