The Truth About
the CELDT 2004
California English Language Development Test
The Legislative Analyst's Office Report
Jill
Kerper Mora
San Diego State University
The Magic Bullet is a Blank:
Updated April 2005
In February 2004, the Legislative Analyst’s Office (LAO) of California released a report on the data from the California English Language Development Test (CELDT) titled “A look at the progress of English Learner students.”
Subsequently, the LAO released an update in January 2006 on the CELDT data from 2002-04 addressing the issue of reclassification of English Learners as Fluent English Proficient titled "The progress of English Learner students."
The LAO is a nonpartisan office that provides fiscal and policy information to advise the Legislature. The purpose of the analysis in this report is to inform legislators in making decisions about programs and policies regarding English Language Learners (ELL) in the California public schools. Consequently, the LAO report merits a careful and thorough reading, as well as cautious interpretation because of the vital importance of formulating wise and effective policies for educating the 25% of California’s public school population classified as ELL. The report uses comparisons of rates of English acquisition among different ethnic groups that comprise California’s student population, including 83% native Spanish speakers. The LAO concludes that “…the Legislature should focus on learning about the mix of services that work for particular groups of EL students.” The report suggests that an “effective mix of services,” including outreach or other services to families or communities, may look very different from the current services provided by public schools for ELL. There is a refreshing acknowledgement that the data indicate that there need to be different programs or "mix of services" to meet different students' needs, a step away from the "one-size-fits-all" mentality. However, the overall premise of the report appears to be the traditional "How can we move them more quickly into English?" focus.
Please refer to my analysis of the CELDT test results for 2001 and 2002 and discussion of the media coverage of the scores by clicking here on The Truth About the CELDT. Also, this analysis of the SAT-9 reading scores for ELL from 1998-2001 is informative, since based on these test scores I accurately projected the academic learning curve that the LAO now documents. See this webpage for an analysis of the Redesignation Debate. I also recommend a review of this webpage for a relevant analysis of the political and policy issues behind Proposition 227.
The LAO Report: A Step in the Right Direction
First, it is a good thing to have a more enlightened discussion about how long it takes to learn English based on the state's own test data. The CELDT has opened the door for us to talk about English language acquisition and academic development with a data base that illuminates the complexity of the educational and policy issues surrounding schooling for language minority students. The interpretation of the larger body of CELDT data by experts in second-language (L2) acquisition and in models and programs for language minority students is necessary, especially to counter the many misinterpretations and abuses of data that occur. See, for instance, my analysis of the media response to the CELDT data from April 2003. We can also observe that the CELDT data confirm what bilingual and English as a second language educators having been saying all along: Different aspects of language and language-based academic processes (listening, speaking, reading, & writing) develop at different rates, ranging from 3-5 years for oral skills development to 5-7 years for academic language proficiency. Now that test data from the California Department of Education provide evidence of the veracity of this L2 acquisition timeframe, we can begin to truly and honestly examine the implications of this reality.
Currently, under Proposition 227 passed in 1998, California’s public school programs for ELL are based on a model that requires specialized English language instruction for a one-year period of “transition,” after which ELL are to be placed in mainstream or regular classrooms. Specifically, Proposition 227 says, “Children who are English learners shall be educated through sheltered English immersion during a temporary transition period not normally intended to exceed one year.” In other words, as a matter of law, the assumption is that language minority children will acquire a level of English proficiency sufficient for them to succeed in school after only one year. The theory behind the Proposition 227 instructional mandate is clearly exposed as somewhere between Mission Impossible and wishful thinking. The voters of California might as well have passed a law mandating that all children have a full set of permanent teeth by age 6 or that all babies walk by the age of 6 months. However, dental maturity and infant mobility are not “hot” political topics, as is the education of language minority students.
We must examine the wisdom and cogency of the LAO report on the CELDT data to determine whether the policy implications it draws and the recommendations it makes come closer to the reality of educating California’s large and growing language minority population. The LAO's focus on “a mix of services” for English language learners is a tacit recognition of an important demographic reality: At least 41% of the total school population in California is comprised of students who are either currently classified as limited English proficient or who have been reclassified as fluent English proficient. In other words, each student who is a member of this large portion of CA’s student population is at some point along the continuum of English language and literacy learning.
