Reading Scores Present Complicated Student Profiles : Standardized Test Shows Disturbing and Unexplained Trend: a Drop-Off in Skills Between 8th and 9th Grades

Of all the consistent trends bubbling up in the state’s new standardized test, the most puzzling was the sudden drop in reading scores at ninth grade.

One reason for the low scores, educators concede, is that California high schools do not provide enough basic reading instruction for teenagers. After noticing a growing number of students with poor reading skills, more school districts are working on offering vocabulary drills and remedial reading programs.

But statistics from the test’s publisher indicate what might be a more complicated picture: Students taking the same test in other states show the same drop at the same grade level.

The Stanford 9 exam was given to most California public schoolchildren in grades two to 11 for the first time last spring. Full percentile scores were released this summer. Percentiles show where students ranked compared with a national group of students selected by test publisher Harcourt Brace Educational Measurements. By definition, 50th percentile is average.

In California, eighth-graders ranked at the 44th percentile in reading. Though that was below average, it was 10 points higher than the average score of ninth-graders. Scores for Arizona, Alabama, West Virginia and the District of Columbia--which all use the Stanford 9--also showed drops of several points between those two grades. And in every area but West Virginia, the ninth-grade scores were below 50th percentile.

The troubling trend has left some education officials wondering whether the test is somehow flawed.

“The pattern doesn’t seem indicative of something weird happening in the schools, but that there’s something weird happening with the test at the high school grades,” said Gwen Stephens, the California department of education’s director of standards, curriculum and assessment.

Standardized tests like the Stanford 9 are developed by selecting a pool of test-takers that accurately reflects a cross-section of the nation’s students. For the Stanford 9, the national sample was formed in 1995 by testing 250,000 students in 48 states.

Those students were chosen based on various socioeconomic, geographic and ethnic characteristics. Then, their test scores were distributed over a scale to basically set a national grading curve.

Scores of students who subsequently took the Stanford 9 were compared against this original testing pool.

Some educators said this systemwide trend of low high school reading scores suggests that the norming group could have been made up of mainly strong readers at the upper grades.

“It can happen,” Anaheim testing consultant Jim Cox said. “If you get a norming group that happens to be of strong students, it’s going to make the rest of them look worse.”

Bob Schaeffer, with the watchdog group FairTest in Massachusetts, said if that were the case, then the Stanford 9’s scoring procedure is not sound.

“That would violate the standards that the norming pool should accurately reflect the nation’s population,” said Schaeffer, the group’s director of public education. “It certainly puts the burden on the test-maker to explain this.”

Harcourt Brace officials said that they cannot explain the sudden drops in several states and that they would study them further. But they defended the test’s soundness and said scores from the other subject tests bear them out. Scores in math and language did not drop at ninth grade.

“If there were something wrong with the norming group, why did it not cause problems for students on the math and language tests?” said Tom Brooks, Harcourt Brace’s manager of applied research.

For whatever reason, Brooks added, this phenomenon has popped up in other standardized tests. Scores from a test used previously in the San Francisco and Los Angeles school districts also showed a sharp dip in reading scores between the eighth and ninth grades, he said.

California school administrators said they are working with Harcourt Brace to identify the reasons for the ninth-grade drop.

Arizona’s education officials have been puzzled by the same ninth-grade drop for the two years they have been using the Stanford 9. This past school year, eighth-graders ranked at the 54th percentile, ninth-graders 10 points lower.

“It just makes no sense,” said Kelly Powell, Arizona’s testing expert.

Arizona officials, however, added that they will continue to use the Stanford 9 because they are more interested in looking for a growth in scores than in comparing with the national average.

California school administrators also are examining other possible reasons for the poor showing at the ninth grade, saying that separate from the test, they have been noticing declining reading skills at the high school level.

About eight years ago, California high schools abandoned reading programs and, instead, mandated that all students take literature-based English courses. The idea was to expose students to classic and contemporary writings so that they could build analytical skills such as identifying themes and story inferences.

Vocabulary building and reading comprehension--two skills measured in the reading section of the Stanford 9--have been de-emphasized over the years.

The result, some educators said, has been that students’ reading skills have slipped dramatically. Simple passages are hard for some to understand.

“Our teachers were noticing that the comprehension levels of the kids were getting lower with each incoming class of freshman,” said Bonnie Barksdale, a librarian at Fullerton’s Sunny Hills High School, a high-achieving school with an International Baccalaureate program, yet one where reading scores surprisingly hovered around only the 50th percentile.

Addressing this concern, Sunny Hills administrators last year implemented in the school’s library the Accelerated Reading Program, which promotes recreational reading by giving students class credit for reading and testing well on various books.

Orange Unified also is seeing a growing number of students who do not understand what they read. Last year, the district hired private specialists to hold intensive remedial reading courses for junior high and high school students who are far below grade level.

Irvine Unified officials are offering summer remedial reading courses for students. More teachers also are being trained on how to help students with poor literacy skills.

“We have a major reading crisis in our state,” said Yvonne Larsen, president of the California Board of Education. “We just have to turn things around because reading is really the cornerstone to all academic achievements.”

(BEGIN TEXT OF INFOBOX / INFOGRAPHIC)

Sudden Drop

A nationwide look at Stanford 9 test results shows a perplexing trend of reading scores falling dramatically at the high school level. On the math section, however, scores tend to increase slightly at the ninth grade. The numbers represent percentile rankings, or how a student’s score compares against a national pool of test-takers.

Reading Scores

*--*

STATE GRADE 8 GRADE 9 Alabama 54 41 Arizona 54 44 California 44 34 District of Columbia 44 31 West Virginia 60 53

*--*

Math Scores

*--*

STATE GRADE 8 GRADE 9 Alabama 53 54 Arizona 52 57 California 45 50 District of Columbia 40 44 West Virginia 62 63

*--*

Source: Test publisher Harcourt Brace Educational Measurements