Summary of the New Elementary and Secondary Education Act

The key testing and accountability provisions of the new, 1000-plus page Elementary and Secondary Education Act ("No Child Left Behind Act") are in Title I, the part of the Act historically intended to improve education for "disadvantaged" children.

Each state is required to have content and performance standards in both reading and math, with assessments based on them, and add science later. The achievement standards must establish three levels - advanced, proficient, and basic – leaving a "below basic" level as well. Every state but Iowa now has standards and at least some state-mandated assessments.

By the 2005-06 academic year, states must assess each child every year in grades 3-8 and once during high school in math and reading/language arts based on the content and achievement standards. Only 14 states now assess both subjects in all these grades, of which only 9 use standards-based tests in all the required grades, while 13 assess only three times from grades 3 to 12. Until 2005-06, annual testing in reading and math once in three grade spans (3-5, 6-9, 10-12) is required. By 2007-08 states must add an annual science assessment in the three grade spans.
Commercial norm-referenced tests will be allowed if items are added to ensure the tests cover the state standards. State assessment systems that are a mix of state exams and local assessments are also allowed – Nebraska, Maine, Rhode Island and Vermont will be doing this. These assessments will be the "primary" method of determining progress toward the goal of all students reaching the "proficient" level within 12 years (by 2014).

The new law says the assessments are to be valid and reliable for the purposes for which they are used and "involve multiple, up-to-date measures of student academic achievement, including measures that assess higher-order thinking skills." The assessments also must "produce individual student interpretive, descriptive, and diagnostic reports...that allow parents, teachers and principals to understand and address the specific needs of students." FairTest and other reviews of state tests find no standardized state or commercial tests now used by any state fully meets these requirements. FairTest concludes no single, centralized state exam will be able to meet them.

Assessment results must be broken down by race/ethnicity, English proficiency, disability, economic background, gender, and migrant status. Accommodations and alternatives are to be available for students with disabilities. Limited English proficient students are to be assessed "where possible" in "language and form most likely to yield accurate data" on student achievement, until they reach proficiency in English; and to be assessed annually for English proficiency if not yet proficient.

Some funding is authorized to help states develop and administer new assessments, though the amount authorized is far less than what some experts say will be required. The shortfall increases incentives to use inexpensive, multiple-choice tests. If sufficient funding is not appropriated for test development in any year, states may postpone the required test administration. Funds are also provided for states to innovate and to develop multiple measures.

State progress will be evaluated using the National Assessment of Educational Progress (NAEP), which will be administered every other year to a sample of students in every state in reading and math. However, no sanctions apply to states that fail to make adequate progress on NAEP.

States must have a "single statewide accountability system" to ensure districts and schools make "adequate yearly progress" (AYP). Graduation rates (for high schools) and one other measure for elementary schools (at state discretion) must be included in accountability reports - but will not affect the AYP determination except that they can lead to additional schools being identified as not making adequate progress. Within each school and district, defined groups – low-income, limited English proficient, racial/ethnic minority, and special needs students – must make AYP.

For a school to make AYP, 95% of the students in each subgroup must be assessed. This will make parent and student test boycotts more difficult because schools and districts will have a powerful impetus to ensure all students take the tests. The pressure on schools and districts may also lead to further demands for high-stakes testing of students to determine grade promotion and graduation in order, for example, to ensure students take the tests seriously.

The accountability process begins with data from the 2001-02 school year, the base year (except for schools already under improvement or correction status). Within 12 years (2014), all students are supposed to attain the proficient level on the various state assessments. All students means all except those with severe special needs and students who have not been in U.S. schools for 3 years and who have LEP status.

States determine their starting point by choosing the formula that produces the higher of either the percentage proficient in the lowest performing demographic group, or the percentage proficient in the school that scores at the twentieth percentile among all schools in the state. For example, because some demographic groups have fewer than 20 percent at proficient, Mass. will use the latter formula. Based on last year's tests, 27% of the students at the school at the 20th percentile reached proficient (tests results from this year will be used to determine the actual starting point).

The gap between the percentage of students who are at the starting point this year and 100 percent must be closed by an average of 1/12 per year. For Mass., that means 100 less 27, or 73, divided by 12, or a bit over 6 percent per year more students must reach proficient in every school and district. A quick look at the Mass test (MCAS) results shows this has not been happening.

States can use multi-year averages. The first target is in two years, and subsequent targets must be set at intervals of no more than three years thereafter. In Mass, there will have to be about 12 percent more students reaching proficient by the end of the 2003-04 school year. An additional 18 percent must reach proficient by spring 2007.

