I have Excel 2003 - but doesn't make much difference for this question.

I know you aren't supposed to do an average of averages but we do.
Here's what we do and why we do it... is there a better way?

We have a number of units that do a monthly audit of things in patients
chart.
there are 26 questions on this audit and they have to do it on 20 charts
each month.
They turn in 5 audits a week.
the questions are all answered with yes, no or n/a. examples of questions
a 1. Is the admission form completed. 2. Was the database signed by an RN.
3. If the patient's pain greater than 4, was the intervention documented.

Ok, so a score is assigned on each audit like this:
each audit is on 5 charts so there are 5 answers in each row.
If the question is answered "Yes" for each chart that would equal 100%
Each no = minus 20 so if there were 2 no's the score would be 60%
(seems like tough scoring, but all the answers should always be either yes
or n/a)
N/a does not count off anything.

then the scores for all 26 questions are averaged for the unit's average
weekly score.

These audits are done each week and at the end of the month, the average
scores for each week per unit are averaged for a score for the month. (that's
the 2nd time the numbers are averaged)

Each month the monthly averages are averaged for a YTD score for each unit.
(average #3)

Now, we aren't looking for an exact number. Just a kind of baseline score to
show how each unit is doing on these different questions - so is the
averaging of the averages a big deal?

We don't actually call them "averages" we call them "scores" and make a bar
chart as a "dashboard" of how things are going on each unit.
And in the long run, all the units come out well, but is there a better way
to do this without making it into something that is way too much work for a
simple check on how they're following procedures?

Thanks to anyone who has hung in this far :)