Difference between revisions of "Ambiguity test"

From Pearl Language
Jump to navigation Jump to search
(The beginning of wisdom is the definition of terms—Plato)
(Completeness++ + Source)
Line 8: Line 8:
|wish=You want clear, unambiguous and intelligeble requirements, expressing needs and wishes.
|wish=You want clear, unambiguous and intelligeble requirements, expressing needs and wishes.
|background=Tom and Kai Gilb talk about ''User Stories: A Skeptical View''.
|background=Tom and Kai Gilb talk about ''User Stories: A Skeptical View''.
 
'''Question''': How many words in the 'requirement', “We want the most intuitive system possible.are potentially ambiguous?
The 'requirement':
'''Answer''': ''All.''
:"We want the most intuitive system possible."
 
*How many words are potentially ambiguous? ''All.''
 
Collect interpretations, and you will find everybody has quite different interpretations, none are identical.
Collect interpretations, and you will find everybody has quite different interpretations, none are identical.
|therefore=Ask how many words in the requirement are potentially ambiguous. Next, collect and compare interpretations from different individuals.
|new=An alternative way to prove unintelligibility is counting defects in relation to the following standard using the Spec QC review method.


An alternative to prove unintelligibility is counting defects in relation to the
Rules:
|therefore=Ask how many words in the requirement are potentially ambiguous. Next, collect and compare interpretations from different individuals.
#The specification will be clear enough to test. Not later, but in itself! Now!
#The specification will be unambiguous to all intended readers, anywhere, anytime (including lawyers, and expert wit- nesses in your lawsuit).
Now using the spec, “We want the most intuitive system possible.”, how many of the words potentially violate those rules? {{author|Tom Gilb}}'s and {{author|Kai Gilb}}'s personal answer is 7, but even 1 disqualifies the spec as useful.
}}
}}
==Sources==
*[http://www.gilb.com/dl461 Tom Gilb & Kai Gilb » Gilb Papers » User Stories: A Skeptical View]


{{Source
{{Source

Revision as of 14:14, 13 February 2013

The beginning of wisdom is the definition of terms—Plato

…preparing product backlog items and other forms of requirements and specifications.

✣  ✣  ✣

{{{wish full}}}

Tom and Kai Gilb talk about User Stories: A Skeptical View. Question: How many words in the 'requirement', “We want the most intuitive system possible.” are potentially ambiguous? Answer: All. Collect interpretations, and you will find everybody has quite different interpretations, none are identical.

Therefore:

{{{therefore full}}}

✣  ✣  ✣

An alternative way to prove unintelligibility is counting defects in relation to the following standard using the Spec QC review method.

Rules:

  1. The specification will be clear enough to test. Not later, but in itself! Now!
  2. The specification will be unambiguous to all intended readers, anywhere, anytime (including lawyers, and expert wit- nesses in your lawsuit).

Now using the spec, “We want the most intuitive system possible.”, how many of the words potentially violate those rules? Tom Gilb's and Kai Gilb's personal answer is 7, but even 1 disqualifies the spec as useful.


✣  ✣  ✣

Sources