"Everything should be made as simple as possible, but not simpler." - Albert Einstein

Seminar Announcement

Thinklink Test Validation and Alabama Proficiency Level Predictions

Basil Conway, STAT-DD-MS Candidate, Colorado State University

Thursday, December 2, 2010

1:00 pm., room 213, Statistics Bldg

ABSTRACT

The Auburn City School District in Alabama has incorporated a predictive assessment to
monitor student performance before state and national testing in compliance with the No Child Left Behind Act and high expectations for all students. This testing is to monitor and predict student performance before standardized testing and ensure teachers target student deficiencies before formal standardized testing. This predictive assessment is called Thinklink and is provided by Discovery Education. The paper will incorporate two useful analyses for the district: predictive assessment validation through an ANCOVA model with heterogeneous and homogenous slopes involving one factor and predict student performance from the predictive assessment to the standardized state assessment through multiple discriminate analysis techniques including logistic regression, linear, and quadratic discriminant analysis. Open-ended questions will arise throughout the paper and be summarized in the last chapter.

Advisory Committee:

Dr. Myung Hee Lee, Advisor, Statistics Department, CSU

Dr. Dan Cooley, Committee Member, Statistics Department, CSU

Dr. Gene Gloeckner, Outside Member, School of Education, CSU