E-Book on No Child Left Behind legislation - how NCLB Affects Your Student

No Child Left Behind

E-Book (revised May 2005)
To print this book use the "Print" feature in your browser.

Table of Contents

Introduction: The ABCs of NCLB
Chapter I: State Accountability Testing
Chapter II: The Achievement Gap
Chapter III: Annual Measurable Objectives (AMOs)
Chapter IV: Adequate Yearly Progress (AYP)
Chapter V: Unintended Consequences
Appendix: Glossary of Search Strings
Bibliography

Introduction

U.S. Public Law 107-110--called "No Child Left Behind" or NCLB--is a reauthorization of the federal "Elementary and Secondary School Act" (ESEA) of 1965, one of the last of President Johnson's "Great Society" programs. President Clinton called his 1994 reauthorization of this act "Goals 2000." When President Bush signed his reauthorization in January 2002, he called it "No Child Left Behind." Although there has been a U.S. Department of Education since 1867, the curriculum has, for better and worse, been maintained by the states Parents though have the ultimate responsibility: the most basic question is whether the federal executive and legislative initiative called NCLB promotes the general welfare, especially that of your student: It's a question that parents ask; its answer is their judgment call. NCLB is an omnibus law--covering student reading and math achievement and teacher preparation, as well as drugs, prayers, pregnancy, and violence in the schools. The focus here is on student academic achievement and the tests used to measure it. The goal here is to give parents the information they need to make informed decisions about their student's educational opportunities.

Historical backdrop: In the 1990s, state legislatures throughout the U.S. passed laws after which state citizen committees articulated performance objectives in reading, writing, and mathematics. States contracted with a test publisher to develop and standardize achievement tests explicitly based on their objectives. Individual students were tested at least once in elementary, middle, and high school (grade spans 3-5, 7-8, 10-12); Individual schools were held accountable 1 by summing or aggregating their student test scores and then calling the resulting number a "grade" or "school report card" (what was actually done with these numbers varied from state to state; a handful of states had no accountability testing programs in place prior to the NCLB). Annually aggregated student scores like these do not measure learning.

Although it does not require writing specific tests, the net effect of the NCLB is increased accountability testing:

All states now have their own standardized reading and math achievement tests:

Those states that implemented their accountability testing system after the NCLB had no need to develop separate standardized writing tests (there is nothing in NCLB, however, that prohibits them from doing so).

Within each state, the rate of student testing is increasing:

NCLB requires that each state test at least 95% of its students on its own standardized tests, and more grade-level tests will be required. By 2006-07, reading and math tests will be administered annually in each of grades 3 through 8 and, by 2007-08, science testing in each of the three grade spans (there is nothing in NCLB that prohibits states from developing and administering their new tests earlier).

NCLB requires that student reading and math test scores be aggregated by schools, which most states had already been doing. 2 It also requires that schools report their aggregated student achievement test scores by the specified demographic groups (subject matter scores are often averaged over different grade-level tests). The following paragraphs are about these achievement test scores and how NCLB uses them to make annual decisions about local schools that affect individual educational choices: quickly and efficiently find out how your state is implementing NCLB and learn what you can about those options you have.

Student achievement tests are a small part of the NCLB's broader goals: they are the sole focus here. The following text is organized around six questions about the NCLB: Each short answer contains a link to a longer response or chapter. Each chapter concludes with footnotes that examine the issues in greater detail (it's not necessary to understand the details discussed in the footnotes to understand the basics). The Appendix lists and defines helpful NCLB jargon to use as search queries for state-specific documents. It's not necessary to read the following text sequentially (although it is organized systematically): if you already know what an "Annual Measurable Objective" is, you don't need to be told repeatedly or drilled (there won't be a test).

How does the NCLB affect what's taught? Chapter I describes how the different states independently articulated their own content and performance standards--typically at three grade levels--in reading, writing, and math. The states also independently contracted with a test publisher to develop their own standardized tests of these performance objectives (some states held their test publisher to a higher standard than others). All states with such tests required that student scores be aggregated over individual schools: even if the tests are reasonably reliable and valid, doing so does not measure learning--a phenomena occurring over time within individual students.
The states are responsible for articulating their own performance objectives for the forthcoming science achievement tests: one state may have objectives about evolution and another about creationism. The states also individually define achievement levels like proficiency. To the extent that time spent being tested could be spent learning other subjects, NCLB does affect curricular decisions. It is also expected that what's being taught will center around what's being tested: in the language arts, that means more emphasis on reading and less on writing skills. The reliability and validity of the state achievement tests should not be assumed.
What problem does the NCLB address? Chapter II discusses the "achievement gap" and one possible explanation for it: a "climate of low expectations." The gap in achievement test scores is between whites in affluent suburbs and primarily everyone everywhere else. The achievement gap in test scores has been documented in numerous independent empirical studies using different standardized tests: while some argue that the gap is just an artifact of standardized tests--that such test scores signify little to nothing about actual achievement levels--the gap between average test scores is, unlike WMD, sufficiently well documented to be called fact. Everyone is in favor of eliminating this gap.
Receipt of Title I funding--over 95% of U.S. public K-12 schools receive some--is now contingent on USED approval of state accountability plans. The NCLB requires that states report their standardized student test scores by different demographic groups within individual schools. The NCLB does not address the learning of individual students within these groups: NCLB shows relative levels of achievement by aggregated demographic subgroups at given points in time.
How does the NCLB propose holding all student achievement in a given year to the same standard? Chapter III discusses "annual measurable objectives" (AMOs) and the working definitions of "proficiency" developed by the states. Each year the same AMO--unlike factory emissions, AMOs cannot be traded--is used to judge the achievement of each demographic group within the state: although comparing aggregated student scores with AMOs doesn't measure learning, it does report or show relative proficiency levels of the different aggregates at any one time. ³
How does the NCLB determine whether an individual school has met the state AMO? Chapter IV examines the options the law gives parents whose local school has been classified as not having made "adequate yearly progress" (AYP). Each state wanting to receive Title I funding subtracts its annual AMO from the actual percentage of students scoring at the proficient level at each of its schools. If a school's actual percentage proficient is significantly--from a statistical perspective--lower than the AMO, the school is classified as not having made AYP that year. The effects of not making AYP are, for the school, increasingly severe.
Why is there still an achievement gap? Chapter V lists some of the likely unintended consequences of mass accountability testing ... The goal of President Johnson's 1965 Elementary and Secondary School Act (ESEA)--and all its subsequent reauthorizations, including NCLB--has been to eliminate the achievement gap. Could it be that accountability testing in general and NCLB in particular works like kudzu--or many another creeper vine--strangling the students and schools that feed it? This final chapter examines some other likely--but, as usual, largely unanticipated--consequences of NCLB.
How does the NCLB directly impact my student? Appendix is a glossary of useful search phrases for locating public documents relative to such educational choices as you currently have: the search engines crawl state and federal public documents word-for-word so, with the right jargon as your query terms, you can get manageable search engine results pages with a dozen or so links that directly address your question. Laugh at your search engine's annoying message that you can get more results by not using parentheses as you open up one of the links on your results page in a new window and then use your software's find function--bypass irrelevant information--to zero in on the answer to your question.

