Filename Statistics and education August 15
Date: 30 August 15
By: Frank Gue
For: SQE Board
Re: “How to” support our mission statement of
We quickly dropped the subject of statistical process control as a possible “how to” support our very good mission statement. That was a mistake.
You will recall the bar chart I circulated long ago showing that the poorest performing (by EQAO) five schools delivered less than half the value per dollar that was being delivered by the best five, and delivered only 64% of the average of all 25 schools in the sample (which was a patch of schools of nearly identical socio-economic status here in Burlington).
That bar chart was derived from a classic bell curve, condensed to emphasize the potential it presents to us.
With no apology whatever for using an industrial analogue, let me tell you what I did in a similar situation in heavy machinery manufacturing:
In our steel fabricating shop we had a frightful clutter of parts, only a few of which were needed today. In our Ship On Time project, we had the computer tell us the standard deviation of our departures from schedule. We identified the workstations (a steel shear is a workstation) with the widest deviations. We improved material flows there, but more importantly we made it clear to all concerned that they had to stick to schedule sequence; make only what’s needed tomorrow, make nothing not needed tomorrow.
· Magically, half of the parts vanished from the no longer cluttered floor.
· Labor cost in the assembly department dropped 20% because they spent less time hunting for and moving stuff. (How do you “lose” a 20-ton fabrication? Easy – you lower a 30-ton fabrication over it to save space and then go off sick.)
· Sandblasting and cleaning costs dropped sharply because nothing had to be stored outdoors to rust.
· On-time performance rose from 60% to 90%.
OK, now where is the education analogue?
· Work with the schools delivering poor value-per-dollar; ignore the schools doing best, they’ll continue to do fine.
· Transfer technology (not only computer technology!) by learning what the best schools do that the poorest can do too.
· And so on.
Results to expect:
· Higher average because the low tail of the bell curve gets cut off.
· Far fewer kids left behind
· And so on.
The above is not meant to displace any of our other good “how” points but, leaping ahead (sorry, Cal), I suggest that this will turn out (in our prioritizing process) to be a must, not a want. Meantime, let’s think about it.