How simple is splitting
Yet how many times do they get bundled together?
If they're mixed up you can't audit what's going on properly.
For example suppose you go to the doctor with a sore throat. What's the appropriate
observations to make? Is there a scale for 'redness' and 'swelling'
and is 'history' something that should be checked out? Now
suppose 'because there's a lot of it about' the doctor jumps to the
conclusion that you've got a streptococcus infection. That could be
a poor decision making process not supported by evidence (you
may have been shouting too much at the children). Even if the
diagnosis is spot-on that still leaves a possible failure in prescribing
the right treatment and communicating it accurately to you.
ODA applies to all decision making - especially when under
pressure. Get the facts, apply an appropriate decision making
process - now take the appropriate action and follow it through.
Another term for the same thing is Transparent Decision Making.
Splitting ODA means you can train and test on each part
separately. For example a production line might have inspection
stations. Do these test the right things and are the instruments
accurate and do results get reported in good time. That's pure
observation which can be checked with test pieces. If people are
involved the training the inspectors to measure properly is not
related to D or A. How is this data analysed and presented and
converted into 'do this' or 'do that' etc? Possibly this decision
protocol gets adapted with experience - if it does it shouldn't
involve the 'O' or the 'A'. Do the indicated actions get executed
properly? When alarm bells ring is anyone listening? When the
system fails each of O-D-A can be examined separately to identify
the root causes and in particular prevent blame-shifting.