Footnotes

¹ All tests are "accountability" tests: since no one spelled "publicly" correctly on last week's spelling test, the word will be on this week's test. Many tests are standardized: everyone gets the same spelling words.

There is a lengthy historical debate about the value of standardized testing. The arguments run from nothing can be measured this way to everything can (the phrase "standardized test" is often used interchangeably with "multiple-choice" test).

Since it is more expensive to score free response or open-ended questions, most state reading and math achievement tests contain primarily multiple-choice items.

"Accountability testing" differs from traditional classroom testing primarily in its use of test scores. Traditionally, test scores have been grounded in the context of the tests on which they're based.

You look at yesterday's multiplication test: the teacher wrote 85% at the top of the test and a short note at the bottom--"learn the 8s and 9s." A quick scan of the teacher's check marks on the test shows that your daughter has her times tables mastered through the 7s (all the errors occur in the test questions about the 8s and 9s): in your brief scan of the test, you've triangulated on three sources of observable information: your student's grade or score, your review of the test items, and the teacher's note: Given a goal of complete student mastery, your course of action is immediately apparent. With accountability testing, parents only receive--some months after the testing--test scores-- often broken down by the state performance objectives: these scores, particularly when accompanied by other diagnostic materials, can be informative. But seeing a score independent of the items on which it was based is not as informative as seeing both (in some cases, students--like those taking the federal NAEP tests--receive no feedback).

The goal of NCLB accountability testing is not to directly facilitate the learning of individual students but to evaluate their school and then, on the basis of that decision, perhaps make additional services available. As discussed in the following note, some states used their high school scores to sanction their students (sometimes called "high school exit exams"). NCLB contains no sanctions against individual students.

² "Norm-referenced" tests are a classroom standard: teachers "grading on a curve" assign grades based on how well different students do relative to each other. The state accountability tests are criterion-referenced: Their concern is not discerning how well students do relative to each other but with establishing how well they do relative to a predefined criterion or achievement level. Although most states used their students to produce the test scores which could be aggregated to hold individual schools accountable, some also held individual high school students accountable by making them "pass" their state writing, reading, and math tests. Passing was often defined as "minimum competency." Some states defined this achievement level arbitrarily (how about 60%? ...no ...65%? ...great!). Others defined achievement levels more systematically with a "bookmarking" process. After field testing different sets of content items, the state test publisher rank ordered each set by difficulty level (the percentage of students answering incorrectly). It presented these rank order items to the individual members of individual state citizen committees, typically in the form of a paper booklet.

Each committee member was asked to independently draw a line in his or her individual booklet of rank ordered items at the place where s/he stopped encountering items s/he thought students at that achievement level should be responsible for knowing. First-round results are then discussed, followed by more rounds of individual bookmarking and discussion (definitions between committee members can converge within as few as three rounds).

NCLB requires that states define three achievement levels for reading and math (it calls them "basic," "proficient," and "advanced").

³ Implementation of the NCLB is confounded with increases in computer hard- and software between the ESEA's previous reauthorization (1994) and the present. Most if not all state departments of education maintained a central repository of digital student records by 1994: schools then probably backed up their data on a floppy and mailed it to their district. Today's schools transfer data over the web.

Chapter I: Accountability Testing

Many state legislatures implemented accountability testing in the 1990s: their state departments of education convened committees--consisting of educators, parents, and other community members --to establish content and performance objectives in the domains of reading, writing, and mathematics. The resulting public documents or files--most states posted them online--represent each state's best attempt at articulating what it thought its students should know and be able to do. Each state's achievement tests are based on its performance objectives. ¹

Prior to the NCLB, states used scores from their standardized achievement tests--aggregated by schools--in different ways: Some states rank ordered their individual schools according to these scores. Others subtracted the previous year's test score from the current year's (and then used that number as the school's "grade"). One state erased its students' scores from its computer after having aggregated them over schools. Another state adjusted its achievement definitions for measurement error. Some states used their standardized test scores for diploma sanctioning--while others who had done so for a few years abandoned the practice. Some states began administering their tests at the beginning of the 1990s and some near the end. A handful of states had no such statewide testing system in place prior to NCLB.

NCLB requires that states wanting Title I money administer their own standardized reading and math achievement tests and publicly report the results. Each state outlined its plans to do so in an "accountability workbook" submitted to and needing the approval of the U.S. Department of Education (USED). One impact of the NCLB is more testing--new standardized tests will be needed--and more classroom time will be devoted to accountability testing (although NCLB does not require a writing-specific test). Beginning in

2002-03, the first academic year after it was enacted, states were to either continue or commence testing students in reading and math--each state using its own achievement tests--at least once in each of the three grade spans. Additionally, biannual samples of state 4th and 8th graders must now take the state and national National Assessment of Educational Progress (NAEP) reading and math tests (participation had previously been voluntary).
2006-07, students will be tested in reading and math in each of grades 3 through 8 and once in the 10-12 span.
2007-08, students will be tested in science once in each of the three grade spans.

There is nothing in NCLB that prohibits states from administering additional tests earlier. Pre-testing items for all the new tests will begin at least a year before NCLB requires that scores on them be reported. In addition to the successive biannual samples of 4th and 8th graders taking the NAEP tests, the NCLB initially called for six annual test administrations (two content domains; three grade spans). By 2006-07, it will require fourteen tests (two content domains; testing in six grades--3, 4, 5, 6, 7, and 8--and one grade span (10-12)) and, by 2007-08, seventeen tests (the preceding plus science once in each of the three grade spans). One of the controversies surrounding the NCLB is whether it adequately funds what it costs to develop, administer, score, and report such tests. Another is the instructional time devoted to testing and test preparation. NCLB does not though tell the states what to teach and what to test (USED does though tell states how to report their aggregated student test scores and how they should format their individual digital student records).

Since NCLB uses the state achievement tests- more of which are on the immediate horizon--to make annual "high stakes" AYP decisions, parents are naturally concerned about the reliability and validity of these tests: everyone knows to expect more error when measuring student achievement with standardized tests than when measuring height with a ruler. Error and test reliability are inversely related: The greater the error component, the less reliable the test--and the less reliable the test, the less sound a basis it is for any subsequent decisions (whether about individual students or schools).

The basic concern of reliability is repeatability:

do different scorers--the first pair of the state's test publisher's employees who score the item--agree how individual student open-ended responses should be scored?
are test items in a test's different versions sufficiently equivalent that a student would get the same scores regardless of which test version s/he took? are successive cohorts of students at a given grade level taking the same test?

The basic concern of validity is whether a test really measures what we think it measures. In the context of the NCLB, the question is whether the state's standardized tests adequately measure the state's performance objectives (and so whether scores on them form a basis for any kind of inference). 2 At the loosest level, a test has "face validity" if a state committee has reviewed its items and judged them appropriate for what was being tested. All the state standardized tests meet this preliminary criteria. Few states have held their test publishers to a higher standard.

The less consequential the decisions based on the test scores, the less the need to be concerned with questions of reliability and validity.3 The only systematic source of information about the individual state standardized ⁴ achievement tests are the technical reports--consisting of statistical data analyses of student scores--the test publishers have submitted to the states (most states keep at least the most recent versions online).

