The Educational Value-Added Assessment System (EVAAS) was the first
consulting group to provide value-added measures for K-12 education across the nation.
This group was formed by Dr. William Sanders at University of TN Knoxville
during the 1980s and later moved to the SAS institute in 2000. Value-added provides districts,
schools, and teachers a measure of progress for groups of students on standardized testing.
EVAAS itself provides many other metrics through a web application to all levels of educators.

Inherent in any standardized student testing data is measurement error in the
tests as well as missing test data for students. The multivariate models used in
EVAAS accommodate these two challenges in student testing data by taking into
account all available data on each individual student while not excluding students
that are missing any possible testing information. This talk will discuss the idea
behind value-added measures, the statistical methodology used by the EVAAS group,
the computational challenges of these models, and how this information is delivered
out to educators across the nation.