Academic Performance of Elementary Students with Limited English

Policy Research Report Page 1

Academic Performance of Elementary Studentswith Limited English Proficiency in Texas Public Schools

Highlights

What appears to impact TAAS performance of LEP and non-LEP students?

Socioeconomic Status. Economically disadvantaged students had lower TAAS passing rates thantheir non-economically disadvantaged classmates regardless of English proficiency.

Ethnicity. Hispanic students had lower passing rates than non-Hispanic students among bothLEP and non-LEP students.

Mobility. Students who remained on the same campus for Grades 1-5 had higher TAAS passing ratesthan their mobile counterparts; this was especially true for non-LEP students.

Attendance. Both LEP and non-LEP students who failed the Grade 5 English TAAS missed anaverage of one or one and a half more days of school each year in Grades 1-5 than students whopassed the test.

Retention. Retained students did not perform as well on the Grade 4 English TAAS as students whowere not retained, even though the retained students had received an additional year of instruction.

Campus Poverty. Percent economically disadvantaged students on a campus may have a negativeimpact on student academic performance independent of the influence on individual students of beingeconomically disadvantaged.

What additional factors appear to impact TAAS performance of LEP students?

Mathematics. LEP students performed better on the mathematics test, which is less languagedependent than the reading test.

Prekindergarten. LEP students who attended prekindergarten had higher Grade 5 TAAS passingrates than those who did not attend.

Prior TAAS Participation. LEP students taking the English TAAS for the first time did not performas well as those who had previously taken the test in English. There is evidence of high growthbetween the first and second years of testing in English.

Special Language Programs. LEP students who were still in bilingual education or ESL programshad lower English TAAS passing rates than former LEP students who had exited those programs.

Academic performance was not examined in relation to type of special language program becausedifferent program goals effect at what grade students exit those programs and history of participationon the English TAAS.

Campus Effectiveness. The performance gap between LEP and non-LEP students on the reading testis smaller on campuses with the highest campuswide TAAS performance; however, those campusesalso retain larger percentages of LEP students than lower performing campuses.

Page 2 Policy Research Report

Introduction

In 1990, just over 6 percent of school-age children in the United States spokea language other than English at homeand had difficulty speaking English(NCES, 1995). Texas was secondonly to California in the number ofschool-age children with limitedEnglish proficiency (LEP) in 1990,and was one of only five states inwhich more than 10 percent of theschool-age population had limitedEnglish proficiency. The other fourstates were California, Hawaii, NewMexico, and New York. Public schoolenrollment of LEP students by statemirrors these numbers (NCES, 1997).

Over 500,000 students enrolled inTexas public schools in 1996-97 wereidentified as having limited Englishproficiency. Although they representover 13 percent of all students, previ-ously little was known about theparticipation of LEP students state-wide in special language programsover time. Furthermore, analysis ofacademic performance of LEP stu-dents was difficult because, once thosestudents achieved proficiency inEnglish, they were no longer identifiedas having limited English proficiency.

This report presents a longitudinaloverview of academic performance ofstudents entering first grade in Texaspublic schools in 1992-93. The reportprovides a demographic profile ofGrade 1 students in 1992-93, andfollows those students through the1996-97 school year. Five-yearpatterns of student enrollment, campusmobility, participation in specialprograms, and grade-level promotionare examined. Analysis of 1996-97Texas Assessment of Academic Skills(TAAS) reading and mathematicsresults focuses on participation andperformance of students who wereidentified as having limited Englishproficiency when they entered Grade 1five years earlier. A profile of 1996-97 Grade 1 students identifies changes

in the demographic characteristics ofstudents entering Texas publicschools over the past five years.Finally, Texas policy related to LEPstudents is reviewed.

