Standardized Tests are Discriminatory
The SAT is biased and academically unsound. Its practical effect is to artificially reduce the number of Latino, black, Native American, new Asian immigrant, other underrepresented minority, and poor students of all races who are eligible for admission to selective colleges. Motivated, talented and intellectually capable people are excluded because of this unfair admission criterion. There is no difference in intellectual capacity between the so-called races – and educational institutions should not base their admissions on a system that simultaneously expresses, exacerbates and ratifies racism and inequality.
In a study of thousands of applicants to UC-Berkeley’s famous Boalt Hall Law School, researchers found that black students with the same grade point average in the same major in the same college scored 9 points less on average on the LSAT than their white counterparts –demonstrating that there are great racial disparities in the test scores of students who perform identically in a college setting.
The tests also discriminates on the basis on family income. SAT-takers score on average 30 test points higher for every $10,000 in parental annual income. However, disparities arising from socio-economic background are overshadowed by disparities based on race. Black students from wealthy backgrounds score lower than white students from poor families.
The SAT Does Not Predict Academic Success
The Universityof Texassystem’s abandonment of the SAT requirement for the top 10 percent of every high school (according to GPA) has led to an increase in academic success. In 1997, the UT system began admitting the top 10 percent of each Texashigh school, regardless of their SAT scores. Since then, the average SAT of these top 10 percent students has gone down from 1242 to 1212, but their average first-year GPA has risen. Also, these top 10 percent students, in terms of academic performance in college, have outperformed non-top 10 percent students with SAT scores that are 200 to 300 points higher. 
In “The Shape of the River” former presidents of Harvard and Yale Derek Bok and William Bowen present extensive data which demonstrates that underrepresented minority students who have low SAT scores and gain admission to and attend a highly selective university go on to graduate and professional schools at comparable rates to their white or Asian peers. Latino and black law school graduates who attend a top-tier law school are able to make the political and social contacts necessary and to gain the confidence needed to seek and attain positions of power in this society once reserved exclusively for white men.
A study of the period 1968-1987 conducted by UC-San Diego School of Medicine professors shows that initial gaps in performance between students admitted under the regular and affirmative action admission standards narrowed and then disappeared over the course of medical education.
Until the early 1970’s, female students outscored male students on the verbal section of the SAT; the verbal section was then “balanced” to produce minimally lower scores by women then men. The math section, which has consistently reflected a significant gap in favor of men, has never been “balanced”.
The purveyors of the SAT themselves admit that test scores do not correspond to general academic or intellectual ability.
Disparities in SAT scores used to reinforce myth of racial inferiority
Despite all the research that demonstrates the racial bias of the SAT, the mistaken equation of SAT/ACT scores with “academic merit” not only deprive Latina/o, black and Native American students of their right to be admitted to our nation’s flagship universities – they are actively encouraging the development of a racially hostile campus climate for the few minority students who do enroll by promoting the racist view that these students don’t belong on these campuses (hence the frequent question – “what sport do you play?” – endured by even the most unathletic black or Latina/o student) and the subtle and not-so-subtle presumption that we are not the intellectuals equals of white students.
Can the differences in average scores be explained by socioeconomic status alone?
No. Income plays an important role, however, there are persistent, substantial score gaps between the races and between women and men, even when the socioeconomic status of those groups is constant and controlled for. The SAT in practice functions as both a racial and a class barrier to higher education. In 1997, black students from families with incomes between $80,000 and $100,000 scored lower on the SAT than did white students from families with incomes of less than $10,000.”
Don’t the score differences simply reflect differences in academic preparation? Shouldn’t we focus on equalizing K-12 education?
Differences in average test score by group reflect more than unequal access to academic preparation. Women, for instance, as a group score lower than men on the SAT, but do not receive less formal “academic preparation” in K-12 education. Additionally, the inadequate academic preparation that results from attending an embattled, impoverished school should not be the occasion for punishment by the SAT, or any other standardized test. We should strive to achieve equality and integration in K-12 education; this is in no way counterposed to eliminating the racial barrier that is the SAT.
How to explain the racial gap in test scores?
The differences in average group scores reflect the complex racism and inequality of our society. Many factors that correlate and are bound up with race and racism play a role, from segregation in K-12, to stereotype anxiety.
Because of the method of test question selection, particular questions from the “experimental” (non-credited) portion of the SAT on which black and Latino students score higher than whites are, by definition, questions not chosen to be part of the actual SAT. The questions deemed by the test writers as effective are those questions that tend to be answered correctly most often by those individuals who tend to score well on the test as a whole. The test writers know from their collection of basic demographic data that there are wide, unexplained group differences in performance on particular questions. Questions that tend to be answered correctly by a minority of individuals who, taken as a whole, tend not to score well, are questions that are not used in the actual SAT. In other words, experimental questions (questions that are not scored but may be used on a future test) that black, Latino, and other underrepresented minority students do better on than whites are automatically discarded as statistically inconsistent. This self-referencing test development method hurts demographic groups who have scored worse on average in the past. While not the product of conscious racist bias, the discriminatory impact of this process is very harmful and very important.