The Flawed Requirements for Limited English Proficient Children of the No Child Left Behind Act

The Flawed Requirements for Limited English Proficient Children of the No Child Left Behind Act A lthough much has been written generally about the achievment and choice requirements of No Child Left Behind (NCLB), little has been written about the provisions of the bill that pertain to Limited English Proficient children. The purpose of this paper is to do that.1 Both Title I and Title III of NCLB speak to the education of limited English proficient (LEP) children, also known as English language learners (ELLs).2 The most important, and perhaps the most controversial, part of NCLB is the requirement in Title I that all students, including the economically disadvantaged, those from major racial and ethnic groups, those with disabilities, and those with limited English proficiency, must achieve the state’s proficient level on “challenging” academic standards by 2013/14.3 However, the test score goals for LEP children are not merely unrealistic—the most common criticism of NCLB—they are illogical. Unlike the black–white achievement gap and the poor–affluent achievement gap, which can at least theoretically be eliminated, the achievement gap between the category of children designated limited English proficient and the category of children designated fluent English proficient (FEP) can never be eliminated. That is because an LEP child is not just a child from an immigrant or non-English-speaking family. An LEP child is a child from an immigrant or non-Englishspeaking family who scores low in English. If you define a group by their low test scores, that group must have low test scores or someone has made a mistake. Thus, the original NCLB legislation, and virtually all of the analyses of NCLB, incorrectly treat limited English proficiency as if it were the same kind of achievement disadvantage as race or ethnicity or poverty—that is, something that can be corrected with enough money or expertise or willpower. It is not. Black and Hispanic students remain black or Hispanic no matter how high their scores get. Poor children similarly remain poor no matter how high their scores get. LEP children, however, are removed from the category called LEP once their scores reach a certain level, leaving only low scorers in the LEP group. In other words, if an LEP student meets the state’s standards for “proficient” on the state achievement test that all children have to take, they are not LEP—that is “limited in English.” In a nutshell, the essential flaw of NCLB is that it does not recognize that LEP children as a group can never be at the state’s proficient level if that level is high and challenging, even if every child formerly in that group becomes a high-achieving English speaker. Annual measurable objectives Title I requires that the plan to achieve 100 percent of limited English proficient students at the state’s “proficient” level by 2013/14 include, beginning with 2001/02, annual objectives expressed as percentages of students achieving at the proficient level. The starting percentage is the higher [emphasis added] of either the percentage of students at the proficient level who are in the state’s lowest achieving subgroup of students (economically disadvantaged, major racial and ethnic groups, students with disabilities, and students with limited English proficiency) or the school at the 20th percentile in the state, based on enrollment, among all schools ranked by the percentage of students at the proficient level. I do not understand the logic of this formula nor have I found anyone who does nor, amazingly, anyone who questions it. All of the states use the starting point of the school at the 20th percentile since it is always higher than the lowest achieving subgroup. Illinois’ accountability plan4 illustrates the problem with this method using their 2002 achievement data. As shown below, the percent proficient of the school representing the cumulative 20th percentile in enrollment (when rank ordered on percent proficient), shown in the first column below, is substantially higher than that of any subgroup except Asians and whites. This means that LEP students, low-income students, students with disabilities, blacks, and Hispanics have a starting point that is substantially higher than their actual starting point. Illinois 2002 20% method All American Indian Asian Black Reading 40.86 59.3 59.2 68.5 36.8 Math 39.68 60.0 54.9 79.2 32.2 Hispanic White IEP Low Income LEP Lowest Group 37.1 72.2 27.4 38.4 24.1 24.1 41.0 72.7 30.3 39.2 31.9 30.3 Source: Accountability Workbook, Illinois Board of Education, Revised May 2004, p. 7. If a Local Education Authority (LEA) has failed to meet their achievement objectives for two consecutive years, the state agency must require the LEA to develop an improvement plan. Schools with large numbers of the types of students whose statewide starting point was substantially higher than their actual starting point could be unfairly targeted as needing improvement because they did not meet the statewide annual target even though they made substantial improvement. This can occur because the statewide starting point was too far above their actual starting point. An affluent high-scoring school, on the other hand, could actually have a decline