Background

In bilingual education programsstudents receive part of their instruc-tion in their native language. InEnglish as a second language (ESL)programs, they receive all instructionin English, but the English is simpli-fied and the content enriched to makeit more understandable. Otherapproaches focus on the developmentof English language skills rather thanacademic content. Studies of bilin-gual education and ESL programsrely on a number of research tradi-tions – cognitive aspects of schoollearning, program evaluation, andresearch on schooling and classroomeffectiveness (August & Hakuta,1997). Within the program evaluationtradition, the case study methodologycontinues to be considered appropri-ate for in-depth study of the complexrelationships between the manystudent, classroom, and school levelvariables in bilingual educationprograms (Carter & Chatfield, 1986;Baker, 1990). However, researchersare advocating a change in the focusof program evaluations to one thatemphasizes identifying programcomponents that are effective in agiven context (August & Hakuta,1997; Baker, 1990). Past studies havefocused on efforts to determine whichtypes of programs (bilingual educa-tion or ESL) are most effective. Thischange in focus has grown in partfrom the inconclusive findings ofmajor studies of bilingual educationand ESL programs and from criti-cisms of the methodologies used inthose studies.

The increased emphasis on schoolaccountability in the 1990s, bothnationally and in Texas, has also ledto an effort to include students with

limited English proficiency in largescale national surveys and assess-ments such as the National Assess-ment of Educational Progress (Olsonand Goldstein, 1996). In Texas, thePublic Education Information Man-agement System (PEIMS) has beenused to collect individual student leveldata on Texas public school studentsfor a sufficient number of years toconduct longitudinal analyses ofeducational progress of LEP studentsfrom the time they enter school untilthey participate in the TAAS, the statecriterion-referenced testing program.Also, Spanish versions of the TAAShave been developed for Grades 3-6 toincrease accountability for studentswho previously were exempt from theTAAS due to limited English profi-ciency.

This report is the first component of alarger study of academic performanceof LEP students in Texas publicschools that will be conducted by theTexas Education Agency (TEA) overthe next year. The study focuses onissues that are of primary interest tostate and local education policymakersrelated to the performance of studentswho enter school with limited Englishproficiency in a system with highperformance standards. The firstcomponent is primarily descriptive innature and limited to the elementarygrades. Longitudinal retention ratesand performance of LEP students whoparticipate in the TAAS are examined.Later components of the study willlook at educational progress of olderLEP students, including high schoolcourse-taking patterns and schoolcompletion, and provide more in-depth analysis of academic perfor-mance.

LEP Policy Development

Describing the educational experi-ences of millions of “Mexican Ameri-can” children in the Southwest ascruelly discriminatory, U.S. SenatorRalph Yarborough from Texas (1992)

(Continued on page 4)


There were 514,139 LEP students enrolled in Texaspublic schools in 1996-97, representing over 13 percent ofthe total student body. This does not include former LEPstudents who were no longer identified as having limitedEnglish proficiency in 1996-97. Between 1992-93 and1996-97 the LEP population increased by 29 percent,compared to an 8 percent increase in total students overthe same period. About 91 percent of LEP students speakSpanish. In 1996-97 over 87 percent of LEP studentswere economically disadvantaged compared to 48 percentof non-LEP students. Students with limited Englishproficiency are less likely to be receiving special educa-tion services (8 percent compared to 12 percent for non-LEP students); however, 87 percent are receiving bilin-gual education or ESL services.

Enrollment of LEP students is highest in major urbandistricts (24 percent of all students) and lowest in ruraldistricts (5 percent). Although LEP enrollment is higheston elementary campuses, there are also substantialnumbers of LEP students on middle and high schoolcampuses. Districts and campuses with overall lowTAAS performance have higher LEP enrollments than

1996-97 LEP Student Enrollmentthose with high TAAS performance. Districts and cam-puses with high percentages of economically disadvan-taged students also have high percentages of LEP students.Campus effectiveness and campus poverty are two factorsthat may be related to student achievement for all studentsregardless of their English proficiency.

Statewide 49 percent of LEP students are in bilingualeducation programs and 38 percent are in ESL programs.About 53 percent of Spanish-speaking LEP students are inbilingual education programs compared to only 10 percentof LEP students with Asian and other languages. Also,LEP students in middle schools and high schools and indistricts with under 1,600 students are more likely to be inESL programs.

In 1996-97, districts budgeted $395 million dollars, or 4percent of their total instructional operating expenditures,for bilingual education and ESL programs. These funds,which include federal, state, and local revenues, coverexpenditures beyond the cost of providing a regularinstructional program. Seven percent of teacher full timeequivalents (FTEs) are allocated to bilingual education andESL programs.

