RE: procedures for procedures?

["Edwards, Richard W" <Richard.Edwards@PSS.Boeing.com>]

One approach that I have used to good effect is the "intelligent idiot"
approach. I don't think I would write a nuclear power plant start-up
procedure this way, but I find it useful for most radiation safety and
laboratory procedures. It is particularly effective when documenting
something for the first time or when writing a procedure from scratch. I
work it something like this.

1.  Get an "intelligent idiot." This is someone who is generally
knowledgeable in fields related to the process or task but not
specifically trained in its performance.

2.  Teach this person the task. Answer the inevitable flood of
questions.

3.  Wait a day or more. Answer any additional questions that may arise.

4.  Have them tell you what to write.

     a.  Focus on the body of the procedure rather than equipment lists,
responsibility lists, etc. Add to these as those things come up. (Note:
I use a lot of leading questions like "Okay, how do I do that?")

    b.  If what they say seems wrong or meaningless, stop and get the
correction or clarification in right away. This will probably mean a
fair amount of discussion.

    c.  If they miss something, toss in a pointer and go on. (This tends
to cover a lot of those "what if" scenarios).

    d.  Go back and fill in the pointers in the same fashion. (Note: I
tend to omit the delay long delay between teaching and scripting and cut
it down to a lunch or coffee break instead.)

5.  Test the newly written procedure by following it verbatim and seeing
if it works.

6.  Fix the errors and omissions.

7.  Clean up the format.

8.  Print the procedure and send it out for peer review or signature or
whatever else it needs.

There's one element that I haven't addressed...the use of regulations,
standards, technical manuals, internal guidance and all those other
things that help define what and how we do things. I've tried addressing
those during the teaching step (step 2) and giving them out before and
after that step. None of these seems to be consistently better or worse
than the others.

Rick Edwards, Analyst
The Boeing Company
richard.w.edwards@boeing.com