SAT: The Maine Initiative

January 13, 2006

Susan A. Gendron
Commissioner of Education
Maine Department of Education
Burton M. Cross State Office Building
111 Sewall Street
Augusta, Maine  04333-0023

Dear Commissioner Gendron:

I am writing in response to criticisms from the Maine Counseling Association that oppose the use of the SAT by the state for measuring student achievement and growth, as well as student and school accountability.  The Maine Counseling Association states that such use of the SAT is in violation of the American Counseling Association’s (ACA) Code of Ethics.   The primary basis for this conclusion appears to be two contentions: (1) that the SAT has not been shown to be a valid, reliable or appropriate test for such purposes, and (2) that the SAT is “culturally biased.”  Both conclusions are inaccurate and unfounded.

The College Board has endorsed all major professional standards that govern the development and use of educational assessments.  The Standards for Educational and Psychological Testing (American Educational Research Association, American Psychological Association, and National Council for Measurement in Education, 1999) are the definitive professional standards concerning the appropriate use of tests.  A second relevant document is the Code of Fair Testing Practices in Education (Revised, 2005; Joint Committee on Testing Practices).  

The Ethics Code of ACA does address testing issues, and is generally consistent with the requirements of the above-mentioned “Standards.”  However, it is the Standards that have been cited in federal and state legislative language, court decisions (including a Supreme Court decision) and used by researchers, educators, policymakers and other key stakeholders when reviewing the appropriate development and use of tests.  The Standards outline many requirements, well beyond those cited by the Maine Counselors Association in the ACA document.   The College Board (and its contractor ETS) have been active in the development of these professional standards and constantly work to provide evidence to support the appropriate use of their tests.

I am writing as the Vice President of Research and Analysis for the College Board.  Prior to joining the College Board, I served as the Associate Executive Director of Science for the American Psychological Association.  In that role I was project director for the current revision of these Standards and am very familiar with the requirements imposed on testing organizations and test users.  I have also been one of the dozen or so scientists involved in the development and revision of the Code of Fair Testing Practices (available at NCME.org).  The College Board adheres to all professional standards, and the concerns that the SAT is not valid and reliable for the uses intended by Maine are unfounded and ignore the scientific evidence that has been accumulated on the SAT over decades.

Hundreds of validity studies have been conducted by the College Board, ETS, and independent academicians and researchers that consistently demonstrate that students’ performance on the SAT is related to: (a) high school grades and GPA, (b) the rigor of courses taken in high school, (c) students’ self perceptions of academic ability, (d) freshman and cumulative college GPA, (e) college grades in specific courses and the need for remediation when entering college, and (f) persistence and graduation.  The most recent studies are available at www.collegeboard.com/research/home, and independent academic researchers in psychometrics and educational measurement can provide testimony to the accumulated validity evidence for the SAT with all groups of students.  The SAT is not culturally biased, and predictive studies also demonstrate that the SAT is a strong predictor of success for all ethnic and racial groups.  A claim that any measure that demonstrates differences between groups is culturally biased is clearly an uninformed and unscientific claim.  Research demonstrates large group differences between ethnic groups on educational tests from the earliest grades; substantial group differences are found on course grades, GPA, high school graduation, the rigor of academic courses completed, college grades, and college graduation.  Research has shown that group differences on all these measures, including the SAT, are largely based on differences in the educational quality, courses completed and the academic rigor of courses. 

In addition to the hundreds of predictive validity studies, content evidence has been accumulated through comparisons of performance on the SAT and other educational tests.  Studies have shown the SAT and ACT correlate at .90 or above, so clearly there is no basis for concluding these two tests differ in content or constructs in meaningful ways.  We have conducted studies that demonstrate student scores on the SAT correlate highly with scores on state scores and other achievement measures.  Alignment studies conducted by independent researchers in Vermont and Michigan have shown the SAT, as with any test, provides substantial coverage, but not total coverage of state standards.  The SAT was developed to reflect standards developed by national curriculum organizations (NCTM, NCTE) and based on curriculum surveys with faculty in high school and college to ensure the content measures skills and knowledge considered important for success in college. 

The new SAT Reasoning Test provides highly reliable measures of critical reading, mathematics, and writing.  Traditional test-retest correlations are .91-.92 for critical reading and math, and .88-.91 for writing.  ETS, which develops the SAT, uses a number of procedures to ensure that test items are fair and do not disadvantage any particular group.  We are more than willing to provide research studies and summary reports that provide evidence to support the valid, reliable and fair use of the SAT for all students in Maine. 

Sincerely,

Wayne J. Camara, Ph.D.
Vice President, Research and Analysis