California Standards Tests (CSTs) Leave Most Children Behind
Susan Notes: We need more of this kind of research--evidence from the field that the Standardistas' measurements are corrupt.
NOTE: The authors sent their study to all 50 CA County Superintendents and received not one return email.
By Gregg Nelsen and DeWayne Mason
"Far Below Basic." That's our rating of California's Standards Tests (CSTs)-measures that annually lead hundreds of schools to receive highly questionable sanctions and unwarranted community criticism. Recent statistical analyses, slighted by State Superintendent Jack O'Connell, show most CST passages far above grade level in readability-underestimating learning and, for most California students, setting unrealistic if not miraculous expectations.
Every year since they were first administered, most teachers we know have judged CST readability as too difficult. Despite sharing our confidence that the Department of Education had conducted proper readability studies, teachers' concerns persisted. When released CST items were placed on the Department's website, we analyzed the 2003 and 2004 Language Arts tests, focusing on grades 2-6-where seeds of academic self-concept and motivation are most sowed. We used the Flesch Reading Ease Scale and the Flesch-Kincaid Grade-Level Scale-highly reputable measures.
We were shocked that 81% of the 36 passages showed readability indices that were above grade. Fifty percent exceeded grade by one year or more; and nearly 20% were above grade by at least two years. Shamefully, several passages surpassed grade by three years or more-setting up most children for failure.
Informed of our findings, O'Connell conceded that reading levels were "sometimes slightly above grade." He reported the Department's contractor ensures "...passage difficulties remain as constant as possible year-to-year." Disturbingly, he also noted that "due to the rigor of California's academic standards, the types of reading used on the tests, and the passage details required to provide sufficient material on which to base questions, the readability levels of the passages will sometimes be challenging to many students."
O'Connell's responses are disappointing, deplorable, and defenseless. As administrators we have defended our State's "world-class" standards and "one-size-fits-all" assessments. We have also defended, albeit reluctantly, a State accountability program that tends to sanction schools with large numbers of impoverished children, children who originally spoke languages other than English, or both.
We cannot defend, however, the State's use of CST reading passages largely above grade. Nor can we defend our State's stubborn stance to maintain test passage "constancy" or their apparent aversion to revising standards or test specifications. Well over 25% of California's students struggle to learn English, more than 10% endure one or more handicaps, and 49% face socio-economic disadvantages. Can anyone who understands student motivation provide a defensible rationale for these mismatched measures?
Creating a State program of educational accountability is a complex process that deserves careful construction. We're confident the Department considered many factors and used various statistical criteria in developing CSTs. Any assessment, however, deserves ongoing scrutiny and, when problems are identified, revision. Because revision in this case is needed, we recommend CDE take three actions: (1) when selecting test passages, use a readability scale (e.g., the Flesch Scales we used accompany Microsoft Word.); (2) select passages within a reasonable grade-appropriate range (e.g., plus or minus one-half year); and (3) since it has been 7 years since CSTs were developed, review and revise to get them right. Other states have made revisions, and their websites show readability indices much more aligned to our recommendations.
No good reason exists for measuring challenging standards with reading passages that are overly difficult. California's current CSTs, which frustrate students and underestimate learning, will eventually lead most schools and districts to sanctions. It's time to revise our CSTs-if not our standards, test specifications, and proficiency levels. Claiming "constancy" or portraying passage readabilities as "slightly" above grade leaves a flawed system leaves most children behind. It also leaves California's assessment program two familiar judgments: "Far Below Basic" and in need of "program improvement."
Gregg Nelsen, former teacher and administrator in Jurupa Unified School District, is an educational consultant. DeWayne Mason, former UCR professor and JUSD assistant superintendent, is an art teacher and Co-Principal Investigator of Mathematical ACTS-a National Science Foundation Mathematics and Science Partnership Grant.
Gregg Nelsen and DeWayne Mason
INDEX OF RESEARCH THAT COUNTS