Date: 2/18/93
     Time: 1:30 pm - 3:30 pm
  Subject: Minutes of RAID QIT Meeting
 Recorder: Andy Oakland

Attendees: Adriana Christopher
           Andy Oakland
           Gordie Noseworthy
           Harold Pakulat
           Naomi Schmidt
           Rocklyn Clarke
           Ron Parker

We began by presenting the results of our information-gathering calls
in the past two weeks.  These data were generally in accordance with the
earlier data we had gathered; i.e. IBM mainframe clients seem to be
generally satisfied with the level of support they receive, and there
do not appear to be any systematic flaws in the way IBM problem reports
are handled.

[To avoid making this mail message too enormous, the aggregate data and
simple ASCII graphs of it have been sent in a separate message.]

Many interviewees spontaneously commented that IBM support had recently
improved, with about one year ago being the turning point.  This was
independently mentioned by people spoken to by at least four different
team members.

Adirana wondered if the lack of dissatisfaction with problem handling
could be due to the fact that the users we polled simply don't report
problems very often.  Gordie countered this, saying that these are
"key" users, who are familiar with the system and wouldn't be asking
trivial questions.  Furthermore, as we found about twenty users who
report problems at least once a month, it seems likely that any systematic
flaw in the problem-handling system would have been reported by them.

We speculated as to why this QIT had been formed, given that the vast
majority of IBM clients seem contented.  One possibility is that there
were a few dissatisfied, but very vocal, clients, whose complaints
gave the impression that there was something wrong with the entire system.

Naomi pointed out that, as there is no systematic problem for us to
investigate, there is no root cause to search out.  The assertions made
in the original charter are not supported by the data we have gathered,
and it appears that this perceived problem is not one suitable for TQM
techniques.

The next step is for us to write and present a paper describing what we
have discovered in our investigations.  Naomi and Adriana are going to
circulate some sample TQM reports to use as guidelines.

One lesson we wish to pass on is that QITs must not be formed haphazardly.

Adriana will be archiving our raw data, in case any questions arise
concerning it in the future.  Team members are asked to send her an email
message containing their data by Monday the 22nd.