Accountability In Education

The cry of the public classroom teacher is heard throughout the United States and in many parts of the world. With all the testing performed to ensure accountability in our educational system, there are not enough hours in the day to teach. Teachers become constrained to "teach to the test," thereby impeding certain modalities of transmitting information to students, reducing the number of hours that students can take part in "unnecessary" activities like art and music, and increasing the amount of homework to fill in the gaps. What is the answer to the conundrum of maintaining standards of excellence in education without creating inordinate stress for teachers and students? Further, what assurances do we have that testing amounts to something more than making clever comparisons across States and nations? Is there a methodology of testing that leads to results that can be efficiently translated into better education for our children? The answer to these questions and more were considered at the International Symposium for Methodological Tools for Accountability Systems in Education at the Joint Research Center [JRC] from February 6-8 in Ispra Italy.

Attended by 90 international experts from 20 countries, leading minds gathered to resolve academic disputes and identify a few solid robust statistical practices upon which the widest international consensus can be reached. The sponsor of the symposium, the Unit for Applied Statistics and Econometrics at JRC was mandated to bring together a diverse team of educators, social scientists, economists and statisticians in a multi-disciplinary approach to education research. "Our Unit builds scales for the knowledge society," said Daniele Vidoni, one of the conference organizers, "and our ability to bring together all these experts in one room may well be the key to focusing what path educational testing will take and what role it will have in cultivating our children's future." "Statistical techniques which allow quality of education to be measured are available," added Andrea Saltelli, Unit Head, "but not all countries have the same experience of their use in education and training." Trevor G. Bond, Hong Kong Institute of Education, Andreas Schleicher, OECD, David Andrich of Murdoch University, Australia, Enrico Gori, University of Udine, Italy, Gage Kingsbury of Northwest Evaluation Association in the United States and several others were on hand to discuss the benefits of Rasch analysis in obtaining measures of student achievement that can be compared through space and time.

The importance of generating a multilevel modeling approach to understanding the actual influence of gender, social and family conditions in student achievement was also explored. Experts also discussed the utilization of longitudinal measures of student achievement as necessary to an in depth understanding of student progress. All these factors are expected to yield, in practice, the ability to track an individual student's progress over time, thus evaluating the quality of the education distinct from socio-economic location. Above and beyond any intellectual discussion of statistics and the accuracy and efficiency of the methodologies used, which are certainly crucial to all forms of testing, experts agreed that whatever methodologies are utilized in the coming decades, they must yield expeditious results. Children are not meant to languish while adults read test scores that require excessive time periods to decipher or incorporate into educational practices.

Adaptive and interactive testing that gives teachers direct access to the "front end" of testing results, already in practice in countries like New Zealand, may be the sought compromise that puts results quickly and efficiently into the hands of those most able to make sure that no child is "left behind.".

Denise Hummel is a native of New York, who moved to Italy with her husband and children. She directs a public relations/ communications business focused on intercultural issues.