[Local-Maine-Schools] statistical malpractice

[Gail Marshall]

Brian,
I ask you consider for a moment your assertion regarding graduation  
rates. Public schools have to play the hand they are dealt. Some are  
in areas where there is higher income, higher educational attainment  
among the population as a whole, higher expectations of schools, and  
greater opportunity to provide the resources to make graduation a  
more likely outcome. Sure the quality of the education provided is  
critical, but a school with a higher graduation rate may be as much a  
recipient of good luck than anything else. Further, making the per  
pupil cost look artificially high by focusing on graduation rates  
sounds to me a bit like NCLB punishing schools who fail to make AYP  
by cutting funding despite the fact they have much bigger problems in  
their communities and much fewer resources than other schools. It  
strikes me that the decision to focus on graduation rates is yet  
another example of shallow bean counting rather than a reasoned  
discussion of what students need to succeed.
They are having a debate in Portland now. There are two high schools:  
Deering, in the suburbs, and Portland, in the center of the city.  
Apparently the board does a strict cost per pupil division of money.  
In my decades long memory, Deering has always been regarded the  
better school.  Teachers at Portland High argue that since their  
school is full of a greater percentage of poor kids, many of whom now  
are refugees, and/or ELL students,  and a higher percentage of whom  
bring serious problems to school every day, they must have a higher  
per pupil investment in order to succeed. Makes sense to me.
Of course this is not unique to education. There are some statistics   
that show some doctors' patients have higher mortality rates. Bad  
doctors? Maybe. But more often it's because they are the best doctors  
who take care of the most critically ill patients. Same problem.
Gail
On Apr 9, 2007, at 1:13 PM, Brian Hubbell wrote:

> Ralph,
>
> Thanks for retrieving these graphs, which certainly confirm that there
> are a lot more variables affecting the expense of education than the
> size of schools.
>
> Two things should be noted here.
>
> The first is that, for overall comparison across high schools, rather
> than simple per-pupil cost Silvernail decided instead to use the total
> budget divided by number of graduates.  This seems like a commendable
> idea because it adds in at least some measure of school performance --
> in that schools with higher graduation rates effectively are credited
> with lower per-pupil costs than those with more dropouts, as the
> expenses for educating dropouts are essentially wasted in
> inefficiency.
>
> But, perhaps even more significantly, these graphs remind that
> Silvernail's analysis here was based on schools and not administrative
> units.  Nowhere in this discussion (except for Donaldson) have I heard
> any question of the validity of extrapolating models for efficiency as
> a function of school size into models for efficiency as a function of
> school unit governance and administration.  Rather, the distinctions
> seem to be blithely conflated.
>
> I think everyone understands how closing small schools would save
> money (and also devastate communities).  But pretending that somehow
> you can save money by continuing to operate the same schools with the
> same teachers and same programs just through drawing them into
> different administrative organizations seems at best wishful thinking,
> and at worst calculated toward a different and unstated agenda that
> the real hatchet work will have to be done subsequently by others.
> _______________________________________________
> Local-Maine-Schools mailing list
> To send a message to the list, write to:
>  Local-Maine-Schools at lists.svaha.com
> To unsubscribe or change subscription settings:
>  http://lists.svaha.com/mailman/listinfo/local-maine-schools
>
>