Elem. MiddleSchool

HighSchool

Under21.5%

21.5%-38.6%

38.7%-55.6%

55.7%-78.1%

Over78.1%

Under52.2%

52.2%-64.0%

64.1%-72.4%

72.5%-80.4%

Over80.4%

0%

5%

10%

15%

20%

25%

30%

35%

40%

Source: TEA PEIMS 1995-96 – 1996-97; TAAS 1995-96 Spring, Year-round and Spanish Tests

Campus Type Campus Poverty% Economically Disadvantaged

Campus Performance% Passing All TAAS Tests Taken

LEP as % of'Enrollment


introduced legislation in 1967 torectify the “folklore ... that everyonehas an equal chance to succeed”(p. 323). The most promising area forprogress, he said, was in the field ofeducation. A year later, enactment ofTitle VII of the Elementary andSecondary Education Act, whichbecame known as the BilingualEducation Act of 1968, marked thebeginning of federal efforts to meetthe “special educational needs of largenumbers of children of limited Englishspeaking ability in the United States.”The Act made funds available directlyto school districts with high concen-trations of language-minority studentsfrom low-income families. Thesefederal grants could be used to de-velop and operate bilingual educationprograms and related activities, suchas teacher training, early childhoodeducation, adult education, vocationaleducation, and dropout reduction.

In Texas, which had always beenhome to one of the largest populationsof Hispanics in the country, severalschool districts across the state werealready operating preschool programsunder a 1959 law intended to improvethe communication skills of non-English speaking children before theyentered first grade (Act approved June1, 1959; TEA, 1962). By 1964, twoSouth Texas school districts, LaredoUnited Consolidated and San Antonio,had begun experimenting with bilin-gual education programs in theelementary grades (Leo, 1985). It tookfive more years, however, before the61st Texas Legislature passed HB103, the state’s first bilingual educa-tion bill (Act approved May 22,1969). Before doing so, lawmakerswould have to repeal the “EnglishOnly” statute of 1918, which made it amisdemeanor for any teacher oradministrator to use a language otherthan English in school or to prescribetextbooks not printed in the Englishlanguage, except in high schoolforeign language classes.

House Bill 103 began by acknowledg-ing English as the primary language ofinstruction in school, but went on toemphasize “the fact that instruction inthe earlier years which includes theuse of language the child understandsmakes learning easier.” Accordingly,the legislation allowed, but did notrequire, school districts to providebilingual instruction through Grade 6.TEA approval was required before adistrict could offer bilingual educationin the secondary grades. Although nostate funds were appropriated forimplementing the bill, by 1970 federalTitle VII grants totaling almost $2million were supporting some 27bilingual programs in Texas schooldistricts (Leo, 1985).

In Washington at this time, govern-ment officials began considering theimplications of recent civil rightslegislation for children with limitedEnglish proficiency. The Office forCivil Rights (OCR) in the Departmentof Health, Education, and Welfareexpressed concern that compliancereviews conducted under Title VI ofthe Civil Rights Act of 1964 hadrevealed practices in some schooldistricts that effectively denied anequal educational opportunity to“Spanish-surnamed” students. In amemorandum issued May 25, 1970, toall federally-funded school districtswith “more than five percent nationalorigin-minority group children,” OCRclarified that: “where inability tospeak and understand the Englishlanguage excludes national origin-minority group children from effectiveparticipation in the education programoffered by a school district, the districtmust take affirmative steps to rectifythe language deficiency in order toopen its instructional program to thesestudents” (OCR, [1975]). This corre-spondence set the stage for a series offuture judicial actions that woulddramatically change the course ofbilingual education in the UnitedStates.

The following year, during the 62ndTexas legislative session, Representa-tive Carlos Truan introduced two bills(HB 495 and HB 1024) in a push tostrengthen the state’s bilingual educa-tion laws. Although neither billpassed, the efforts led the State Boardof Education (SBOE) to issue a morecomprehensive policy statement onbilingual education (San Miguel, Jr.,1987). Under the Revised StatewideDesign for Bilingual Education, eachbilingual program was required to:introduce the school environmentusing the child’s first language;develop the child’s language skills inboth the first language and English;teach subject matter and conceptsusing both languages; and help thechild develop a positive self-imagethrough an appreciation of his or hercultural heritage (SBOE, 1971). Soonafter the policy was approved, TEApublished a resource manual to helpschool districts implement bilingualeducation programs (TEA, [1972]).The guide described the importance ofincorporating the student’s nativelanguage in the educational process,asserting that “bilingual education isnot merely using the first language ofa child as a bridge to English and theneliminating the first language asproficiency in English is attained. It isthe total development of the childbilingually so that he can functionwithin his own capabilities in twolanguages.” (p. 2)