The 2004 CELDT Data: Misleading by "Percentage Points"
Obviously taking their cue from mortgage brokers, the California Department of Education released its now-annual Good News report about the results of the CELDT for 2004 by borrowing a term to describe increases in the percentages of ELL students scoring at the higher levels of proficiency: Percentage point increases. According to the CDE, the CELDT results tell us that "we're moving in the right direction" because the percentage of the total ELL population of students tested who scored at the advanced level increased from 11% to 15%. The media presented this as an increase of four percentage points in the early advanced and advanced categories of the CELDT, although there was no increase in percentage of students scoring at the early advanced level. In the eyes of educators and policymakers, one must wonder why a 4% increase in one category of scores seems merits so much crowing about "progress." Tthe possibility exists that this percentage increase in the advanced level scorers could be attributed to fewer students being reclassified from last year's cohort, which might, in and of itself, explain the increase in the percentage of "advanced" scorers over last year's EL population.
We must be very skeptical of the CELDT spin and hype. The so-called increases in the numbers of students scoring at the EA and A levels can be explained by demographic trends and shifts within the ELL student population that have nothing whatsoever to do with improvements in the education of English language learners. Words of caution came from the Legislative Analyst's Office senior analyst Paul Warren (Sturrock, San Francisco Chronicle, February 9, 2005) who pointed out that since the test results report on "shifting" groups of students each year, it is impossible to track the progress of individual students to determine if they are making gains.
In fact, a single factor that is unrelated to the quality of instruction or reclassification criteria can explain the CELDT statistics: students' age. According to the CDE fact sheets, 1,344,091 ELL took the CELDT in 2004-05. Among these, 609,992 or 45% of the total ELL population, were in grades 6-12. This is an important statistic because we can assume that students in the upper grades have had more years of schooling in English and are therefore much more likely to score in the upper ranges of the proficiency scale than students in the lower grades. Consider this in light of the fact that the average ELL takes six years to achieve fluent English proficient status. Some percentage of 6-12 students are new arrivals to the United States and to California who are likely to have lower levels of English. However, we can safely predict that the majority of 6-12 students have been in school in CA for a number of years and are therefore more likely to have gained English proficiency beyond the intermediate level. Therefore, it is not particularly startling or peculiar to find that a similar percentage of ELL test in this range on the CELDT. The four "percentage points" increase in students scoring at the Advanced level on the CELDT could therefore be explained by shifts in the age demographics of the limited English proficient student population. We don't know because the CDE's statistics are a doily of information about the demographic characteristics and academic achievement scores of ELL.
Meanwhile, the media continues to equate Early Advanced and Advanced scores on the CELDT with "fluency" and based on this misdefinition, to question the state's reclassification rate of 8.3% (in 2004, up .6 from 2003's rate of 7.7). In their rush to generate controversy where none exists, the focus of the CELDT reporting has shifted to speculation that students may be missing out on educational opportunities because they are not classified as Fluent English Proficient (FEP). If this is the case and there is a "backlog" of un-reclassified students who are being denied opportunities because of this, shame on the school districts! If these students are not ready for reclassification and would be denied necessary educational services and funding to help them close the achievement gap between them and their native-English speaking and FEP peers and classmates, then there is nothing to make a fuss about. The policymakers who determine reclassification criteria should not be pressured into changing these criteria to satisfy those with particular ideological or political agendas.
The important question to ask and answer is this: Are all ELL in grades K-12 receiving the most effective and appropriate instruction to take them to the next step in their English language fluency and in their academic achievement in literacy and the content areas? This is a much more challenging question than raising unfounded concerns about reclassification criteria. The greatest likelihood is that ELL who are scoring at EA and A levels are not yet reading and writing at a level equivalent to their native English speaking (NES) peers. Nor are they likely to have the grasp of the content areas that their NES peers have acquired as they move up through the grades at a "normal" pace. Contrary to what many reporters would have us believe, EA and A level scorers on the CELDT are NOT fluent in academic English. Therefore, they should not be deprived of the services and support systems in place for them by virtue of their ELL status.
