To Test or Not to Test: Questions on Standardized Testing
The high stakes and high costs have parents and educators wondering if we’re focused too much on testing and not enough on teaching.
Since No Child Left Behind in 2001, standardized testing has become increasingly excessive in our school system. The high stakes and high costs have parents and educators wondering if we’re focused too much on testing and not enough on teaching. In 2011, a cheating scandal was exposed involving 178 Atlanta Public Schools’ teachers and principals. While admittedly one of the largest, it was not the first nor the last in response to the increasing pressure placed on schools by standardized tests. So… to test, or not to test? That is the question.
The High Stakes of Standardized Testing:
- school funding
- teacher pay and employment
- grade promotion
- college admission
The High Cost of Standardized Testing:
- cost of standardized tests in all states combined:
- 2001 before No Child Left Behind: $423 million
- By 2008: $1.1 billion
- narrowing of curriculum/“teaching to the test”
- larger negative effects on disadvantaged students
- test anxiety
- low teacher morale
- cheating and policy change to increase scores
History of Standardized Testing
- Mid-19th century: Access to education was expanded beyond the American elite to the general public. Standardized testing was used to:
- efficiently and uniformly test students
- classify students
- evaluate progress of students and schools
- 1845: First reported use of a written examination was administered by policymakers in Boston public schools and the surrounding areas for comparison. Policymakers used the low test scores to shame and fire educators.
- 1938: Kaplan Inc. was founded as a tutoring service for tests. At the time, the SAT was unfairly used to exclude Jewish applicants from some colleges. Stanley Kaplan, son of Jewish immigrants, cracked the code on the SAT and coached his students on the ideology behind the test, giving them a better chance at a higher education.
- 1967: Hobson v. Hansen was filed on behalf of a group of Black students in Washington, DC. The court ruled that the policy of using tests to assign students to tracks was racially biased because the tests were standardized to a White, middle class group.
- 2001: President George W. Bush signed the No Child Left Behind Act into law. This tied federal funding of schools to standardized test performance and annual academic progress.
- April 2014: Hundreds of principals, teachers, parents, and students protested the state standardized tests on the streets of New York City.
- May 2014: A group of teachers filed a law suit against the Houston Independent School District over a policy tying pay and employment to standardized test results.
The Declining SAT Scores
Founded in 1926, the SAT is a measure of college readiness that serves as one of the leading considerations for a student’s admission to college. As nation-wide SAT scores continue to decline, research calls into question the efficacy of high-stakes testing.
Averages SAT Scores of College-Bound Seniors, 2003 – 2013:
|Year||Critical Reading||Mathematics||Writing||SAT Combined Score|
*In these years the SAT was on a 1600 point scale. We scaled them up to the current 2400 point scale for comparison.
The Racial Bias of the SAT
Research shows that the SAT might unfairly put minority students at a disadvantage.
Every year on the SAT there are a group of experimental questions that don’t get scored. Instead, they are analyzed to determine if they will be included in the next year’s questions:
- If the question results change the SAT bell curve, the question is discarded.
- If the question maintains the SAT bell curve, it is included in the next year’s test. The SAT has had the same bell curve since 1926 [example of the bell curve here: http://www.testmasters.net/satabout/scoring-scale].
- Since over 50% of SAT takers are White [see chart below], this methodology ensures that each year’s tests will have even more questions that Whites answer better.
- In 2012, an analysis of 2 years worth of experimental questions revealed that none of the 156 questions that were included in the SATs were questions in which Blacks answered better than Whites — even though Blacks did answer better than Whites on many experimental questions. When Whites didn’t perform well on these questions, they had a greater effect on the bell curve, and the questions were therefore discarded.
- This resulted in:
- very few questions that are neutral (all demographics answer it consistently good or bad).
- many questions that are biased (different demographics answer it differently).
- Whites having an advantage since they affect the bell curve the most:
SAT Averages and Percentages of College-Bound Seniors by Race/Ethnicity, 2013:
|Percent||Critical Reading||Mathematics||Writing||SAT Combined Score|
The Economic Bias of the SAT
There is a strong correlation between family income and SAT scores. Possible causes:
- access to test preparation materials
- condition and quality of schools
- educational attainment of parents
SAT Averages of College-Bound Seniors by Family Income, 2013:
|Family Income||Critical Reading||Mathematics||Writing||SAT Combined Score|
|$0 – $20,000||435||462||429||1326|
|$20,000 – $40,000||465||482||455||1402|
|$40,000 – $60,000||487||500||474||1461|
|$60,000 – $80,000||500||511||486||1497|
|$80,000 – 100,000||512||524||499||1535|
|$100,000 – $120,000||522||536||511||1569|
|$120,000 – $140,000||526||540||515||1581|
|$140,000 – $160,000||533||548||523||1604|
|$160,000 – $200,000||539||555||531||1625|
|More than $200,000||565||586||563||1714|
The Gender Bias of the SAT
High school GPA is usually the best predictor of college performance. Women have a higher GPA in high school than men, and they also have a higher college graduation rate. However, women score lower than men on the SAT:
|Gender||High School GPA||College Graduation Rates||Critical Reading||Mathematics||Writing||SAT Combined Score|
The Movement from Standard to Optional
Over 800 four-year colleges and universities have instituted “test score optional” policies in response to the research on standardized tests.
There is no significant difference of college GPAs and graduation rates between those that submit standardized test scores and those that choose not to:
|High School GPA||3.35||3.28|
- Non-submitters earned GPAs only 0.05 lower than submitters.
- Non-submitters have graduation rates only 0.6% lower than submitters.
Currently, 571 organizations and 18,955 individuals have signed a National Resolution on High-Stakes Testing. They are calling on policymakers to reduce testing mandates and reexamine public school accountability in order to improve the quality of learning in the classroom and reduce the negative effects on students, teachers, and campuses.
It’s time to take a time out from high-stakes testing!