The NCLB addresses questions about the reliability and validity of the state standardized tests by saying they should be "scientifically valid and reliable" and that states should base instructional decisions on "scientifically based research."

What USED wants is accurate, standardized, and modern management of student records: while schools used to back up data on floppies which they mailed to their central offices, most schools' data entry programs now have web interfaces. Over the years, states have modernized their digital student record repositories: implementation of the NCLB has accelerated that process. Similarly, the NCLB is increasing the amount of time devoted to accountability testing--more than double by the 2006-07 academic year.

Footnotes

¹ After articulating its objectives, each state hired a test publisher and charged it with developing, administering, and scoring standardized tests measuring student mastery of them: Each new test begins with a "test blueprint" or cross-tabulation of the state's performance objectives by number and type of test items (multiple choice or free response).

An example is "3_Ia 5 4 1": the 3rd grade test would measure mastery of objective Ia with student responses to 5 individual test questions, 4 of them multiple-choice (students select correct answers from amongst incorrect ones) and 1 free response (they write down their answer).

² In November 2002, USED eliminated its former Office of Educational Research and Improvement--citing lack of scientific rigor--and replaced it with an "Institute of Education Sciences." The National Council on Educational Statistics--which conducts the NAEP tests--now reports to its director. Reading the new IES reports suggests that USED is reasoning that, since patients are randomly assigned to conditions in clinical trials, it is possible to randomly assign students--the treatment condition could be reading selections from the Freire pedagogy book and the placebo, passages from a CTB McGraw-Hill technical report to one of its state clients about inter-rater reliability--to gain "scientifically valid and reliable" knowledge of, to name one, best teaching practices.

³ Being reliable is a necessary but not sufficient reason for validity: Tests can be reliable (but not valid) but never the converse. Scores from a math test--individual item content ranging from single-digit addition to double-digit multiplication problems--administered to a group of K1-6 students would distinguish older from younger students--and so reliably predict which were better readers: Such a test though would not be a valid indicator of reading achievement.

⁴ USED has compared percentages of 4th and 8th graders scoring at the nationally defined proficient level on the federal NAEP reading and math tests administered every other year with the percentages of 4th and 8th graders scoring at the state defined proficient level on the state standardized tests and interpreted the observed discrepancies as problematic, indicative of a lack of commitment to eliminating the achievement gap. USED cannot, however, meaningfully interpreted such discrepancies, because performance objectives, achievement tests, and achievement level decisions vary across states. That USED has used these numbers as a weapon is fodder for students of ethnostatistics everywhere.

Chapter II: The "Achievement Gap" and the "Climate of Low Expectations"

The "achievement gap" the NCLB addressed is between the aggregated scores of white students in primarily affluent U.S. suburbs and primarily everyone everywhere else in the country. The achievement gap the NCLB addressed is the same one that led President Johnson to enact the Elementary and Secondary Education Act (ESEA) in 1965: the law's subsequent reauthorizations have also been intended to equalize educational opportunities throughout the country. The NCLB's way of addressing the achievement gap between different demographic groups is to require that states report their standardized achievement test scores by those groups (in a manner approved by the U.S. Department of Education (USED)). 1 As mandated by the NCLB, receipt of Title I funding is now contingent on ongoing federal approval of state accountability plans.

There have been many explanations of the achievement gap. The one the NCLB tries to address is a "climate of low expectations."

One problem is that some students aren't taught.

During the late 1950s left-handed K-5 students were dismissed from penmanship classes, because they were thought incapable: never taught, they never learned to write well (fortunately for them, keyboarding became a more useful skill).

Another problem is that high achieving students are not recognized as such.

In the late 1980s, the U.S.'s Educational Testing Service (ETS) accused students in Mr. Jamie Escalante's Garfield High School calculus class of cheating on its advanced math test (reasoning that students from an inner city school couldn't get scores like that). As made famous in the movie Stand and Deliver, those students retaking the ETS math test achieved scores similar to their original ones. Had those students not been from an inner city school, their scores would not have been doubted.

The most commonly proffered explanation for the achievement gap shown on most standardized tests is poverty: In addition to the anticipated effect of scarce to nonexistent resources, poverty is confounded with other variables like ethnicity and English language proficiency (e.g., almost all white students are also native English speakers; many students whose families are poor do not speak English as their first language). Confounding makes interpretation of even professionally appropriate analyses difficult. ² The most commonly mentioned causes of the achievement gap--in addition to a climate of low expectations--are:

poverty;
English language proficiency;
the curriculum wars of the 1990s: their cumulative impact on student achievement has not been positive (entire generations lost to one kind of postmodern math or another);
local schools are burdened with unfunded federal mandates;
today's teachers are unqualified (just so, state teachers think state legislators); another is that today's educational system is suffering from
too much standardized testing (and too little);
different students from different backgrounds have different learning styles that are not always taken into account or adjusted for by normative instructional and testing methods;
the growing digital divide, and
all of the above.

The NCLB plans to eliminate the achievement gap by having states annually aggregate and report their standardized test results by different demographic groups: the census 2000 ethnicity categories--African-American, American Indian, Alaska Native, Asian-Pacific Islander, Hispanic, and white--as well as gender, poverty, language, migrant, and disability status (note that these categories are not mutually exclusive: e.g., scores from female disabled students who are the children of poor non-English speaking Martian parents will be included in the tallies for all but the demographic subgroups): States do so now when "feasible."

Feasibility means at least 11 students. Before the NCLB, the federal "Family Educational Rights and Privacy Act" or FERPA (http://www.ed.gov/offices/OII/fpco) did not allow states to report scores for schools with 10 or fewer students: NCLB a priori categorizes these small schools as making "adequate yearly progress" (AYP)--to avoid conflicts with FERPA.

Reporting group percentages proficient does not, of course, measure the learning of any child within these groups. ³ NCLB's reporting procedures though will highlight annual across-group differences in proficiency.

Footnotes

¹ Comparing the scores of nationally representative samples of U.S. students to those of nationally representative samples of students in other countries shows another reflection of the U.S.' "achievement gap:" aggregated U.S. student scores do not reflect this steady progression relative to their peers in other countries (http://isc.bc.edu).

² Another example of confounding is implementation of the NCLB and the enormous leaps in computing power--increased digital data storage and management possibilities--since the ESEA's previous reauthorization in 1994 (is USED currently helping states standardize student digital records because of the NCLB--to better serve migratory students--or because doing so is now possible? Although SEC. 1308 of NCLB does explicitly instruct states to track migrant students (http://www.ed.gov/policy/elsec/leg/esea02/pg8.html#sec1308), SEC. 9631 prohibits USED from creating a nationwide database (http://www.ed.gov/policy/elsec/leg/esea02/pg112.html#sec9531).

