Re: [arn-l] pre-school accountability

[George Sheridan <learn@jps.net>]

At 10:38 AM 3/17/2006 -0500, Monty Neill wrote:
> Sam Meisels of the Erikson Institute just sent me an email:
> In my continuing efforts to bring some sanity to this field, I've finished 
> another chapter on early childhood accountability. If you'd like to see it 
> or tell others about it, you can find it on our website:
>
> <http://www.erikson.edu/files/nonimages/opmeisels2006.pdf>http://www.erikson.edu/files/nonimages/opmeisels2006.pdf
>
> I have not yet read it.


Meisels is a well-respected scholar and his conclusions will be important 
in designing accountability systems for public preschools - a significant 
issue if California voters enact the Reiner Initiative for Universal 
Preschool this June. The following paragraphs are excerpted from the 
Executive Summary.

*

How did a testing approach originally developed for middle-and high- 
schoolers come to be applied to very young children? Are the results of 
such tests reliable, and if they are, can a narrow range of information 
about a single child at a particular time be used to evaluate teaching or 
curriculum? In this paper, written as a chapter in the forthcoming School 
Readiness, Early Learning, and the Transition to Kindergarten (R.C.Pianta 
et al.,Eds.), Samuel J.Meisels examines the genesis of accountability 
testing in preschool and refutes the quality-assurance, production-model 
assumptions that underlie its use with young children. Citing the best 
available research, he summarizes the arguments against such testing in 
early childhood:

·       the practical problems of measuring the developmentally unreliable;
·       unintended but real consequences for teaching and learning;
·       the failure of such tests to account for tremendous differences 
across the preschool population in prior opportunities to learn;and
·       the demonstrably weak association between academic/cognitive 
measures in preschool and like measures in first and second grade.

Meisels goes on to examine how each of these facts or circumstances 
contributed to the failure of Head Start ?s National Reporting System, one 
of the largest-scale examples of early childhood accountability testing to 
date. Finally, Meisels takes up the question of how to measure program 
effectiveness and program quality. He argues for program evaluation: 
collecting data on

·       structural and dynamic characteristics of programs (child-staff 
ratios, staff training, developmentally appropriate practice, positive 
interaction between children and staff, parental involvement,etc.),
·       key demographic variables, and finally,
·       programs? impact on children.

To measure the latter, Meisels proposes creating an assessment based on 
item response theory (IRT), using a metric that describes children's 
relative position on a developmental path. Such an assessment will not only 
indicate whether children are learning. It will enable the analysis of 
program elements, pedagogical techniques, and child outcomes to determine 
whether particular aspects of a program or child and family background are 
more or less strongly associated with child outcomes. By learning what 
works for whom, we can move beyond simply identifying a particular program 
?s outcomes to determining what we can do to help that program ?and the 
young children it serves ?succeed.



Knowledge which is acquired under compulsion has no hold on the mind. 
Therefore do not use compulsion, but let early education be rather a sort 
of amusement.

        - Plato