Under either formula, some schools and districts will start the process below the starting point. For example, in MA there are many schools with well under 10 percent of their students at proficient. The law is not explicit as to how they should be treated, but the MA Department of Ed. is assuming that by the end of the first two years of this program, these schools and districts must have reached the starting point plus the required gain. Using our example, a school with 10 percent at proficient would need to close the gap to 27 and add 6 percent more per year, for a gain of 29 percent more students at proficient to ensure 39 percent of its students are proficient (the state target for 2003-04). The same annual gains are also expected of schools which already have most students at the proficient level. For example, a school with 80 percent at proficient must add six percent more per year. However, MA is one of the states seeking modifications of that requirement, so that high-scoring schools would have to gain, but at a slower rate.

Results from each subgroup must be reported only if there are enough students in the group to enable statistically sound reporting. This is often taken to be 30 students, but that number is almost certainly too small to produce statistically sound data for reliably determining score gains. Since in most if not all schools, these groups contain a lower percentage of students who have reached proficient, they will have to make a large initial two-year gain, much as do whole schools that are below the starting line.

There is a "safe harbor" provision in the law that may somewhat minimize the effect of this provision, so that fewer schools will be identified as needing improvement. This provision requires that in a given year 10 percent fewer students in a group fail the test. This implies a rate of improvement necessarily faster than any possible statewide rate (which cannot be higher than a bit more than 8 percent), but would enable schools to not have to ensure all students in all the groups make that initial two year leap.

Score volatility
Small schools have very volatile changes in annual test score gains due to small numbers of students (less than 68 makes for a lot of random variation). A study of North Carolina found that there is far more "noise" than real change in student learning in the data on annual school score changes. (See Summer 2001 FairTest Examiner, Presumably, this fact is also true for groups of students within a school. Further, the research found that schools that are more integrated are more vulnerable to not making AYP, meaning segregation is rewarded.

Congress authorized multi-year averages in part to solve the problem of erratic annual data. However, the research suggests that even three-year rolling averages will only partially reduce the problem. In North Carolina, a school which desired to predict its current year reading score gains would be better off to simply pick the state's average score increase rather than to use its own previous four years of score changes.

State are to create a recognition and reward program, possibly including financial incentives, for schools which perform well. Districts will "identify for school improvement" any school receiving Title I funds that fails for two consecutive years to make AYP, and states will similarly identify districts. Students at such schools will be eligible to transfer to another public school in the district that is making AYP. Note that schools which were under improvement mandates from the previous authorization of ESEA remain under those mandates. This means that in many districts, sizeable numbers of schools must make the transfer option available this year, and districts are required to provide transportation for them.

On July 1, the Department of Education reported that 8652 schools now face some form of sanction based on earlier, less stringent, testing requirements. This is about 9 percent of the nation's approximately 91,000 schools. Because these are based on each state's tests, the number of schools listed in each state varies wildly. Michigan has the most (1513), though its students score above average on NAEP, while Arkansas, near the bottom on NAEP, has no schools facing sanctions.

Some states are considering lowering the definition of "proficient" in order to avoid massive numbers of schools facing sanctions. Connecticut, for example, has created a new definition of "proficient," ensuring most of its students already meet that standard. However, many urban districts are quite certain to have a large share of their schools facing "corrective action."

An identified school (or district) must craft an improvement plan listing "specific, measurable objectives" for ensuring all students in each group make AYP. Once approved, the school will immediately implement its plan. At least 10% of the school's funds are to be spent on professional development.

Local education agencies are supposed to provide technical assistance throughout the plan's duration (as will the state for an identified district), but funding is not likely to be sufficient. A study of the previous ESEA found that half the 9100 schools identified as underperforming received no extra assistance.

If a school does not make AYP after one year under its improvement plan, the district will make tutoring available. This can include parent choice among district-approved private companies such as Kaplan or Sylvan. District are required to spend up to 20 percent of their Title I funds for transportation or tutoring programs.

If after the second year of the improvement plan the school is not making AYP, the district will identify the school for "corrective action" and take at least one of the following steps: "replace the school staff"; "institute and fully implement a new curriculum"; "significantly decrease management authority at the school level"; "appoint an outside expert to advise the school"; "extend the school year or school day for the school"; or "restructure the internal organizational structure of the school." If a district identified for "improvement" by the state does not make AYP in two years, it too faces "corrective actions" that can include a state takeover.

If after one full year of corrective action a school continues to not make AYP, the school will undergo "restructuring," to include at least one of the following: reopen the school as a "public charter school"; replace "all or most of the school staff"; enter into a contract "with an entity, such as a private management operate the school"; turn "the operation of the school over to the State;" or "any other major restructuring of the school's governance arrangement that makes fundamental reform." The last, vague option was possibly added because none of the listed options has a track record of success – even at raising test scores.

If a school (or district) identified for improvement, corrective action, or restructuring makes AYP for two consecutive school years, the requirements for improvement will cease.
- Regulations on implementing the law are to be completed by the start of 2003, and have been drafted for the assessment provisions of the law:
- The full legislative text can be found at