You may be interested in a report from NASA, that evaluates six tools, that analyze requirements written in a natural language. They spent a year evaluating these tools, using criteria like completeness,consistency, precision, etc.
The report is avaliable. Mail to:
valerie.s.jones@ivvv.nasa.gov

By the way, I really enjoyed your presentation at the Atlanta IIBA meeting a couple weeks back, and I’ll be tracking the webcasts on requirements.net as well.

http://alistair.cockburn.us/index.php/Parts:_precision%2C_accuracy%2C_relevance%2C_tolerance%2C_scale_in_object_design

Your comment reminds me of the quote/maxim:

Question: “How do you eat an elephant?”

Answer: “One bite at a time!”

I agree that we too often see or hear of the all-or-nothing or “silver bullet” approach to improving processes and performance.

Unfortunately, there is no silver bullet. Fortunately, incremental improvement - through means as small as uniformity of requirement structure - can make a big difference over time.

Spoken like a true engineer!

Thanks for visiting and for chiming in. I can’t speak with any great level of expertise on formal principles of measurement, but agree that the parallels are there; certainly that there are at least principles - if not rules - to writing requirements that help reduce requirement “errors.”