³ When it comes to measuring individual student learning, the proper--the only useful--unit of analysis is the individual student, not some aggregate. And the proper research design is the split plot, longitudinal not cross-sectional, with individual students across time nested within schools (Hayes, 1994; Winer, 1971). Beyond NCLB aggregated numbers not measuring any learning--and knowing that looking at these demographic categories separately ignores the fact that many factors are correlated or confounded--no one should

expect a single number to provide an anywhere near a complete picture of something as complex as a school at a single point in time--let alone across time: at the most obvious level, transferring students will be counted in one school's pile of scores one year and another's the next. Others argue that
a single year is too small a window by which to judge progress: the successes of Mr. Escalante's famous Garfield High School calculus classes resulted in part from what his students had learned in earlier preparatory courses, many of which he designed and taught: the lesson is that success was not a one-semester phenomena--it took years of steady work (http://reason.com/0207/fe.jj.stand.shtml).

What the NCLB delivers is a schematic for states to report annually aggregated numbers--accompanied by the mandate that the numbers be publicly available to all parents. Breakdowns aggregated by school ignore short term demographic changes like students moving between test administrations: When annual cross-sectional school aggregates like these are used, scores from the same student could one year be added to one school's aggregate and another year, another's: a mobile student population yields an even messier pile of numbers.

Chapter III: "Annual Measurable Objectives" (AMOs)

By 2013, the NCLB requires that every student in every classroom in every school in every state score at or above the proficient level. In the interim, the NCLB requires that the states judge achievement with a series of intermediate annual percentages--called "annual measurable objectives" (AMOs). Each state compares its AMO to each year's percentages of students scoring at or above its proficient level on its tests. Central to NCLB's remedy is comparing all aggregated annual student achievement scores in each state to the same AMO--and making the results of these comparisons publicly available to parents.

To say that a state had an initial AMO of 60% is to say that it required at least 60% of its students--or demographic subgroups within schools--score at or above its proficiency level.

In a given year, the exact state AMOs will be somewhat different. Over time, they will all increment: a state with an AMO of 60% one year can't set one of 50% the next (nor can AMOs be traded). The NCLB requires that all converge at 100% in 2013 (which is, while a great goal, impossible--some student somewhere will be off a column on a bubble sheet--and will probably be repealed when the law is renewed). In the interim though the NCLB requires that all student achievement in a given year be held to the same standard or AMO.

Specific AMOs or goal percentages proficient vary from state to state

One state's initial AMO for reading might have been 60% and, for math, 55%--while another's was 62% for both; one's AMOs might increment by 5% odd years and another's, by 6.5% even years. The U.S. Department of Education (USED) would not have accepted a state plan that started with an AMO of 50%, called for 1% increases in the percentage scoring proficient between 2002 and 2012 (and then a 40% increase).

AMOs establish the percentage of students who have to score at or above the proficient level for the school to make "adequate yearly progress" (AYP) that year. If any of the different subgroups--on any of the subject matter tests--score below the AMO, the school is classified as not having made AYP that year. The actual percentages observed scoring at or above proficient are a function of several factors in addition to actual student achievement.

One factor is the performance objectives on which the state standardized achievement tests were based.
Another is how each state defined achievement: NCLB requires that each state define it at least at three levels (which it calls "basic," "proficient," and "advanced").

One state might categorize ranges of scores on its test into the three required achievement level definitions while another used four: basic, novice, proficient, and advanced. It may be that these two state definitions of "proficient" are equivalent--but not necessarily. Even when the same category labels are used, one state's "proficient" might be the label for scores above 50 but below 75 while another's might be scores above 45 and less than 80. And, of course, once "proficient" is defined, basic becomes scores below it and advanced, scores above.

In summary, AMOs are interim percentage goals. Like performance objectives, the state tests intended to measure them--and definitions of proficiency--they vary from state to state. AMOs are compared to the actual percentages of students scoring at or above the proficient level that year on that year's tests and are the sole standard to which all student achievement in the state is be compared that year. The actual percentages proficient observed are influenced by--in addition to student achievement--other factors including definitions of proficiency applied, who is tested when (near the beginning or end of the academic year), and test difficulty. ¹

Footnotes

¹ It is likely that factors other than student achievement played a role in producing the results shown in the U.S. Department of Education's table titled "Percentage of Students at or above Proficient in Mathematics, Grade 8 Public Schools: By State, 1990-2003" (http://nces.ed.gov/nationsreportcard/mathematics/results2003/stateachieve-g8-compare.asp). This table shows percentages proficient by state by year on the federal National Assessment of Educational Progress (NAEP) 8th grade math test (since the percentages reported are aggregated by state by year, these data are relevant to questions about achievement levels at given points in time but not learning, a phenomena occurring within individuals across time). This table's

rows are the individual states; its
columns are years--1990, 1992, 1996, 2000, and 2003--when the NAEP math test was administered to successive cohorts of 8th graders; and its
cells--the intersection of rows and columns--contain the observed percentages of 8th grade students who took the test that year and scored at or above its proficient level (a "--" indicates that a state didn't participate in NAEP testing that year--doing so was voluntary prior to NCLB).

Pick your state as a row and look across its columns, the five testing administrations between 1990 and 2003: notice how the percentages scoring proficient nearly always increased. Eyeball a few other states or rows: Almost all increment by a percentage point or two from one year to the next.

These percentages scoring at or above the federal definition of proficient show 8th grade math achievement throughout the country rising steadily over the last 15 years. One interpretation of this table is that math achievement in the U.S. has steadily risen.

International student achievement tests, however, show that U.S. 8th grade math achievement has not increased: year after year, U.S. students score lower than their foreign counterparts (Chapter II, note 1).

Another possible interpretation of the small but steady increases in percentages proficient shown in this USED table is that state or federal policies on who got tested changed over the years (the reported scores are from a *sample* of students within each state--unlike the state standardized achievement tests, which are administered to all).

It is unlikely, however, that changes in students permitted to take their state tests with accommodations explain the systematic increases in percentages proficient shown in this USED table: Look at the last two rightmost columns in the table, which indicate that "Accommodations" were permitted in the 2000 and 2003 NAEP test administrations but not before (the rationale for this level of detail is presumably that, historically, percentages reported scoring at least proficient have been inflated by not testing students with disabilities and those whose native language is not English).

These steadily increasing percentages proficient reported in this USED table could also be explained by older versions of the NAEP 8th grade math test having been replaced by ones containing increasingly easier items and the federal definition of what constituted "proficiency" having become increasingly less rigorous over the years. Thus, while NCLB holds all student achievement to the same annual state standard or AMO, many factors contribute to the percentages scoring proficient actually observed.

Chapter IV: "Adequate Yearly Progress" (AYP)

The NCLB's mechanism for bringing about school reform is making annual decisions about whether each school is or is not making "adequate yearly progress" (AYP). The NCLB asks both whether the school as a whole and the different demographic subgroups within it are making AYP (states typically aggregate over the specific grade level tests they administer): To make the decision, the same standard is applied to all. If any subgroup doesn't make AYP, the school is classified as not having made AYP. 1 The sanctions for not making AYP vary with the number of consecutive years a school has not made AYP.

States determine whether their schools have made AYP (and then inform parents and the U.S. Department of Education (USED)).

States subtract that year's "annual measurable objective" (AMO) from the percentage of students scoring at or above the proficient level on their subject matter tests at each school (by demographic subgroup whenever possible). If the resulting number is negative--the school's percentage proficient is less than the AMO--the state asks whether the school's percentage is significantly--in a statistical* sense--lower than the AMO: if so, the school has not made AYP.