At the same time the SBOE wasrefining its position on bilingualeducation, the U.S. Eastern DivisionCourt was considering claims involv-ing the San Felipe and Del Rio schooldistricts brought under United Statesv. State of Texas (1971/1972), anongoing desegregation case. One ofthe major issues in the litigation waswhether the school districts wereproviding Mexican American studentsan equal educational opportunity. OnAugust 6, 1971, Judge WilliamWayne Justice ordered that the twodistricts be consolidated, and four

(Continued from page 2)


months later, instructed the SanFelipe/Del Rio Consolidated Indepen-dent School District to implement acomprehensive program of bilingual/bicultural education. The court’s planclosely followed recommendationssubmitted by the Department ofHealth, Education, and Welfare(HEW) that sought to reinforce the“cultural and linguistic pluralism ofthe student body” (Hardgrave, Jr., &Hinojosa, 1975, p. 42).

On the heels of the court’s ruling, andwith support from the Texas Associa-tion for Continuing Adult Educationand the League of United LatinAmerican Citizens (LULAC), the 63rdLegislature passed Senate Bill (SB)121, the Texas Bilingual Educationand Training Act of 1973 (SanMiguel, Jr., 1987). The Act directedeach school district in which 20 ormore LEP students in the same gradeshared the same language classifica-tion the previous year to institute aprogram of bilingual instructionbeginning with the 1974-75 schoolyear. As defined by legislators,bilingual education was to be a full-time program of dual-languageinstruction in all subjects required bylaw. LEP students, however, were notto be segregated from their English-speaking peers in “predominantlynonverbal” classes, such as art, music,and physical education. While the Actapplied only to Grade 1 the first year,it required that one grade be addedeach succeeding year until bilingualeducation was offered in all elemen-tary grades through Grade 6. The Actdid not address the education of LEPstudents in Grades 7-12.

To fund bilingual education, the stateappropriated $2.7 million for thebiennium (TEA, [1974]); districtswere allocated $15 for each LEPstudent to purchase instructionalmaterials (House Study Group, 1981).The SBOE, meanwhile, was directedto adopt bilingual textbooks to bemade available free under the state

textbook program. To address the needfor qualified teachers, the SBOEprepared to promulgate rules govern-ing bilingual teacher certification, andTEA began establishing bilingualeducation training institutes for publicschool personnel. Another priority foreducators was the curriculum, becauseby the end of the first year of man-dated bilingual education, there werestill no guidelines available for theprogram (House Study Group, 1981).

Early in 1974, the courts again took upthe issue of equal educational opportu-nity for language-minority students.The U.S. Supreme Court, in a civilrights suit brought by a group ofChinese students against the SanFrancisco School District, ruled thatthe district’s failure to provide lan-guage assistance for the studentsviolated federal law by denying them“a meaningful opportunity to partici-pate in the education program of thedistrict.” In the unanimous decision,the justices declared in Lau v. Nichols(1974) that “there is no equality oftreatment merely by providing stu-dents with the same facilities, text-books, teachers, and curriculum; forstudents who do not understandEnglish are effectively foreclosedfrom any meaningful education.” Thecourt found, moreover, that OCR hadcorrectly interpreted the Civil RightsAct of 1964, and that the rules laid outin the agency’s 1970 memorandumessentially carried the weight of law.While both the Supreme Court andOCR clearly admonished schooldistricts to provide some form ofspecial language assistance for LEPstudents, it is important to note thatneither prescribed a specific methodol-ogy or type of program (U.S. Commis-sion on Civil Rights, 1975). Bilingualeducation, up to this point, was stillconsidered only one option availableto school districts.