The 2003 CELDT Data: Much To Do About Not Much
The California Department of Education announced the results of the 2003 test administration of CELDT in March 2004. Again this year, the CDE focused on the "good news" about the percentages of ELL scoring at the Early Advanced and Advanced levels on the test. A close examination of the changes of percentages of students at each level of proficiency indicated a minimum level of change from the last administration in 2002. See Attachment A for a table showing these statistics:
Changes in percentages of ELL students scoring at each proficiency level from 2002 to 2003:
Beginning level--minus 3%
Early Intermediate level--minus 5%
Intermediate level--minus 1%
Early Advanced level--plus 7%
Advanced level--plus 2%
Try as they might, the CDE cannot extract an endorsement of the English-only instructional policy established by passage of Proposition 227. The only legitimate basis for comparison of rates of English acquisition from "before and after" comparisons is the rate of reclassification of ELL as fluent English proficient. Since 18% of the total number of students who took the CELDT more than once during the three year period between 2001 and 2003 have reached the fluent English proficient (FEP) levels on the CELDT, we can infer that the exit rate averages 6%. The 2003 test data indicate an "exit rate" or "reclassification rate" of around 7%-9%. This reclassification rate is about the same as last year's (8%) and therefore, predictably stable. The CDE emphasized the larger and growing percentages of ELL who are now in the upper ranges of English proficiency, which is not particularly surprising, given the demographic patterns of the schools. The numbers of new immigrants to the United States is probably lower and the immigrant resident population is growing older and moving up in the grades, which results in larger numbers of proficient English and reclassified students. Students in the upper elementary grades will also predictably improve their scores on the reading and writing subtests of the CELDT. Contrary to what is touted in the press, these shifts do not necessarily mean that students are learning English any "quicker" than before. In fact, the Legislative Analyst's report debunks this popular theory.
Is “Quick” Enough?
The LAO report acknowledges that the challenge for California’s language minority population is two-pronged: learning English and mastering the content standards for each grade of their schooling. For years now, the focus of the debate about the best methods and approaches for educating language minority students has been on how quickly they can learn English. This is based on a belief that the “problem” facing these students is essentially a language problem. Consequently, the belief is that once the “language problem” is solved, these students will progress through the public school system and learn the academic content as if they were native speakers of English. This stance, which is both educational and political, is based on a number of false assumptions. Not the least of these is that we must change language minority students to fit the school system, rather than change the school system to fit the students. Underlying this assumption is lack of recognition that learning can take place in any language other than English. The LAO report appears to take this stance, expressing great concern about the length of time it takes students from different ethnic groups to reach English proficiency. Therefore, much of the policy discussion in this report is centered on how instruction and programs can possibly accelerate English learning.
There is nothing inherently wrong with the concept of “accelerating” English language learning through high-quality instruction. However, there is danger in creating the illusion that instruction and programs alone can hasten students’ acquisition of English. By focusing on the speed of acquisition of English language skills, the report sidesteps one of the central issues in language minority education: The educational equity of a system where students must acquire English language skills before they can have access to the academic content (the sequential model).
The LAO’s analysis bemoans how long it takes to learn English and the effects and disadvantages this "delay" has on students' ability to keep up academically, learn the California Content Standards (CCS), and engage in understanding "complex concepts" and "complex topics" required in the "...higher level portions of the state standards." However, the LAO report does not contain any discussion of an instructional model in which complex concepts and topics in the CCS could be understood and learned through instruction in students' native language--i.e., bilingual education. The same principle applies to reading and writing instruction. There is no mention of bilingual program students' Spanish Assessment of Basic Education (SABE) scores as an indication that a normal statistical distribution of students are acquiring on-grade level reading and writing skills in Spanish, which compensates for the "disadvantages" of not being able to perform at this same level in English.
All Things NOT Being Equal
The LAO Report mischaracterizes instruction in bilingual education and contradicts itself in presenting data regarding English language acquisition of students in bilingual programs. See the section under Other Findings--Instructional Services. In what may be merely a problem of overly generalized definitions and labels, the report states that students in bilingual programs do not receive English Language Development or "special instruction in academic subjects that is designed for students who are still learning to speak English (probably referring to SDAIE). This statement gives the erroneous impression that bilingual programs are not developing English and that probably all content instruction is in the native language (Spanish). This is not in keeping with the model of transitional bilingual education that is commonly accepted as effective in transitioning students into English.
