Research-Based Policy Statements on Assessment

• Testing the mean proficiency levels of students’ reading ability to determine Annual Yearly Progress of schools (as NCLB does) is not statistically defensible because it uses descriptive statistics to make causal inferences. 
(Raudenbush, 2004)

• Sample-based assessment is less intrusive and more capable of broader coverage of subject matter than continuous mass testing of all students.
 (Barton, 1999)

• The majority of existing tests have a harmful effect on the way students are taught to write.
(Hillocks, 2002)

• Assessment plays a prominent role in educational reform because tests are relatively inexpensive compared with enhancing the material conditions of instruction or improving professional development.
(Linn, 1995)    

• High-stakes testing can marginalize and contain students’ engagement with literature.
(Anagnostopoulos, 2003)

Anagnostopoulos, D. (2003). Testing and student engagement with literature in urban classrooms: A multi-layered perspective. Research in the Teaching of English  38 (2), 177-212.

Barton, P. E. (1999). Too much testing of the wrong kind; too little of the right kind in k-12 education: A policy information perspective. New Jersey: Educational Testing Service.
http://www.ets.org/research/researcher/PIC-TOOMUCH.html

Hillocks, G. (2002).  The testing trap: How state writing assessments control learning. New York: Teachers College Press.

Linn, R. L. (1994). Assessment-based reform: Challenges to educational measurement. New Jersey: Educational Testing Service.  William H. Angoff Memorial Lecture Series.
http://www.ets.org/research/researcher/PIC-ANG1.html

Raudenbush, S.W. (2004). Schooling, statistics, and poverty: Can we measure school improvement?  New Jersey: Educational Testing Service William H. Angoff Memorial Lecture Series.
http://www.ets.org/research/researcher/PIC-ANG9.html