In August, Congress enacted the EqualEducational Opportunity Act of 1974.The legislation allowed an individual

to initiate civil action if he or she wasdenied equal educational opportunity.Among the circumstances expresslydefined in the Act as constituting sucha denial was “the failure by an educa-tional agency to take appropriateaction to overcome language barriersthat impede equal participation by itsstudents in its instructional program.”The law thus obliged all schooldistricts, not just those receivingfederal funds, to comply with Title VIof the Civil Rights Act and the OCRguidelines of 1970 (Arizona Depart-ment of Education, [1977]).

Also that year, the federal BilingualEducation Act of 1974 superseded the1968 Act. Congress expanded thescope of the program by deleting therequirement that children served comefrom low-income families. In addi-tion, the new law for the first timeprovided a federal definition of anappropriate bilingual educationprogram. Such a program, in part, wasto offer instruction given in the nativelanguages of LEP students “to theextent necessary to allow [them] toprogress effectively through theeducational system.” Combined withevents at the state level, the wave offederal actions that took place in 1974forced many school districts to beginreevaluating not only their educationalgoals, but their obligations as well.

In Texas that year, TEA received aU.S. Commission on Civil Rightsreport that was critical of bilingualeducation programs in several South-western states, including Texas (U.S.Commission on Civil Rights, 1974).The commission found that MexicanAmerican students in these states werestill frequently subject to discrimina-tion and ethnic segregation. In addi-tion to being seriously underfinancedand reaching only a fraction of theLEP student population, many of thebilingual programs failed to ad-equately address Mexican Americanculture and history. The State Boardof Education, before submitting its



The number of LEP students enrolled in Texas publicschools grew from 398,633 in 1992-93 to 514,139 in1996-97 – an increase of 115,506 students. Thehighest growth in both numbers of LEP students andLEP students as a percent of total enrollment tookplace in districts located in major urban areas andtheir surrounding suburbs. Major urban districts arethe state’s eight largest metropolitan districts serving

LEP Enrollment Trendsthe Houston, Dallas, San Antonio, Fort Worth,Austin, and El Paso areas. About one-third of theLEP students in the state are enrolled in majorurban districts. The LEP populations in thesedistricts increased from 19 percent of total enroll-ment in 1992-93 to 24 percent in 1996-97. Majorsuburban districts are other districts in and aroundthe major urban areas. Although these districtsexperienced high growth in total enrollment as wellas LEP enrollment from 1992-93 to 1996-97, theirLEP populations increased from 8 percent of totalenrollment to 11 percent over the 5 years.

There were fewer LEP students enrolled in districtsin the Corpus Christi (Region 2) and Abilene(Region 14) regions in 1996-97 than in 1992-93,and almost no change in LEP enrollment in theLubbock region (Region 17). Enrollment of LEPstudents increased at a slower rate than total enroll-ment in the San Antonio region (Region 20) and atthe same rate in the Edinburg region (Region 1). Inall other regions LEP enrollment increased at afaster rate than total enrollment. The largest in-creases in number of LEP students were in themajor urban regions of Houston, Richardson, FortWorth, and El Paso. However, the number of LEPstudents almost doubled in the Huntsville, Kilgore,and Waco regions.

Districts in which 40 to 80 percent of the studentsare economically disadvantaged had greater in-creases in numbers of LEP students from 1992-93to 1996-97 than districts in which there are more orfewer economically disadvantaged students.Although districts in which 20 to 30 percent of thestudents are economically disadvantaged had asmaller increase in number of LEP students, thisrepresented a 72 percent increase from 1992-93to 1996-97. By comparison, there was only a7 percent increase in districts in which morethan 80 percent of the students are economicallydisadvantaged.

LEP Students 1996-97

Change From

1992-93Percent Change

District Type Major Urban 176,207 42,500 32% Major Suburban 119,332 41,538 53% Other Central City 96,739 14,526 18% Other CC Suburban 55,503 5,779 12% Independent Town 21,631 5,186 32% Non-metro Fast Growing 7,619 2,313 44% Non-metro Stable 28,018 1,640 6% Rural 8,468 1,453 21%

ESC Region 1 Edinburg 115,816 7,612 7% 2 Corpus Christi 6,488 –1,687 –21% 3 Victoria 2,529 478 23% 4 Houston 125,293 33,775 37% 5 Beaumont 2,251 744 49%