The report then goes on to say this: "Second, bilingual students improve more slowly in learning English than other students. By fourth grade, our analysis showed bilingual students about a year behind students who received other services." In statistical terms, this means that a majority of students in biingual programs (12% of the total ELL population) are .75 English proficiency levels on the CELDT below their peers in English-only programs. It is not possible or correct to draw the inference from two test administrations of CELDT that CA fourth graders enrolled in a bilingual program are learning English at a slower rate in bilingual programs. Earlier, the report states that kindergarten students in bilingual programs are much less fluent than students have a lower level of proficiency than students "receiving other types of services." We must keep in mind that students with lower levels of English proficiency may be enrolled in a bilingual program BECAUSE they have less English proficiency. We cannot conclude that they have a lower level of English proficiency BECAUSE they are in a bilingual program. No cause-and-effect relationship can be inferred from this statistic.
The LAO report states that one-half of all kindergarten students in bilingual programs score at the lowest CELDT level as compared to one-fourth in all “other services.” Therefore, the CELDT data indicate that kindergarteners in bilingual programs have a different English proficiency profile than students in other types of programs from the onset of their schooling. The LAO Report does not take into account the fact that many students in bilingual programs who reach higher levels of English proficiency “exit” the program or transfer into English-only classrooms before they reach fourth grade. In fact, the report gives this caveat statement: “It is not possible to track student progress based on the types of services they receive over the years.”
In spite of its own caveat statement, the LAO then states the following: “…bilingual students improve more slowly in learning English than other students…. Since the bilingual model introduces English more slowly than the other services, this finding is not especially surprising. The data, unfortunately, do not allow an assessment of the consequences of this delay on the ultimate goal of achieving English proficiency.”
It is important to note that the LAO suggests that we assess the "consequences of delay" in students' learning English in a bilingual program, should in fact that delay exist, but does not suggest that we investigate the consequences of the inevitable delay in literacy and academic content learning that occurs in monolingual English instruction for students with no knowledge of the language. This is the other side of the proverbial coin: How much learning is lost when ELL cannot understand the language of instruction? Consequently, how much remediation will be required for them to "catch up" with what they have missed out on in the process of acquiring English?
The analysis of the CELDT data in the LAO's report is based only on statistical projections from two test administrations, Therefore, we must exercise great caution in interpreting the statements about the length of time students take to acquire full English proficiency in different programs. No valid inferences about the time it takes students enrolled in bilingual programs can be drawn without additional information, which are not available from only two administrations of the CELDT. The LAO captures only two points in time, draws a line between two test administrations, and then projects the slope of that line based on statistical modeling of the data. When examining the CELDT data from bilingual program fourth graders in 2003 who were bilingual program third graders in 2002, we see only one year of reported language growth. We cannot say that the students examined in this report in fourth grade have been enrolled in a bilingual program continuously for four or five years. How many of these students enrolled in the program for the first time, with a level one proficiency, in second grade, or third grade, or fourth, who are new immigrants from their country of origin? The report does, in fact, point out that students may be enrolled in different types of “services” at different points in their academic careers. We can reasonably assume that students who are enrolled in a bilingual program in fourth grade are in the program because they have a lower level of proficiency in English. Perhaps they have not exited the program for a variety of reasons or they are recent arrivals to the United States. Therefore, we must emphasize that the CELDT data do not make a statement about the effectiveness of bilingual programs or bilingual instruction.
Reinforcing the Model Minority Myth
The data from two cycles of administration of the CELDT project that there are differences among ethnic-linguistic groups in the number of years students require to be reclassified as proficient English speakers. The longest time span is for native speakers of Hmong (7.4 years) and the shortest for native Mandarin speakers (3.6). Native Spanish speakers, who comprise 83% of all ELL, appear to require 6.7 years to attain full proficiency (CELDT level 5).
We must use caution to avoid having the comparisons among the academic performance of different groups based on native language spoken reinforce the "model minority" myth. We must think more deeply about the socio-cultural and economic circumstances of different ethnic groups, such as the contact and interaction different groups have with native English speakers. An important element in the rate of acquisition of English among Spanish speakers is the pervasiveness of Spanish within our society, especially in southern CA where 25% of the ELL attend school, and the transnational nature of their lives that promotes higher levels of bilingualism and the maintenance of Spanish.
