Reading Between The ACTs

The new ACT 2007 College Readiness Report, released August 15, congratulates American educators once again for improving student scores “on all four subject-area tests: English, mathematics, reading and science.” A closer examination of the data reveals that of the approximately 1.3 million students took the curriculum-based, national ACT college placement exam this year, average scores have been increasing incrementally within each subject. Between 2003 and 2007, average composite scores increased .4 points on a 36 point scale, from 20.8 to 21.2 points.

African-American composite test scores, which constitute about 11.7% of the 2007 sample, have wavered between 16.9 and 17.1 points since 2003. ACT Readiness Benchmarks calculate how likely a student is to succeed in college, and provide a 50% probability that a student will garner a B or higher in a related college course, or 75% probability that the student will get a C or higher. African-American test-taker readiness proved consistently low, ranging from 5% of African-Americans meeting Science benchmarks to 37% of African Americans demonstrating college-readiness in English. Caucasian scores ranged from 33% readiness in Science to 78% readiness in English.

These test results seem to indicate that African-Americans face seemingly insurmountable obstacles to college success. African-Americans, of whom only 3% passed ACT benchmarks in all four categories, have 1.5% chance of receiving a B or higher in all their core college subjects. They have a 2.25% chance of getting a C or higher in their core coursework.

Nonetheless, the ACT press release asserts that “Scores have trended upward for nearly all racial/ethnic groups since 2003. . .All groups with the exception of African-American graduates posted an increase on their ACT composite score this year compared to last…” It does not mention, however, that minority test-takers have systematically lower scores than Caucasian students. Three percent of African-Americans were college-ready in all subjects, while 11% of American Indians and Alaska natives were college-ready in all four categories. Similarly, 10% of Hispanic test-takers, who composed 7.2% of the sample, were college ready in all categories. Thirty-seven percent of Caucasians, about 59.9% of the sample, met each college-readiness benchmark.

Only Asian-Americans and Pacific Islanders escaped the negative score effect, with 32% of this demographic passing college-readiness benchmarks in all four categories. Asian- Americans and Pacific Islanders surpassed every other ethnicity—including Caucasians— in every college benchmark except reading, even though they only composed .03% of the sample.

Despite dramatically poor scores, with only 23% of all test-takers passing college-readiness benchmarks, Richard L. Fergusan, ACT’s chief executive officer and chairman of the board, asserted “These upward trends show more students are graduating from high school with the academic skills they need to succeed in college and workforce training programs.” The national press release also notes that “core courses offered in U.S. high schools, by and large, still need more rigor” to prepare students for college.

Actually, even if the number of students passing ACT benchmarks in all four categories doubled to 50% readiness, the average test-taking student would have a 25% probability of getting a B or higher in all their core college subjects. At current levels, they have an 11.5% of getting a B or higher, and a 17.25% chance of getting a C or higher.

The ACT 2007 College Readiness Report reveals that the number of students passing college-readiness benchmarks has increased since 2003 from 20% to 23% but the ACT board may be violating its own scientific standards by emphasizing steady improvement. Page 2 of the 2007 report reminds educators to “focus on trends (3, 5, 10 years), not year-to-year changes. Such changes can represent normal—even expected— fluctuations.” However, the ACT “Policy Alert Message,” which is emailed out to educators, prominently highlights a graph demonstrating the purported four-year “increase” in readiness among all four categories. This graph, located to the right of improving math and science scores, is actually flat in all but one year. Only 2007 data shows an increase in scores for this four-year bracket. The real, demonstrable increases occur within each subject, rather than across the board.

If taken out to the five-year level, the percent of students reaching ACT benchmarks in all four categories between 2003 and 2007 increases sporadically with 20%, 21%, 21%, 21%, 21%, and 23% annual readiness, respectively. But if these results are measured at a 3- or 4-year bracket, or with 2003 data omitted, the trend disappears. Some would consider such variability far from conclusive.

Bethany Stotts is an intern at the American Journalism Center, a training program run by Accuracy in Media and Accuracy in Academia.