in achievement and still meet their Title I Annual Measurement Achievement Objective for that year because the statewide starting point was well below their actual starting point. • LEP children are also affected by Title III, which is devoted solely to limited English proficient and immigrant students. Title III requires that each state: • establish English language proficiency standards aligned with the state curriculum and standards; • conduct an annual assessment of English language proficiency using a test designed for children learning English; • define two annual measurable achievement objectives for LEP students’ English proficiency: – annual increases in the percentage of students making progress, and – annual increases in the percentage of students attaining English proficiency; • hold LEAs accountable for meeting the above Annual Measurement Achievement Objectves; and • monitor formerly LEP students for two years after being redesignated FEP. Thus, not only do states have to achieve Title I goals for LEP students, but LEAs that receive Title III funds also have to meet Title III goals. In short, they have a double set of hurdles to overcome. No Child Left Behind specifically states that Title III neither mandates nor precludes the use of a particular curricular or pedagogical approach to educating LEP children. However, students can be tested in their native language on the state reading and language arts achievement tests for up to five years and perhaps longer,5 although they still must take an English language proficiency test each year and show progress on it, and they must take the state math assessment in English. NCLB is unrealistic, however, with regard to math achievement. LEP students may be excused from taking math tests for their first year in the United States, but it is assumed that after one year there is no language barrier. This is simply not true. Since most K–12 math is language-based, it is not possible for a child who is truly LEP to understand a language-based lesson in math at the same level as a child who is fluent in English. What is a limited English proficient child? No Child Left Behind treats this issue as most educators do, as if it were a simple decision. It is not. The determination of the English proficiency of a child is conducted by the LEA, but regulated by the state. The process typically includes two or three steps, of which the first is the home language survey (HLS). That survey, filled out by the parents of all entering students, asks several questions about the home language, including (1) the language that the student first acquired or the language most often spoken by the student, and (2) the primary language used in the home. If a parent indicates a language other than English on any of the questions, the child is sent to have his/her English proficiency assessed by an oral test if the child is young and assumed to be preliterate, or an oral and written test if the child is older and assumed to be literate. Before NCLB, children who came from a language minority background as determined by the home language survey had their English proficiency assessed by any one of a dozen or so commercial English proficiency tests, such as the BINL, BSM I/II, Pre-IPT, IPT I/II, pre-LAS, LAS I/II, and the Woodcock-Muñoz Language Survey. These tests are normed on a mix of English-speaking students and high-level (i.e., high-scoring) LEP students6 and a cut-score is chosen that assigns children to a specific LEP level based on their score or FEP if they score above it. Since NCLB, the trend has been to replace the dozens of tests used in a state with a single, statewide English proficiency test. In a few instances, states have entered into multi-state consortia to develop a test used by the members of the consortia. Developing a single statewide English proficiency test is a step in the right direction because it improves monitoring by increasing coordination and continuity. It does not, however, solve the essential problems of the tests themselves. First, any cut-score above zero will identify some percentage of children as limited English proficient who are in fact fluent in English (Rossell and Baker, 1988, 1996; Rossell, 2000a, 2000b, 2002). I call these “false positives”—that is, children are designated as limited English proficient when in fact they are fluent in English, but do not know the answer to many of the questions. In one experiment conducted in Chicago two decades ago, the LAS test identified 78 percent of above-average English monolingual kindergarten students as limited English proficient, and across all grades 45 percent were identified as limited English proficient (Pearlman and Rice, 1979). Luckily, the home language survey is used as an initial screening device or three-fourths of American kindergarten students would have been classified as LEP by these tests. The inclusion of high-scoring LEP students in the norming population in recent years has probably reduced the false positive rate, but I could not find any information on what the current rate is nor any interest in the topic among those I called, which included one test-publisher representative. In light of the low redesignation rates in the seven states (Arizona, California, Florida, Illinois, Massachusetts, New York, and Texas) that