Re: GOAL: All of our children above average!!

[Bob Clements <clementz@kos.net>]

Regarding a British editorial condemning testing, Brian asks, "Wasn't
that refreshing?"

My response is Yes, and No. 

Yes, some few of us will share a good yuk at confused statements like
"all above average", when what is meant, perhaps, is "all who are
capable of meeting standard, meeting it". A laugh at the expense of
another's confusion or ignorance is a cheap one, no?

Yes, we can all be pleased that some editor somewhere demonstrates an
insight Deming would hold dear. The abject ignorance of news reporters
put them high on his list of problems, no?

Yes, we might deride the classical curriculum shaped for scholarship.
But buried in the article is a similar dismissal of "practical
mathematics and science". I wonder what is  meant by the word practical,
and if it might entail an understanding of variation and the reading of
control charts. Nevermind, not important. Why interrupt a good rant on
the evils of high stakes testing. But, the problem is not so much with
testing as it is with a perversion of the purpose of testing, no?


Karl Close, chief inspector with the Gwent police in Wales asks, "Are
they just so dull they don't realize the difference between the two?"
This, about a mob who attacked the home of a pediatrician upon confusing
the word with pedophile. (http://www.thestar.com/back_issues/ED20000831/news/20000831NEW17_FO-PEDO.html)

The chief inspector  (gotta love his job title) is laboring under the
assumption that people are dull, not simply illiterate. His assumption
is not unlike that of too many teachers who find a myriad of reasons why
children grow up semi-literate, when the simple reason is that they were
not taught to read. 

In my jurisdiction fully 50% of all grade 3 students cannot read
independently at grade level. This number declines (predictably) to 40%
in grade 6. This year we'll learn how densely the mob is forming by
grade 10. Pediatricians everywhere are on anxious alert; cast bronze
sign makers stand at the ready.


Instead of spinning our wheels deploring the high stakes, can't we
develop some traction by analyzing the data? They might speak to system
performance, no? And if we are interested in system performance, getting
a handle on its current capacity seems a good place to start, yes?


Yes, yes. I know no.  
First get an aim. 

Well, assume we have one, and now we have some data. Scholtes advises,
"first plot the data". Balestracci advises more forcefully, "plot the
bloody data". And with high stakes involved you can bet it is fairly
bloody for some. I give you the damned fool pediatrician whose classical
shingle might better have read "children's doctor".


If management is about prediction, why not try predicting from the data?
Why not use the analytical lens, and look for outliers. Why not take
management on a field trip to an outlier site, a site that is likely out
of the system and offering a look back in?

Every cloud has a silver lining.
=======================================================================