Although with some caution, the LAO concludes from these data that it is possible for ELL students to acquire English rapidly, given the conditions. The report accurately points out that many of the “right conditions” for acquiring English are beyond the control of the public schools. In the report titled "How Long Does It Take English Learners to Attain Proficiency" by Kenji Hakuta, Yuko Goto Butler, and Daria Witt (2000, January) found that the strongest variable associated with the rate of English acquisition is poverty. Poverty as a factor in the quality and effectiveness of the schooling and the home environments of different ethnic groups is itself complex and multifaceted. However, socioeconomic differences between the Mandarin-speaking community and the Spanish-speaking community cannot be dismissed as inconsequential in an analysis of English acquisition patterns and opportunities, both in and out of school.
The Double Message
Instead of entering into the turbulent and uncharted waters of ethnic group comparisons, the LAO focuses instead on the “policy implications” of the different communities’ support of their children’s rapid acquisition of English. The report states: “Some language groups may require schools to work more closely with parents and communities about the importance of quickly gaining proficiency in English and ways parents can support their children in that process.” This statement raises a number of problematic issues, given the current political and policy environment regarding education for language minority students. Proposition 227 unequivocally stated to the Latino community that if parents wished to foster Spanish language skills and literacy learning among their children, they must do so outside the context of the public schools. The complex and obstacle-ridden procedures for securing waivers for bilingual instruction included in the law severely limit access for parents who value bilingualism and biliteracy to obtain services. The Latino Spanish-speaking community accurately concludes that the heritage language and culture are not valued in the schools, so many parents reinforce the use of Spanish in the home to compensate.
Now, the LAO suggests that the schools should also “work more closely” with parents to stress the importance of English acquisition. This double message to the Latino community is both confusing and insulting since it implies that students’ development of linguistic and academic skills in Spanish is not supportive of their children’s academic achievement. It also states once again that bilingualism is not valued by the schools, and that the community should adjust its value system to fit the schools’ narrow views and prejudices, rather than the schools responding to the legitimate values and concerns of the bilingual Latino community.
There is a credible body of scholarly research to refute the premise that bilingualism has a negative impact on academic achievement and to support the benefits of bilingualism and acculturation on long term academic achievement, especially among second-generation immigrant students. For example, Rumbaut and Portes (2001) found in their research with 5,200 second-generation immigrant youth, that the most successful and most acculturated youth were those who maintained their bilingualism and whose families maintained a strong cultural identity and sense of community with the ancestral culture of origin. This was true of all immigrant youth and their families, not just Mexican-Americans. This research suggests that further study of acculturation patterns among different ethnic groups who show faster rates of English acquisition could prove to be productive. For instance, researchers might find that Mandarin and Korean native speakers have a higher level of availability of “de facto” bilingual education through community-organized native language schooling and support for bilingualism and biliteracy learning in the home, with accumulated benefits to in-school achievement. Such findings would suggest that rather than focus on converting the Spanish-speaking community over to a “quick” English learning model, the schools, and the larger society, would be further ahead to encourage and support family and school bilingualism and biliteracy.
A Match Under the Thermostat?
It is fascinating to contemplate the meaning of another of the CELDT statistics. According to the CDE report, 43% of the ELL tested in 2003 scored in the early advanced to advanced levels overall of the test. On the surface, this would mean that nearly half of all the students in California who are classified as English Language Learners are in fact not lacking in English proficiency. This is an amazing finding, if it is in fact the case, with some very profound implications for the California public schools. If this number of ELL are proficient in English, then how does the CDE explain their lack of academic achievement as measured by standardized tests? The answer to this question certainly puts a damper on the jubilance over the supposed gains in English proficiency. It also means that we must be giving instructional strategies and program structures and implementation at all grade levels very serious examination. There are also many of those of us with expertise in second-language acquisition, among us our esteemed colleague Stephen Krashen, who question whether or not the CELDT scores are inflated. Professor Krashen has suggested that this is much like lighting a match under the thermostat to claim that the temperature in the room is rising. Certainly, there has been enormous pressure placed on administrators and teachers to raise scores on the CELDT. It is also obvious that most students will score better on a test the second and/or third time they take it. We must also research the characteristics of the level of proficiency in academic English that the Early Advanced and Advanced level scores on the CELDT purport to define and measure.