---

* In calculating AYP, some states have set the probability of a Type I error or false positive--i.e., making an AYP decision that classifies a school that really is doing well as one that is not doing well--at 1% and others at 5%. 2 Setting the Type I error rate at the more restrictive 1% results in fewer false positives or schools mistakenly classified as not making AYP.

If the percentages proficient are significantly less than the AMO, the school has not made AYP:

A state's math AMO is 60%. School A's percentage scoring at or above the proficient level on the state math test is 59%. School B's percentage proficient is 30%. School A will probably be classified as making AYP and school B, not, whatever the confidence interval the state uses. ³

Schools can not make AYP on more than achievement indicators (state standardized tests of state performance objectives using state-defined achievement levels). Schools can also not make AYP by testing less than 95% of their students. ⁴

If over 5% of a school's parents decide to opt their student out of being tested-- some states allow parents to exempt their students from certain testing mandates--the school will not make AYP.

NCLB also required that the each state propose an additional indicator in its "accountability workbook:" many have set attendance goals--USED requires graduation rates from high schools--and a few have used their writing test results. Schools not making AYP face consequences.

The following applies to schools receiving Title I funds: the greater the amount of Title I funding involved, the greater the impact of the NCLB.

The first consequence is that, if their school has not made AYP, parents are notified (many states were reporting their school report cards to parents prior to the NCLB). If a school is judged as not having made AYP, the NCLB also requires that it use some of its Title I funding to pay for transportation--so parents can transfer their children to another perhaps better school in a neighborhood further away, if they want. The school must also develop a two-year self-improvement plan (within three months of notification that it hasn't made AYP).

Schools not making AYP two consecutive years go onto the USED's "needs improvement" list. Once a school is on this list, it must make AYP for two consecutive years--in all categories on all indicators--to be removed. Schools making AYP on everything the next year remain at the same sanction level that year, typically until the next annual testing window.
Those not making AYP a third consecutive year must provide, in addition to transportation, "supplemental educational services" like after-school tutoring (partly paid for out of their Title I funding). Each state selects its own such service providers.

Parents must explicitly agree--opt-in--to having their student's data shared with the service provider. Under FERPA, the federal law protecting student record privacy, since the provider is not considered an agent of the U.S. Secretary of Education, explicit parental consent is mandatory.

Many schools now offer supplemental services. Schools that haven't made AYP for three consecutive years must also develop a "corrective action plan" (options include implementing a new curriculum and decreasing the school's management authority).
Some schools will soon find themselves not having made AYP for four consecutive years. In addition to the preceding, they will be required to develop a "restructuring plan."
Any school not making AYP five consecutive years must implement its restructuring plan: "alternative governance" includes reopening as a public charter school, replacing all or most of its staff, and contracting with a private management company.

Schools not making AYP must inform parents. A school must explain why it didn't make AYP, what it's doing to address the academic issues, and the options from which parents can choose. If, for example, there are no schools to which a student can transfer, supplemental services are probably offered.

Footnotes

¹ USED has made several exceptions to the general rules:

Each state is allowed to administer 1% of its tests "with accommodations" (now 3% for "sincere" states).
Students learning English can be tested in their native languages for their first three years.
If its percentage of students scoring below the proficient level is reduced by at least 10%, a school that would not otherwise make AYP is classified as having made it (called "safe harbor" by USED).
Schools with less than 10 students are not given public "report cards" to preclude the possibility that student anonymity will be compromised (the NCLB categorizes such schools as making AYP).
The scores of demographic subgroups with less than 30 students cannot be reliably reported (different states use somewhat different minimum group sizes).

² The NCLB's main academic indicators are scores on the different states' reading and math achievement tests. Testing errors should be considered when interpreting these scores. One question is whether the scores have been accurately reported: Over the years, several states have reported that their publisher has incorrectly assigned scores to students. Although test publishers have quality assurance procedures, some error is inevitable, both in the reporting of scores and in the test scores themselves: In the context of annual NCLB decisions, anything contributing to a student's score that isn't related to mastery of the state's performance objectives is error. Instructions alone can be a source of error: Contrast what you'd expect as answers to questions like "circle an answer if you think it is correct" and "don't circle an answer if you don't think it is not correct." There are many kinds of testing error:

Some testing error is systematic. If a state only wants students to be able to subtract numbers and thinks what the results are called unimportant, labeling differences-- one school teaches that the amount remaining after one quantity is subtracted from another is the "remainder" and another school, that it's the "difference"--will increase error (similarly for "adjectives" and "modifiers").
Some testing error is random. Wouldn't you expect a student whose cat hadn't died the day before the test to score higher than a student of equal achievement whose cat had?
Most testing error is not anticipated. On one administration of the multiple-choice National Latin Exam, Carl--a solid "D-" student--scored "cum laude." How? That year, the scoring key was imbalanced, with C almost always the correct answer (Carl guessed C for his name whenever he was in doubt, which was most of the time).

In the context of the NCLB, when a school notifies its parents that it hasn't made AYP, it will not list error as one of the reasons why its percentages proficient didn't meet the state's AMO. Parents though should consider that possibility. Scores on the state standardized achievement tests are the NCLB's main academic indicator: AYP decisions are "high stakes" decisions for schools. The positivist assumption that test scores absolutely reflect school achievement is not tenable.

³ See Education Commission of the States' reviews of the state accountability workbooks and USED decisions ( http://www.ccsso.org/content/pdfs/StatewideEducationalAccountabilityUnderNCLB.pdf; http://www.ccsso.org/content/pdfs/RevistingStatewideEducationalAccountabilityUnderNCLB.pdf); see also USED's NCLB Policy Letter Updates to States ( http://www.ed.gov/policy/elsec/guid/stateletters/index.html).

⁴ Parents deciding they want to explore the alternatives can use the names of transfer option schools and "supplemental education service" providers as new search phrases (links to them may also be provided on the state page listing schools on the "needs improvement" list).

One parent has a student who needs more practice with long division. S/he reads the web storefronts of the available "supplemental educational service" providers: none mention math tutoring. The parent though does find the email addresses of the providers--s/he can ask them about their experience providing supplemental math tutoring (or perhaps their company policy about student records).

Chapter V: Unintended Consequences

One of the NCLB's short-term effects is more testing and annual AYP decisions. Another is reporting state standardized achievement test scores by demographic groups within schools. Doing so doesn't say anything about the learning of any child but does provide a picture of achievement scores by different groups at a given point in time, the immediate past pre

sent.

Most states seem spinelessly locked into the attempting to reason and denial phases of a codependent abusive relationship with the U.S. Department of Education. Even Utah's House Bill 1001 called repeatedly for local school officials to lobby federal officials about the NCLB. 1 Students are already being left behind:

The USED's senseless sanctimonious top-down rigid rules about how many tests can be administered with accommodations has already resulted in some students being confronted with grade-level tests whose very questions make no sense to them (surely a humiliating experience for the sincere).
There are also some students somewhere who have already been stigmatized by belonging to a group that didn't make AYP (and so let their school down).
Schools with native language immersion programs have been at odds with what the NCLB tests--English reading proficiency--since the beginning (the impact will be disproportionately felt by languages with the fewest speakers).