6 Huntsville 6,420 3,023 89% 7 Kilgore 7,457 3,702 99% 8 Mt. Pleasant 1,750 729 71% 9 Wichita Falls 1,083 411 61%10 Richardson 74,176 27,468 59%

11 Fort Worth 29,645 10,830 58%12 Waco 5,595 2,699 93%13 Austin 19,610 7,183 58%14 Abilene 1,423 –62 –4%15 San Angelo 3,807 632 20%

16 Amarillo 6,710 2,001 42%17 Lubbock 4,910 8 0%18 Midland 10,621 1,118 12%19 El Paso 52,208 12,921 33%20 San Antonio 35,725 1,350 4%

Pct Econ Disadvantaged Under 20% 23,609 8,603 57% 20% to 30% 24,179 10,112 72% 30% to 40% 17,632 5,491 45% 40% to 60% 115,889 36,314 46% 60% to 80% 213,693 47,013 28% Over 80% 118,515 7,402 7%


legislative recommendations, askedTEA to conduct a study of the defi-ciencies identified by the commission(SBOE, 1974a). In a report to the 64thLegislature in November, the SBOEdescribed the condition of bilingualeducation in the state and urgedlawmakers both to revitalize thesystem with “increased and continuedfinancial support” and to enhance it byincluding kindergarten (SBOE, 1974b,p. 3).

When several bills were introducedthat session amending the bilingualeducation act, key opponents of thelegislation orchestrated what SenatorCarlos Truan later referred to as an“awful tradeoff”(House Study Group,1981, p. 6). House Bill 1126, thepublic school finance bill, passed in1975 with an amendment to thebilingual education laws that addedkindergarten to the mandatory pro-gram, but removed Grades 4-6 (Actapproved June 6, 1975). Althoughstate funds would be provided forbilingual education in Grades 4 and 5,the programs were optional. Bilingualinstruction beyond the fifth gradecould only be offered at the district’sexpense.

In response to the reductions, LULACand the American G.I. Forum joinedthe Mexican American Legal Defenseand Education Fund (MALDEF) inefforts to reactivate United States v.State of Texas, the statewide desegre-gation case that had begun in 1970.New motions filed by the plaintiffs inJune 1975 alleged that the state hadfailed to take steps to remedy discrimi-natory educational practices againstMexican American students, thusdenying them equal educationalopportunity (United States v. State ofTexas, 1981/1982). In seeking relief,the plaintiffs asked the court to requireTEA to implement a plan to provide aprogram of bilingual instruction to allLEP students in Texas. Six yearswould go by before a decision in thecase was reached.

While Texas continued to strugglewith decisions about how best to servea growing LEP population, federalpolicies were quickly becoming morestringent. Following the 1974 SupremeCourt decision in Lau v. Nichols, OCRorganized a task force to developguidelines to help school districtscomply with the court’s ruling (Officefor Civil Rights, [1975]). The resulting“Lau remedies” were distributed to allstate education agencies in 1975, andregional “Lau centers” were estab-lished across the country to providetechnical assistance to districts(Harrington, 1980). Now, each schooldistrict found to be out of compliancewould have to submit a plan to OCRdesigned to eliminate past educationalinequities by meeting the languageneeds of its LEP students or risk theloss of all federal funding under theElementary and Secondary EducationAct (ESEA).

Between 1976 and 1978, federalscrutiny of Texas bilingual programssteadily intensified. TEA officials metperiodically with representatives ofOCR during that time to discussdevelopment of state standards forbilingual education that would satisfycivil rights criteria. Despite theseefforts, however, at least 40 Texasschool districts were cited for noncom-pliance by June of 1978 (Vega, 1983).OCR subsequently moved to withholdsome $14 million in federal fundsavailable under the Emergency SchoolAid Act of 1972, legislation aimed athelping districts with high percentagesof minority students and districtsundergoing desegregation.

As TEA and other state officialsprotested the move in Washington, theState Board of Education was prepar-ing to adopt administrative rules toimplement its “Texas State Plan forBilingual Education.” The SBOE firstadopted the rules on an emergencybasis in June 1978, voting to expandthe mandatory program beyond thestatutory limit of Grade 3 to include

Grades 4 and 5 and to require Englishlanguage development services inGrades 6-12 (SBOE, 1978a). Therules applied to all districts in whichany LEP students were identified, notjust those with 20 or more LEPstudents with the same languageclassification in any one grade level.