Even if we take the 43% proficiency statistic at face value, however, the implications for the California public schools are considerable. We must keep in mind that the type of instruction and the focus of the curriculum that are appropriate and effective for ELL depends upon their level of English proficiency in each of the components of the language arts: Listening, speaking , reading and writing. ELL who are in the more advanced levels of English language development need an instructional focus on developing their reading and writing abilities and learning content, especially through reading in the content-areas. They cannot be expected, however, to perform in their literacy skills as if they were native English speakers. Nor can we assume that they have acquired the same level of content knowledge as their grade-level and age-matched peers.
The Language-Literacy Connection
The LAO report makes the following statement: “The second grade CELDT assesses students in reading and writing for the first time (students in K-1 are assessed only on speaking and listening). Thus, the low improvement rates at this grade suggest that reading and writing skills are slowing the progress of some EL students on the test.” The report then goes on to say, “Some groups of EL students show advanced reading and writing skills on the second grade CELDT. These students appear to become fluent and reclassify much more quickly than students who do not possess early reading and writing skills.” While noting the discrepancy between the rate of acquisition of listening and speaking skills in English and the more complex and abstract skills of reading and writing, the report says little else about the connection between literacy skills and the listening/speaking proficiency of these students. Thus, the LAO glosses over one of the more significant areas of challenge in improving instruction for ELL in the public schools.
The CELDT data regarding ELLs’ progress in reading and writing English as a second language come as no surprise to literacy educators and second-language reading researchers. Learning to read and write in a second language, especially without the benefit of well-established literacy skills in a student’s native language, is a complex and challenging endeavor. Take for example the excellent research article by Meinke Droop and Ludo Verhoeven of the University of Nijmegen, the Netherlands published in Reading Research Quarterly (Jan.-March, 2003) titled “Language proficiency and reading ability in first- and second-language learners.” Droop & Verhoeven (2003) examined five factors in the growth in reading of Dutch, Moroccan and Turkish third and fourth graders to compare the impact of the factors in first- and second-language (L2) reading proficiency. The five factors are decoding, morphosyntactic knowledge, vocabulary knowledge, oral text comprehension, and reading comprehension. Droop and Verhoeven concluded that for the L2 children, the correlations between reading comprehension, oral text comprehension, morphosyntactic knowledge, and vocabulary knowledge were all high, whereas reading comprehension and decoding skills correlated moderately. The data was analyzed using various combinations of these factors. However, the five-factor solution was the significantly better fit. The decoding factor did not stand out alone as an explanatory factor. Furthermore, these five factors explained only “about half of the variance” in reading scores.
An analysis of the CELDT data on the reading and writing subtests in light of the body of research on L2 literacy development raises a number of educational practice and policy issues. The LAO suggests that there are factors that may account for a “slowing” of English acquisition among some students related to literacy learning versus an acceleration of English development in relationship to reading and writing among other students. However, the LAO report is silent on the relationship between scores on the Spanish Assessment of Basic Education (SABE) and the CELDT reading and writing scores. The SABE is administered to ELL who are native Spanish speakers who have less than one year of residence in California and/or who are receiving reading and writing instruction in Spanish. Based on the research in biliteracy and the interaction between native-language literacy and academic achievement (Cummins, 2000), we can project that ELL students who have grade-level Spanish reading and writing skills will advance in English reading and writing at a faster rate. The LAO does not enter into this realm of data analysis, although the report does suggest that a major goal of the California schools is for ELL to learn the academic content of the standards-based state curriculum. Undoubtedly, students can learn academic content in their native language as well as in English. The LAO side-steps the question as to why, as a matter of public policy, most ELL are denied access to academic content learning in their native language while they are acquiring English.
Several researchers (Freeman & Freeman, 1999; Gándara, 2000; Gutiérrez, Asato, & Baquedano-López, 2000; McQuillan, 2000, Quezada, 2001) document the combined effects of reading instruction policy and high-stakes testing on the quality of literacy instruction being provided to language minority students. Cummins (1999) points out how literacy instruction for bilingual learners must be grounded in a solid theoretical paradigm to accommodate the linguistic and cognitive factors that pose particular challenges for these emergent readers. An emphasis on discrete-point phonics and phonemic awareness as a departure point for learning to read in English as a second language for limited English proficient students is ill-advised and potentially detrimental. This is especially the case where other meaningful and effective forms of literacy instruction are suspended or de-emphasized in order to comply with external mandates and directives that do not take into consideration the learning characteristics and needs of bilingual students (Mora, 2002).