What students could have learned during the time they were being tested to produce accountability numbers will never be known: such questions are never asked. One thing all students will come away with is a more finely honed sense of the importance of unquestioningly following directions or orders.

The NCLB's long-term consequences will probably first be felt by students now starting public K-12. The amount and cost of instructional time devoted to the different phases of the testing cycle--field-testing items (think of it as a test's beta phase), item review, student drills, meetings, actual test administration, writing self-improvement plans--increases under NCLB. When the state achievement tests--the NCLB's major academic indicators--aren't sufficiently reliable and valid, scores on them are meaningless and the whole accountability endeavor a waste of time. The states, by and large, have not held their test publishers to very high standards (largely because of the untenable positivist position that the state achievement tests measure what the state legislators want them to measure). Although it is rigid about data formatting, USED has little to say about the state achievement tests other than that they should be "scientifically valid and reliable:" USED creating its new Institutes of Education Sciences does not make education a science like physics anymore than assuming that the state achievement tests are reliable makes them so.

Even in the best of all worlds--the NCLB's mandates are adequately funded; the state achievement tests are reasonably reliable and valid; the test publishers make no scoring errors; the data are rationally analyzed--one still has to ask whether the accountability numbers are worth the resources consumed--money as well as teacher and student time--creating them: the issue is priorities, resource consumption during lean times.

If history is a predictor, as with every other major attempt at reforming U.S. K-12, the NCLB's unintended long-term consequences will be greater than its intended short-term effects.

It is inevitable the curriculum will homogenize around what's being tested: for the language arts, that means emphasis on reading over writing skills. This homogenization of the curriculum occurs in the context of an increasingly homogenized U.S. culture (local weather programs broadcast from national satellites, media consolidation, power entrenchment).

Teaching to the test is the main mechanism for curricular homogenization. It is on the surface an idea with a lot going for it: what could be more reasonable than teaching and testing on the content that you expect to be learned? A state wants students to multiply by grade x; teachers in grade x-1 teach students to multiply; students are tested on multiplication problems when they reach grade x. So what's there to oppose? The problem is that judging schools and teachers solely on the basis of tested subjects results in tested subjects being emphasized over untested ones (in the free market idiom, if my only market is for beans, I'm going to quit planting corn). Note also that, were student scores aggregated over teachers in this example, the teachers in grade x would be assigned scores based on how well the teachers in grade x-1 taught multiplication: that's irrational (even assuming the tests themselves were sufficiently reliable and valid).

Anonymity of public K-12 student data will be compromised. Although NCLB itself contains an explicit prohibition against creating a single nationwide database, one seems inevitable (or perhaps already exists: USED speaks of "complete data harvesting" from the states). Such a database could be created deliberately or through a slovenly series of choicepoint-like accidents. Once created, copies will be made, and the clock cannot be turned back: The repercussions of that identity theft will follow those children for the rest of their lives. At the least their test scores will form the basis for targeted advertising campaigns--more Orwellian uses are also too easily imaginable.

The same technology that makes mass accountability testing and your zeroing in on decision-related documents enables advertisers to target particular groups: don't want your newborn to flunk your state exit exam like you did? sign up today.

Even legitimately obtained USED data are imperiled, because USED servers run on a Microsoft variant: That means equal insecurity on state servers (USED is so rigid * about data formatting that states are forced to use MS products: you may remember MS from its cameo roles in the past two U.S. presidential elections and last century's Department of Justice anticompetitive practices suit).

* USED is as imperious about the format of student data records--the third field will contain your student's telephone number and the fourth, address--don't use hyphens when entering student data; abbreviate "Avenue" AVE---as it is about the percentage of students who can can take the state achievement tests with accommodations.

The achievement gap will not close, but no one will officially take notice. It may even widen, aided by a growing digital divide. Attention will remain on the annually aggregated scores from the tests that may or may not be sufficiently reliable and valid to provide a basis for any inference.

These numbers will be manipulated in various ways whose sole goal will be to keep up the appearance of increasing proficiency nationwide.
Enron-like scandals about NCLB accountability data being fudged will continue. Already people speak of propaganda being packaged as news: how much easier to manipulate data behind the scenes without anyone noticing (the more innumerate the populace, the easier will be the scam).

The original "Houston education miracle" was created by misclassifying--deliberately, not negligently--students who dropped out as students who went on to get their GEDs.
This kind of math has been going on for years: A good example is the school whose score sheets had considerably more erasures than others': forensic investigation revealed that 90% of those erasures came from changing incorrect to correct answers--and that "follow-up investigation of older tests ... showed tests had been tampered with for years" (Bushweller, American School Board Journal, 1997).

So much dust will be raised when the first schools fail to make "adequate yearly progress" (AYP) for five consecutive years--that we can predict this will happen does not make us seers--that at least temporary chaos will ensue, and all parents who can will think seriously about abandoning public K-12 because of their totally legitimate pressing concerns about their own student. To those who say eliminate the achievement gap by providing parents with meaningful economic opportunities--if the past few years are a legitimate guide, this is unlikely to happen--there is no answer.

Parents as information hunter-gatherers

Parents have been more and less part of the process in which the states defined their performance objectives and achievement levels. They have set on the committees that did the work: some have shared their experiences (http://www.parentsinc.org/newsletter/news0700/isetscores.html). NCLB requires that parents be informed when their local schools don't make "adequate yearly progress" (AYP) and be advised of their alternatives. Parents are also free to seek more information than that provided them--they've been looking online for supplemental information long before NCLB.

The same technology that makes mass achievement accountability testing possible--NCLB more than doubles the number of tests administered and the amount of time teachers and students spend testing and being tested--makes it easy for parents to sift through mountains of information to zero in on the handful of documents that pertain to how their state is implementing NCLB and thus its direct impact on their student.

Today's parents can easily become successful online information gleaners, because--in the context of full-text indexing of public documents--all that's needed to get a tailored search engine results page is the right jargon. In addition to self-enabling themselves to make more informed personal decisions about their student, parents can also publicly act on what they have learned--writing congressional delegations and signing petitions, online and off. Parents can also ask questions: since one of NCLB's stated goals is parental choice, most state and federal public NCLB accountability documents contain at least the email address of the page's webmaster. Parents could also self organize:

Anyone can start a new discussion thread or group and others can respond: Many parents might have the same question about a particular "supplemental educational service" provider; although each state independently approves its own list of such providers, most are not regional businesses.
It is possible to imagine ad hoc anonymous intra- and inter-state online communities of parents forming like flash mobs around particular issues--the ones of immediate relevance to them--and then dissolving after solving their presenting problem: USED Secretary Spellings' April 20th letter to the Utah State legislature is an example. The letter passed, from the Secretary to the senior federal Senator and then to a state Senator and then to other State legislature members--and, within short order, someone having scanned the original or a copy, theoretically anyone in Utah who was interested had a copy: horizontal communication like this occurs without preplanning and special collaborative software.
Parents can also, of course, read what local and national journalists are saying by going to any of the automatically generated online news sites. The impact of parents having such unprecedented access to decision-specific information is unknown: certain though is that informed decisions are often better ones. What to do though when the choice is between a rock and a hard place?

