Re: acceptable error rates

[Steven Jong <SteveJong -at- AOL -dot- COM>]

I am firmly of the opinion that quality metrics are a good thing, and that
those who decry them are shoveling against the tide. that said, I would like
to go one step further in the discussion of measuring typographical errors
(which I'd like to be measured on within a Six Sigma frame of reference,
thank you 8^):

If your company thinks that excessive typos are a problem with your
documents, then what you really want to do is to find the root cause of the
problem and eliminate it. (I think that some errors classed as "grammatical,"
such as onvious subject-verb number disagreements, are actually editing typos
and should be treated as such.) I would suggest an analysis of where the
typos are coming from. The Ishikawa ("fishbone") diagram is the tool to use.
Errors arise from the four components of output: Materials, people,
processes, and tools. You would assemble a team and brainstorm possible
causes of the error "excessive typos" using a diagram that looks something
like this (typewriter font needed below!):


          P                  P
            E                  R
              O                  O
                P                  C
                  L                  E
                    E                  S
                                           S
                                                               ___________
                   --------------------------------------------| TYPOS
      |






                                                          -------------------
                    M                     T
                 A                      O
               T                      O
             E                      L
           R                      S
          I
        A
       L

Each possible error can be attached to one of the "ribs" of the diagram,
which lets you categorize and rank the error. (Yes, I know "Dilbert" has
satirized fishbone diagrams 8^(

I lead seminars on documentation quality, and I use "typos" as an example of
Ishikawa diagrams. The possible error causes people come up with are always
revealing. Here are some that I remember; all of them are true:

MATERIAL:
Adding a last-minute change introduced typo
Writer had to transform source material that was full of typos
Source code (software) included typo

PROCESS:
No review/signoff cycle
No time to think, only time to type
No proofreader
No editor
No peer reviewer
Typo in graphic introduced by artist who can't spell

TOOLS:
No spelling checker [I am composing this on AOL, without a spell-checker]
Spelling checker included misspellings [it can happen to you!]
Document so large/spell-checker so slow writer dozes off while holding
[Return] key
Spell-checker so tedious writer doesn't use it
Display screen/zoom factor too small to see errors
Monitor out of focus/dirty/old

PEOPLE:
Writer moved from big company with editors to small company without editors
Writer has carpal-tunnel syndrome; typing is an agony
Writer's vision is deteriorating; writer doesn't know it
Writer can't spell

The thing that strikes me is how many possible error sources are out of the
writer's control. (Dr. Deming said the same about all workers.) Which of
these error sources is the most prominent in your case? How can you correct
them? (In my experience, the first reason I listed--last-minute changes
mis-entered--is the most common cause, at least for me.)

(If you run the analysis and come up with anything else, I'd like to hear
about it!)

I hope this is helpful (and on-topic!).

          -- Steve

================================================================
Steven Jong, Documentation Group Leader ("Typo?  What tpyo?")
Lightbridge, Inc, 281 Winter St., Waltham, MA 02154 USA
<jong -at- lightbridge -dot- com>, 617.672.4902 [voice], 617.890.2681 [FAX]