What’s Been Left Out
The LAO report on the CELDT data contains very little discussion of what the rate of English acquisition means in terms of the continuity and coherence of the K-12 curriculum and the necessity for continued literacy instruction into middle school and high school for current and reclassified ELL. Since the English learning process has been demonstrated to cover the entire span of K-6 education, a system that is based on “standards” and standardized achievement testing for a native English speaking population must be questioned. Invariably, a majority of the 41% of the students who speak a language other than English in the home will take time to close the gap between their oral English proficiency and their reading and writing skills. This means that these students cannot be expected to perform “on grade level” along with their native English speaking peers. The average or “normal” learning curve for language, literacy and academic content must be considered in formulating education policy in order to avoid counter-productive and even harmful use of educational “interventions” such as in-grade retention and mandatory summer school. Such measures, especially when applied based on standardized test data of questionable validity for non-English proficient students, may in fact be unjust and detrimental to students’ long-term academic progress. For instance, Cummins (2000) found that ELL were scoring at the level of chance five to six times more than students who were not limited in English proficiency.
The lack of validity and statistical reliability also has profound implications regarding the High School Exit Exam policy, due to go into effect in 2004. We must consider the equity issues involved in denying a student who enters the California school system at seventh grade or later a high school diploma while knowing that he or she would have to learn English and academic content at an accelerated rate, much above average, despite the lack of provision of appropriate “services” to accomplish this nearly impossible task.
Where Do We Go From Here?
The LAO report is a good beginning to a long-overdue discussion of the policy implications of the timeframe for learning English language and literacy among language minority students. A more recent report titled "English Learners in California Schools" published by the Public Policy Institute of California (Jepsen & de Alth, 2005) builds on the analysis in the LAO report. This report focuses on the issue of redesignation and calls for adequate resources to address the challenges to English language learning and academic achievement. However, we must recognize that we need several more years of data from annual administrations of the CELDT before we will have a complete picture of the learning curve for ELL. We also need in-depth analysis of correlations among different measures of academic achievement. A pitfall to avoid is the unsubstantiated comparisons of program models and faulty inferences regarding program effectiveness based on false pedagogical and statistical assumptions and incomplete or inaccurate descriptions of program components. We in the education and policy-making communities have a great responsibility to demand accurate, complete, and unbiased research and fact-finding to provide equal educational opportunities and effective programs of instruction for current and future generations of language minority students.
References
Cummins, J. (2000). Language, power and pedagogy: Bilingual children in the crossfire. Clevedon, England: Multilingual Matters LTD.
Droop, M., & Verhoeven, L. (2003). Language proficiency and reading
ability in first- and second-language learners. Reading Research Quarterly
38 (1), 78-103.
Freeman, D., & Freeman, Y.S. (1999). The California Reading Initiative: A formula for failure for bilingual students? Language Arts, 76 (3), 241-248.
Gándara, P. (2000). In the aftermath of the storm: English Learners in the Post-227 Era. Bilingual Research Journal, 24 (1 & 2), 1-14.
Gutiérrez, K.D., Asato, J., & Baquedano-López, P. (2000). “English for the Children”: The new literacy of the old world order, language policy and educational reform. Bilingual Research Journal, 24 (1 & 2), 87-112.
McQuillan, J. (2000). Mis-READing the data: Why California's SAT-9 scores don't make the case for English immersion. NABE News, 23 (4), 16-17, 23.
Mora, J.K. (2002). Caught in a policy web: The impact of education reform on Latino students. Journal of Latinos and Education, 1(1), 29-44.
Portes, A. (1995). Segmented assimilation among new immigrant youth: A conceptual framework. In R. G. Rumbaut & W.A. Cornelius (Eds.), California's Immigrant Children: Theory, Research, and Implications for Educational Policy, (71-76). San Diego, CA: University of California Center for U.S.-Mexican Studies.
Portes, A., & Rumbaut, R.G. (2001). Legacies: The story of the immigrant second generation. Berkeley, CA: University of California Press.
Quezada, M.S. (2001). Is sounding out words enough? The Multilingual Educator, 2 (1), 12-15.
Rumbaut, R.G., & Portes, A. (Eds.) (2001). Ethnicities: Children of immigrants in America. Berkeley, CA: University of California Press.