Footnotes

¹ Utah's modest proposal

On April 19, 2005, the Utah State Legislature passed its House Bill 1001 ("Implementing Federal Educational Programs"). This bill instructed school officials to prioritize Utah educational standards and procedures over those of the federal "No Child Left Behind" (NCLB) law and to only participate in federal mandates that were adequately funded (Utah has its own standardized achievement tests for grades 5, 8, and 11). Utah has been proactive about protecting student records (http://attygen.state.ut.us/PrRel/prjan132005studentinfo.htm) so it is not surprising that HB 1001 instructed school officials to "protect the confidentiality of data under state and federal privacy laws" (http://www.le.state.ut.us/~2005S1/bills/hbillint/hb1001.htm). The bill also instructed school officials to, as part of their "normal duties,"

"Lobby Congress for needed changes to the No Child Left Behind Act ... lobby federal education officials for relief from the provisions of the No Child Left Behind Act, including waivers from federal requirements, regulations, and administrative burdens ... lobby Congress and federal education officials for needed resolution and clarification."
"In the case of conflicts between the No Child Left Behind Act and the Individuals with Disabilities Education Act, the parents, in conjunction with school officials, shall determine which program best meets the educational needs of the student." The Utah bill further instructs school officials that:
"Each local school board shall: (b) administer tests, required by the State Board of Education, which measure the progress of each student, and coordinate with the state superintendent and State Board of Education to assess results and create plans to improve the student's progress which shall be submitted to the State Office of Education for approval; (c) use progress-based assessments * as part of a plan to identify schools, teachers, and students that need remediation and determine the type and amount of federal, state, and local resources to implement remediation; (d) develop early warning systems for students or classes failing to make progress." Utah's bill is pertinent to NCLB only insofar as it places state achievement standards over federal ones and eschews unfunded federal mandates. The bill also instructs school officials to protect student records, lobby federal officials, and administer tests to measure student learning (and identify any in need of remediation). The Utah State Legislature asked the U.S. Department of Education (USED) to review its HB 1001.

* Utah's longitudinal assessment method focuses on rates of student learning while the NCLB focuses on levels of student achievement. Utah's method makes sense statistically and allows Utah to examine and report rates of learning across the same demographic subgroups as NCLB requires (assuming that Utah's measures of learning, its UPASS tests, are technically adequate). If there are enough students in a subgroup to report its level of learning--NCLB uses percentage proficient--then there are enough to report its rate of learning: no public accountability is lost.

On April 20th, USED Secretary Spellings wrote * that Utah's bill was "needlessly confrontational" and that:

"The consequences of enacting and implementing this bill would be so detrimental to students in Utah"--at least $76 million in federal funds are at stake--particularly low-income and minority students (http://www.parentsunitednetwork.org/24Apr20052.html). The Secretary's letter concluded that:
"Under the state test, 74 percent of white eighth-graders and 47 percent of Hispanics were reported 'proficient or advanced' in math, while just 34 percent of white and 7 percent of Hispanic eighth-graders were 'proficient or advanced' on the NAEP math test." Since the state definitions of "proficient" are different from each other--and from the federal definition--it would be surprising if the two numbers to which the Secretary referred were the same.
Secretary Spellings also wrote in a press release that same day that: "States across the nation who have embraced No Child Left Behind have shown progress: student achievement is rising and the achievement gap is closing. The same could be true in Utah, whose achievement gap between Hispanics and their peers is the third largest in the nation and has not improved significantly in over a decade."
(http://www.ed.gov/news/pressreleases/2005/04/04202005a.html). The Secretary did not specify the data on which her assertions were based.

Although Utah's bill consisted of a list of instructions to school officials--Utah officials found the wording of the Secretary's letter "quite extraordinary"--Connecticut and the National Education Association have sued USED in the courts: CT argued that developing, administering, and scoring six more tests in reading and math--CT already has tests for grades 4, 6, and 8 (and would need tests for grades 3, 5, and 7)--would cost 8 million over current funding levels (about 1.3 million per test). The NEA's suit also noted that those students most harmed by NCLB are those it was intended to help. While other states and organizations are joining this fray, some, like Alaska, are ahead of schedule, having already developed and administered new grade level tests (despite record windfall oil profits, one of its school districts is laying off teachers, another is ready to go to court over the still unresolved funding formula, and the state education association has issued a strong vote of no confidence).

* USED delivered the Secretary's response--a formal letter--to the Utah State Legislature's question to Utah's senior federal senator. USED's web page hasn't been updated to include this letter yet, but there are numerous copies on fax machines and hard disk drives throughout Utah. Although no mountains have yet been named after Senator Dayton (who devoted the last two years to the several earlier versions of HB 1001), a recent editorial suggests Utah feels embattled: "Two-hundred-thirty years ago, Massachusetts militiamen at Lexington and Concord faced off with British troops, firing the famous shot heard 'round the world .... At Lexington and Concord, and in Salt Lake City, the central players were taking a risk. The colonists were risking the wrath of a global superpower--Britain--and the hangman's noose for treason .... " (Provo Daily-Herald, 21 April 2005).

Appendix: Glossary of Search Strings

to do: incorporate search jargon definitions

Questions about the NCLB's direct impacts on your student are easy to answer, because the NCLB makes documents about them publicly available to parents. Most states began making information about their schools public in the 1990s. Parents can go to their state web sites and find these public documents: parents can, however, find them more quickly using a commercial search engine and then the find function on their browsers. Internet searches that produce thousands of partially relevant results are not particularly helpful (they are also unnecessary, because, by limiting your search to your own state, with the judicious use of jargon, you can zero your search down to a handful of results that explicitly answer your questions about your state's implementation of the federal NCLB and its impact on your K-12 student).

Since one NCLB impact is more testing, you might want to learn how many standardized tests your state now administers each year: specify that you only want results from your state and then search on "testing calendar" or "assessment calendar".

Specify that you only want public documents from your state by indicating that WE.us is the domain to be searched (go to any search engine's advanced interface). Replace the WE with your state's 2-character abbreviation (from AK to WY). Another sometimes useful domain to search for a state's public NCLB documents is WE.gov.

Follow a link to learn which test is administered when (sometimes it's easier to get the current year's calendar by using the current year as a search term and sometimes specifying that you only want pages updated in the last three months). The name of your state's achievement test is also a useful search phrase--as is "accountability workbook", the title of one of the digital documents USED required that the states submit for its approval (about state accountability plans). Most states have revised their workbooks several times.

State tests are based on individual state performance objectives.
You might want to search for both math "performance objective" and math "test blueprint" (or maybe your state uses "blue print"). You could be more specific: third math "test blueprint". Follow any of a handful of links to read your state's K3 objectives in detail or to see them cross-tabulated by the number and type of test items (multiple-choice or not).
If you're looking for a particular objective, use the find function in your browser; if you want to use the document as a study guide, print it; if you have questions, email your state's webmaster (these documents are online to help parents--so questions are welcomed): NCLB calls for public accountability and parental choice. What the search engines make possible is easy, targeted information gathering.
The technical reports the test publishers submitted to the states are the main source of information about the caliber of the state achievement tests (the NCLB's main academic indicator): paste in queries like education test reliability validity to locate documents about the technical caliber of your state's achievement tests. If you know your state's test name, search on it (if you don't limit your search to your state's documents, your search will show you whether anyone else has investigated its caliber).