By the time the SBOE adopted therules on a permanent basis in Novem-ber of 1978, concern over potentialprogram costs and teacher shortageshad substantially weakened the plan(SBOE, 1978b; Vega, 1983). Manda-tory bilingual instruction was againconfined to Grades K-3 in districtswith 20 or more LEP students with thesame language classification enrolledin the same grade, with programs inGrades 4 and 5 being optional.Although districts would now berequired to offer English languagedevelopment services for all LEPstudents in Grades K-12 who were notprovided bilingual education, no statefunds were made available for theprogram.

Finally, the plan allowed a district toclassify a student as non-LEP whenthe student scored as low as the 23rdpercentile on the reading and languagearts sections of TEA-approvedachievement tests, a controversialprovision that sharply reduced thelevel of proficiency previouslyrequired for reclassification. In amemorandum to school districts twoyears earlier, TEA had requiredminimum scores at or above the 40thpercentile for transfer from a bilingualprogram, explaining that “a childwhose primary language is other thanEnglish should be able to demonstrateEnglish proficiency to an extent thathis integration into and participationin the regular school program will inno way be jeopardized by a deficiencyin English language skills” (UnitedStates v. State of Texas, 1981/1982).



The new Texas State Plan for Bilin-gual Education had been in effect lessthan a year when hearings beganDecember 3, 1979, in the bilingualeducation suit brought by MALDEFunder United States v. State of Texas(1981/1982). During eight days oftrial, the court received numerousdocuments and heard extensivetestimony regarding the historicaltreatment of Mexican Americans inTexas schools and the adequacy of thestate’s response to the educationalneeds of LEP children. Although theimportance of special languageservices in helping LEP studentsparticipate successfully in school wasnot questioned by either side in thecase, there was considerable disputeover how the statewide program wasbeing implemented. Judge Justiceordered the parties to submit their finalbriefs to the court by April 1, 1980.

When the Texas Legislature convenedin January 1981, proponents ofbilingual education held little hope ofgathering enough support to passenhanced legislation (San Miguel, Jr.,1987). Bills proposing to extendmandatory bilingual coverage toadditional grades had failed duringboth the 1977 and 1979 sessions. Inaddition, several national reports hadrecently been published, including anevaluation of Title VII projects by theU.S. Office of Education, that raiseddoubts about the efficacy of bilingualeducation for increasing schoolachievement. Just two weeks into thesession, however, news of a decisionin the United States v. State of Texascase quickly transformed the politicaldebate.

On January 9, Judge Justice issued a67-page memorandum opinion,declaring that the state’s educationalprogram for LEP students, “while animprovement over past practices, iswholly inadequate” (United States v.State of Texas, 1981/1982). The judgefound that, through a history of“pervasive, invidious discrimination”

against Mexican Americans, the statehad violated the equal protectionclause of the Fourteenth Amendmentto the U.S. Constitution. Moreover,the state’s failure to take appropriateremedial measures constituted aviolation of the Equal EducationalOpportunity Act of 1974. The courtwent on to enumerate the “myriaddeficiencies of the [state’s] existingeducational program” in areas such asprogram coverage, LEP identificationprocedures, entrance and exit criteria,and monitoring and enforcement.

Based on these findings, Judge Justiceconcluded that an immediate andcomprehensive plan of relief wasnecessary “to eliminate the discrimi-natory effects of the past and to assurefuture compliance with the laws of theland.” The parties were ordered tosubmit a joint plan by March 2, 1981,or, if they were unable to reachagreement, to submit separate propos-als by March 9. As outlined in thecourt decree, certain elements wouldhave to be included in a suitable planof relief. The plan had to require, firstand foremost, that bilingual instruc-tion be made available to all MexicanAmerican LEP students in Texas for“as long as necessary to fulfill theireducational potential.” To ensureadequate staffing, TEA would have toimplement aggressive strategies forrecruiting and training bilingualteachers. A home language survey hadto be administered to every incomingstudent, not just Spanish-surnamedchildren, and English proficiencywould be determined using teacherobservations, as well as achievementtest results.