I have examined in detail, I suspect it is still pretty high. Second, because each test publisher selects different cut-scores for categorizing a child as limited English proficient, the tests do not agree with each other. A child could be identified as LEP by one test and FEP by another (Rossell and Baker, 1988; Rossell, 2000a, 2000b). More to the point, a child could be identified as FEP in September in Kansas and LEP the next June in California because the latter is using a different test. The tests are not only invalid to some unknown degree (that is, they misidentify children as LEP), they are also unreliable (that is, you cannot get the same answer in repeated measures). In several analyses done of language proficiency tests, a large percentage of scores for the same child and test were either no different after one year or lower. Since it is highly unlikely that an LEP child will know less English at the end of one year than at the beginning, one can only conclude that the tests are unreliable (Rossell and Baker, 1988). Redesignation The process of redesignating a child as fluent English proficient has the same problems. It typically involves an English proficiency test and often an academic achievement test. The lack of agreement between English proficiency tests is even greater at the point of redesignation. This is because the problem of the tests being unable to tell the difference between a student who does not know the answer and a student who does not know enough English to understand the question is more likely at the point of redesignation, when the language proficiency exams are more difficult. Several years after the legislation was written and approved, the U.S. DOE recognized the fact that LEP children are defined by their low achievement and the proposed revised regulations allow a state to include “formerly LEP” students within the LEP category in making AYP determinations for up to two years after they have been redesignated fluent.7 While the inclusion of reclassified LEP student scores “for a year or two” is a step in the right direction, it is inadequate. Assuming an LEP student was correctly identified to begin with (which may not be true), he or she will not reach their true level of proficiency absent the language barrier in two years. Recommendations for change Title I AYP Current data, and common sense as well, indicate that it is impossible for all current LEP students as a group to meet any state’s standard for “proficient” if that standard is high and challenging. • Recommendation 1: Current and formerly LEP students should be exempted from meeting the state’s proficient level in any subject matter—math, language arts, science, social science and whatever other subjects are added in the future—until they have been educated in the United States for five full academic years. • Recommendation 2: LEP, and formerly LEP, students should show annual progress in all subject matter, but that progress should be assessed for individual students for whom there are two years of test scores in English and it should be exact scores, not broad LEP categories or levels. – Research should be conducted on samples of students to determine a definition of progress that is both ambitious and achievable. – In order to discourage failure to redesignate, LEP students who score at the proficient level on the state English language proficiency test for LEP students, but who are not redesignated, should not count toward the participation rate nor should their scores be included in the LEP category. – In order to discourage failure to redesignate, elementary (K–5) LEP students who have been educated in the United States for three years and have not been redesignated will not count toward the participation rate, and middle/high school students who have been educated in the United States for five years and have not been redesignated will not count toward the participation rate. Native Language Tests It is a glaring contradiction in NCLB that LEP students can take the state proficiency tests in their native language for up to five years. It is hypothetically possible that 100 percent of LEP students might make the state’s proficient level in every subject (except English language arts) because they took it in their native tongue. But what does that mean? First, tests are not equally difficult in different languages. Second, is a student really at the state’s proficient level if he or she cannot pass the test in the language of this country? • Recommendation 3: Native-language tests should be prohibited by the U.S. DOE. Alternative Tests These may be as misleading as native language tests. What does it mean for LEP students to reach their state’s proficient level if they have taken a different test? Indeed, they have not reached the proficient level and who knows what level they have reached? • Recommendation 4: Alternative tests should be prohibited by DOE. The Starting Point for AYP For LEP students, the state-imposed starting point is about 20 points higher than their actual starting point. Schools with large numbers of subgroups whose starting point was 20 points above their actual starting point could start encountering penalties despite having made as much, or greater, progress as another school that was lucky enough to be populated by a subgroup