NCLB instructs the states to use standardized achievement tests that are"scientifically valid and reliable" and base instructional decisions on "scientifically based research": these two are useful phrases for locating federal NCLB documents.

If your state's test has free-response items, look for high rates of "inter-rater agreement"--the percentage of the test publisher's employees who scored individual student responses the same--also called "interrater reliability". If you find no documents mentioning reliability and validity, your state probably hasn't posted its test publisher's technical report online. To locate documents describing how and where your state has defined the different achievement levels it reports to you on your student's report card, use education test "cut score" proficient or "minimum competence" (some states use "proficiency" and "competency").
The "achievement gap" and a"climate of low expectations" are not state specific phenomena. NCLB requires that schools report their annual aggregated student achievement test scores by different demographic groups. NCLB also requires that states test at least 95% of their students (and that no more than 1%--3% for some states--of the tests are administered with accommodations (e.g., a math test administered in the language a student knew). If you're interested in what alternative assessments your state offers under NCLB, you could do a basic search of your state's public documents on education test accommodation or education test "alternative assessment". To ask a more specific question, append some of the following to your basic search:

For questions about language, add"English language proficiency" ELF, "limited English proficiency"--sometimes "proficient" is the better term-- LEP,"English as a second language", ESL, or bilingual.
For questions about students with disabilities, add"students with disabilities", IDEA, or"disability status".
For questions are about economic status, add poverty,"economically disadvantaged","free lunch" eligibility,"low-income", or"Individuals with Disabilities Education Act" (IDEA).
Other NCLB reporting categories are the Census 2000 ethnicity categories, migrant, homeless, and gender.

To find out your state's "annual measurable objective", search on it or words and phrases like

education test "annual measurable objective" (your state may use the acronym AMO or the phrases "measurable objective" and "annual objective").
Schools whose "percentage proficient" on the state's achievement test are significantly lower than the state's AMO that year are classified as not making "adequate yearly progress" (or AYP or "yearly progress" and "annual yearly progress").

Schools inform parents if their schools are not making AYP. They also post this information online. To find schools that have not made AYP for at least two consecutive years, try a search like school NCLB AYP "needs improvement" (alternatives to the quoted phrase are "identified for improvement", "school improvement","program improvement","corrective action", and"alternative governance").
If you find that your school offers "supplemental educational services" and transfer options, things for you to decide include whether your school's not having made AYP means anything and what are your student's alternatives: Use the names of the schools under the transfer option and the names of the supplemental service providers as new search phrases: even if you have a question that you can't answer by looking online, you'll at least find a monitored telephone number or email address.

Parents will find gathering information about the direct impact of the NCLB on their student relatively easy. Questions that can be answered this way are, of course, not the only kinds worth asking: Many, most important questions can't be answered with searches like the above: is your student better off remaining in your local school or transferring? what supplemental services are useful? These questions you must answer: the preceding can only help with the information gathering phase of your decision making. Other questions are too broad to be answered by simple online searches: How will the NCLB change the schools other than, in the near future, more than doubling testing burden? Will reporting by demographic categories eliminate the U.S.'s achievement gap? Are all the state achievement tests sufficiently reliable and valid? What are the risks to student data privacy? Do annual decisions make sense (especially since aggregated numbers do not measure learning)?

Footnotes

¹ Parents can also search the NCLB itself to better understand how their states are implementing it. Anyone looking at the link to the NCLB's table of contents--maintained by the U.S. Department of Education (USED)--will notice that its links are named "Sec." number this and Sec. number that (and, looking at the file names, "beginning.html" is followed by "pg1.html" through "pg125.html"). Specify ed.gov as the domain to be searched. Use SEC. "elementary and secondary education" as your base search. Add other phrases and terms--from "test publisher" to 2013--to your base search. Between the search engines and the find function on your browser, you can easily locate those parts of the law relevant to your question and then read a paragraph or two to get the gist.

² One clear direct impact of the NCLB on individual students--especially those just beginning public K-12--is more time spent being tested. Individual students are also certain to be at least indirectly affected by being in a school that "needs improvement." Except in the sense that time spent being tested could be time spent learning something else, the NCLB has not changed the curriculum (beyond the same general homogenization around what's being tested, as happened in each of the states after they implemented their accountability testing programs in the 1990s). USED trying to standardize the state definitions of proficiency using the National Assessment of Educational Progress (NAEP) test results would herald an entirely new level of NCLB implementation, one with a direct federal voice in curricular decisions.

Bibliography

References cited in previous chapters:
Preface: http://attygen.state.ut.us/PrRel/prjan132005studentinfo.htm; http://www.le.state.ut.us/~2005S1/bills/hbillint/hb1001.htm; http://www.parentsunitednetwork.org/24Apr20052.html; http://www.ed.gov/news/pressreleases/2005/04/04202005a.html.

Chapter I:
Public Law 107-110: No Child Left Behind (NCLB). http://www.ed.gov/policy/elsec/leg/esea02/index.html; http://www.ed.gov/policy/elsec/leg/esea02/pg8.html#sec1308; http://www.ed.gov/policy/elsec/leg/esea02/pg112.html#sec9531

Chapter II:
a) Boston College. Trends in International Mathematics and Science Study (TIMMS), 1995, 1999, 2003 (http://isc.bc.edu/); b) Family Educational Rights and Privacy Act (FERPA): Summary of rules regarding student record privacy. Each state also has its own student record confidentiality procedures (http://www.ed.gov/offices/OII/fpco); c) Jesness, J. Stand and Deliver Revisited: The untold story behind the famous rise -- and shameful fall--of Jaime Escalante, America's master math teacher. Reasononline, July 2002 (http://reason.com/0207/fe.jj.stand.shtml); d) Hayes, W. L. Statistics. Fort Worth: Harcourt Brace, 5 Ed. (1994), Winer, B.J. Statistical Principles in Experimental Design. NY: McGraw-Hill, 2nd Ed. (1971)

Chapter III:
USED. Table, Percentage of Students at or above Proficient in Mathematics, Grade 8 Public Schools: By State, 1990-2003. (http://nces.ed.gov/nationsreportcard/mathematics/results2003/stateachieve-g8-compare.asp).

Chapter IV:
a) Education Commission of the State' reviews of accountability workbooks and USED decisions ( http://www.ccsso.org/content/pdfs/StatewideEducationalAccountabilityUnderNCLB.pdf; http://www.ccsso.org/content/pdfs/RevistingStatewideEducationalAccountabilityUnderNCLB.pdf) and b) USED's NCLB Policy Letter Updates to States ( http://www.ed.gov/policy/elsec/guid/stateletters/index.html).

Chapter V:
Bushweller, K. Teaching to the Test, American School Board Journal (Sept 1997); Weiss, T. I Helped Set the Exit Exam Scores. Parents News (Mar 00) (http://www.parentsinc.org/newsletter/news0700/isetscores.html)

Appendix:
Joint Committee on Testing Practices (American Educational Research Association, American Psychological Association, National Council on Measurement in Education, Standards for Educational and Psychological Testing, Washington, DC (1999)