SAT Score Decline Damages College Board Credibility, Helps Rival ACT

Status: 
Archived
Subject: 
University Testing

Average scores from the first high school class taking the "new" SAT plunged despite promises by the test's sponsor, the College Board, that results from the revised exam would be comparable to those of previous versions. Scores declined by seven points on the "Math" and "Critical Reading" (formerly "Verbal") sections of the test (see accompanying chart), the biggest drop in three decades. In addition, the number of graduating seniors who had taken the SAT at least once dropped for the first time since 1990.

 

As expected, the addition of a "Writing" test narrowed the SAT's historic gender bias. The male-female score margin fell to 34 points on the combined 600 - 2400 scale. A similar drop occurred when a multiple-choice "writing" section was added to the Preliminary SAT (PSAT) to settle a FairTest civil rights complaint about that exam's use in awarding National Merit Scholarships (see Examiner, June 1999). Nonetheless, the "new" SAT remains profoundly skewed against females who continue to earn better college grades, the outcome the test supposedly predicts, than their male counterparts with higher test scores.

 

Concerns that the test revisions would exacerbate score differences between ethnic groups did not materialize. In fact, some minority groups posted slightly better results this year. But declines in the numbers of students who identified themselves as African American or Mexican American makes definitive comparisons impossible. Average scores generally rise when a smaller pool of students take a test since fewer poor performers generally participate.

 

Combined with ongoing controversies about flaws in the "new" SAT, particularly its "writing" section (http://fairtest.org/univ/newsatfact.htm), the severe mishandling of a scoring error that affected 5,000 students (http://www.fairtest.org/examarts/August%202006/SAT%20Error.html), and the rapid growth of test-optional colleges (see "Test Optional," this issue), the unexplained score drop and decline in test-takers added to test-makers' credibility problems. The College Board initially said that the score differences reflected nothing more than a change in the number of students who took the exam multiple times, thus forfeiting likely score increases from retesting. But FairTest pointed out that the College Board claim could not account for more than a small portion of the decline (http://www.fairtest.org/univ/Score_Release_2006/lttr_to_Caperton.html).

 

ACT Scores and Test-takers Rise

In contrast, the number of students taking the ACT reached an all-time high: more than 1.2 million graduating seniors in the high school class of 2006. The most significant growth was in East Coast states such as New Jersey, Connecticut and Florida. Nearly all colleges that still require test scores will accept either ACT or SAT results.

 

Average ACT score were up or stable in every demographic group (see accompanying chart). This year, however, ACT did not make public score breakdowns by family income and reduced the number of ethnic subgroups for which scores were available by consolidating Mexican Americans, Cubans, and other Latinos under the broad category Hispanic. Nor were ethnic subgroup scores made available for the four ACT subtests - English, Mathematics, Reading, and Science - rendering year-to-year comparisons impossible. Subtest data was reported by gender.

 

Barely a third of students who took the ACT signed up for its new "writing" exam - an optional section, unlike on the SAT where test-takers have no choice - demonstrating that many applicants did not need such scores to apply to colleges. According to an ACT survey, the vast majority of college admissions offices do not require "writing" test results.

 2006 COLLEGE BOUND SENIORS AVERAGE SAT SCORES
Approximately 1.47 million test takers, of whom 53.6% were female
   CRITICAL READING  MATH  WRITING TOTAL 
 GENDER
 Female  502  502  502  1506
 Male  505  536  491  1532
 ETHNICITY        
 Amer. Indian or Alaskan Native  487  494  474  1455
 Asian, Asian Amer. or Pacific Islander  510  578  512  1600
 African American or Black  434  429  428  1291
 Mexican or Mexican American  454  465  452  1371
 Puerto Rican  459  456  448  1363
 Other Hispanic or Latino  458  463  450  1371
 White  527  536  519  1582
 Other  494  513  493  1500
 No Response (9%)  487  506  482  1475
 FAMILY INCOME        
 Less than $10,000/year  429  457  427  1313
 $10,000 - $20,000/year  445  465  440  1350
 $20,000 - $30,000/year  462  474  454  1390
 $30,000 - $40,000/year  478  488  470  1436
 $40,000 - $50,000/year  493  501  483  1477
 $50,000 - $60,000/year  500  509  490  1499
 $60,000 - $70,000/year  505  515  496  1516
 $70,000 - $80,000/year  511  521  502  1534
 $80,000 - $100,000/year  523  534  514  1571
 More than $100,000/year  549  564  543  1656
 No Response (35%)  (scores not reported)      
 ALL TEST-TAKERS  503  518  497  1518
 calculated by FairTest from: College Board, College-Bound Seniors 2006: Total Group Profile Report

2006 COLLEGE BOUND SENIORS AVERAGE ACT SCORES
Approximately 1.21 million test takers, of whom 56% were female

   COMPOSITE SCORE
 GENDER  
 Female  21.0
 Male  21.2
 ETHNICITY  
 African-American or Black  17.1
 American Indian or Alaskan Native  18.8
 Asian-American or Pacific Islander  22.3
 Caucasian-American, White  22.0
 Hispanic  18.6
 Other/No Response (14%)  21.1
 ALL TEST-TAKERS  21.1
 source: ACT, ACT High School Profile Report 2006