whose starting point matched, or was higher, than the single state starting point. This is obviously unfair. Following on this logic, the law’s requirement that all subgroups end at 100 percent proficiency in 2013/14 seems similarly arbitrary and capricious since they started at very different places. Schools explain only a quarter or less of the variation in academic achievement because students start school with very different amounts of social capital and they are in school only 13 percent of their waking hours from age 0–18. The current timetable and standards thus unfairly reward schools with mostly affluent students and unfairly punish schools with poor or limited English proficient students. • Recommendation 5: Each subgroup should have its own starting point for percentage proficient and realistic ending point. A onesize-fits-all plan is not realistic. Title III Standards English language proficiency tests will always over-identify children as LEP because tests cannot tell the difference between a child who does not know English and a child who does not know the answer. And educators are, unfortunately, sometimes too deferential to test results and to the notion that children should not be redesignated until they have fully mastered “academic English.” Therefore, there must be a mechanism to override the English proficiency test results for initial designation and redesignation. Although I would like to substitute the term “language minority family” for “limited English proficient” or “English Language Learner” in order to avoid invalid and unreliable English language proficiency tests and the process of redesignation (see Rossell, 2000a), that is probably too radical to ever be adopted. • Recommendation 6: If an English language proficiency test is used, it should be the same one for every school district in the state (and ideally the United States). • Recommendation 7: Any child who has been educated in the United States for three years who is designated LEP by an English language proficiency test either at entry/designation or exit/ redesignation should have the test overridden and be designated fluent English proficient. Annual Progress in Learning English The current situation of having the progress of cohorts of LEP students defined differently from state to state is untenable. • Recommendation 8: The definition of cohort should be individual students with test scores on the same English proficiency test at two points in time and it should be exact scores, not broad LEP categories or levels. Attaining English Proficiency The goal of attaining English proficiency is fraught with error even if all states were monitoring individual students. There is apparently no way to construct a test that can always tell the difference between a child who does not know English and a child who does not know the answer, and most educators, as well as test publishers, seem not to care. • Recommendation 9: The cohort of students that should be expected to attain English proficiency (different from the state’s proficient level) should be elementary (K–5) students who have been in the United States for three years and secondary (7–12) students who have been in the United States for five years. • States should administer their English proficiency test to English monolingual students to determine the extent of false positives (students fluent in English, classified as LEP) and adjust the percentage of the cohort that must attain English proficiency downward, at a minimum, by the false positive percentage. Right now the states are laboring under an arbitrary one-sizefits-all schedule and timetable that they were forced to adopt and are gamely trying to adhere to. One cannot help but admire their commitment, intelligence, and energy, but they are not going to succeed without lowering their standards unless NCLB is reformed so that it is both ambitious and realistic. Notes 1. A much longer, more detailed analysis was published by the Lexington Institute in September 2005. See Christine H. Rossell, Making Uneven Strides: State Standards for Achieving English Language Proficiency Under the No Child Left Behind Act. 2. Because NCLB refers to LEP children, I will use that term in this paper. However, the more common term in the last few years is English Learner or English Language Learner. 3. This can be found in Section 1111, page 1447 under (F) time line. 4. All of the state accountability plans can be found online at http://www .ed.gov/admins/lead/account/stateplans03/index.html. 5. See Title I, Part A, Section 1111(b)(3)(C)(ix)(III). This is also noted a little more clearly on p. 35463 in the Federal Register, vol. 69, No. 121, June 24, 2004, which can be found online at http://www.ed.gov/legislation /FedRegister/proprule/2004–2/062404a.html. 6. According to the technical manuals, the norming population for the LAS R/W is 32 percent English speaking, the LPTS is not stated, the MAC II is 0 percent English speaking, the IPT R/W is 48.4 percent, the IPT O is 30 percent. The rest are high scoring LEP students. 7. The Secretary’s letter to the states can be found online at http://www.ed .gov/news/pressreleases/2004/02/02192004.html. The revised rules can be found in the federal register online at http://www.ed.gov/legislation /FedRegister/proprule/2004–2/062404a.html.