Last year, the CST aggregate score for Gladstone High School students was in the "below basic" category in U.S History and Algebra 1, and they scored in the "far below basic" category in World History and Geometry.
The school's student body is 89% Hispanic, and 40% of the school's students are on reduced lunches. Although the school boasts a 92% graduation rate with around 30% pursuing higher education, 60% of the students participate in the English Language Learners program.
"These statistics clearly influence the aggregate results for the California Standardized Tests," Dyal said.
Dyal has been concerned that variation in class work grading standards throughout the country makes it very difficult to establish a standardized test. Students who get good grades on their class work may find that their CST scores don't measure up, or vice versa.
In the 1980s, the U.S. Department of Education did a study that examined students' grades in English and mathematics by evaluating their scores on two short tests given as part of the National Educational Longitudinal Study of 1988, according to K. Givens in a report published in Contemporary Education in 1997.
"Students in high poverty schools (schools where more than 75% of students received free or reduced-price lunches) who received mostly A's in English got about the same reading scores as the C and D students in the most affluent schools," Givens said. "The students who received A's in math in the high poverty schools scored about the same on the math test as the D students in the most affluent schools did."
And since the CST scores don't affect students' class work grades or their overall GPA, the students have little motivation to study for the tests, despite school incentives such as class parties or ideal parking spaces. Without adequate motivation to score well, the tests may not adequately measure the students' abilities.
A study conducted by Napoli and Raymond in 2004 on a community college psychology class measured the motivation of two groups of students to take an assessment test. The groups had demographic consistencies that were statistically equal, such as their age, gender, and score on the college admissions test. The first group was graded in the test, and the other group was not graded.
"[T]he graded students had an average of … 64%, whereas the non-graded students scored an average of … 42.9% … The results [also] indicate that none of the non-graded students obtained a score of 70 or above, while a significantly higher 41.3% of the graded students obtained a score of 70 or above," noted the report, which was published in Research in Higher Education.
Because of the inequity with the CST exam, Dyal said it is the teachers who feel the pressure to raise their students' scores.
"Educators … say the current method is unfair because schools don't get credit for making big gains if groups of students still fail to hit testing benchmarks," according to the Associated Press.
To address this inequity, the U.S. Department of Education announced in early December that it will adopt a new "growth model," which will reward schools based on the amount of progress their students make.
The growth model pilot was established by Secretary of Education Margaret Spellings in November 2005 and was included in the Presidents's NCLB reauthorization blueprint earlier this year, according to a press release from Spellings' office. Nine states currently have approved growth model proposals, including North Carolina, Tennessee, Delaware, Arkansas, Florida, Iowa, Ohio, Alaska, and Arizona.
"Our work on reauthorization has shown broad bipartisan support for growth models, and now, many states have improved data systems so they can track individual student growth over time," Spellings said in the press release.
"A growth model is a way for states that are raising achievement and are following the bright-line principles of NCLB to strengthen accountability," Secretary Spellings continued. "It will allow states another effective way of measuring adequate yearly progress (AYP) by measuring individual student growth over time, and it will continue to expand the flexibility available to states under NCLB."
To implement the growth model, states will need to have systems for tracking student scores which will also protect students' privacy, the Associated Press reported.
States that apply for the growth model approach must commit to ensuring that all of their students are proficient by 2014 by setting annual state goals to ensure that the achievement gap is closing for all groups of students identified in the law, according to the Department of Education's press release.
The new model may redefine improvements towards proficiency in terms that may better measure students from various backgrounds, but as far as motivating students to try their best to score well on the CST exams? Perhaps future NCLB models will address that problem.