First of all, yes, this is homework.

I'm hoping someone can direct me to a reasonably authoritative reference with some statistics relating the quality or format of the initial requirements to the time it takes to test the final product and/or the thoroughness of the testing (e.g., with requirements defined in this format, delivered software received x number of bug reports per thousand lines of code within the first year, vs. y number for requirements processed in this other way).

Any direction would be most appreciated.

-- b.r.t.