To measure their progress, studentsenrolled in bilingual programs had tobe evaluated at the end of each year.Program exit criteria would have toinclude the student’s English languagetest scores, oral proficiency in English,mastery of specific language skills,teachers’ judgments, and parentalopinions. As part of the plan, TEA

would have to visit each schooldistrict in the state at least once everythree years to monitor compliancewith state regulations. Districtsconsistently found to be out of com-pliance with state law would face“severe sanctions, including loss ofaccreditation and funding.”

Unable to negotiate a mutuallyacceptable plan of relief, the parties tothe suit eventually submitted separateproposals. Judge Justice, maintainingthe state had used its plan merely toreassert its support of current bilingualpolicies, entered a remedial order onApril 17 (Leo, 1985). The orderclosely followed the plan laid out inthe judge’s memorandum opinion, butadded several important provisions.School districts were required toestablish language proficiency assess-ment committees to review LEPstudent placement, and parentaladvisory committees to monitorcompliance with the court’s order. Inaddition, TEA was to request thatdistricts offer six-week summerschool sessions for LEP students.

About the same time Judge Justiceissued his remedial order, the TexasLegislature received a special taskforce report on bilingual education(Task Force on Bilingual Education,1981). The 15-member group, duringa month-long evaluation, had identi-fied flaws in the state’s existingprogram and agreed on a number ofrecommended changes. However, thegroup had struggled to reach a consen-sus on the best methodology forteaching LEP students at the second-ary level. While seven members of thegroup felt bilingual instruction wasessential in all grades through highschool, seven others believed theprogram should not be extended pastthe elementary grades (San Miguel,Jr., 1987). The chair of the task forceeventually broke the tie, and the groupwent on record supporting mandatorybilingual education only throughGrade 6.


Encouraged by events in the EasternDivision Court, Senator Carlos Truan,meanwhile, had introduced SB 477 inFebruary. The proposal reflected theambitious guidelines presented inJudge Justice’s memorandum opinionof January 9 (House Study Group,1981). At hearings held by the SenateEducation Committee in March,legislators took issue with numerousaspects of the bill, including the scopeof the proposed program and the costsassociated with it (San Miguel, Jr.,1987). The senators voted to send SB477 to subcommittee for further study.

A month later, when the task forceissued its report on bilingual educa-tion, Senator Truan made the decisionto amend SB 477. The compromiselegislation, he hoped, would winsupport by incorporating many of thereport’s recommendations, includingthe limitations on mandatory bilingualinstruction (San Miguel, Jr., 1987).Despite continued opposition thatforced a series of last-minute changesto the measure, SB 477 was finallypassed on June 1, 1981 (Act approvedJune 12, 1981). The bill requiredschool districts to offer bilingualinstruction through the elementarygrades (at least through Grade 5), butretained the provision in earlier lawrestricting the mandate to districts with20 or more LEP students with thesame language classification in thesame grade. In Grades 7 and 8, schooldistricts had to offer bilingual educa-tion, English as a second language(ESL), or “other transitional languageinstruction” approved by TEA.Students in Grades 9-12 were toreceive ESL services. In addition,school districts were encouraged toestablish preschool, summer school,and extended day or week programsfor LEP students. Instruction in allspecial language classes was toconsider the students’ cultural back-ground and learning experiences.

For the first time, the commissioner ofeducation was authorized to grant

exceptions to the mandatory programof bilingual instruction in the elemen-tary grades. A school district request-ing a one-year exception had toprovide detailed evidence of a short-age of bilingual education teachers inthe district and present a plan toincrease staffing to appropriate levels.During the exception period, thedistrict would have to use “alternativemethods” approved by the commis-sioner to meet the needs of its LEPstudents. The SBOE in the meantimewas required to develop a comprehen-sive plan to meet the teacher supplyneeds created by the law and submit itto the legislature by January of 1983.

Borrowing from Judge Justice’sremedial order, the bill requireddistricts to establish language profi-ciency assessment committees(LPACs) to implement SBOE-adoptedstandards for identifying, assessing,and classifying LEP students. Itincluded program exit criteria similarto those stipulated in the order, whilespecifying that a LEP student had toscore at or above the 40th percentileon the reading and language artssections of a TEA-approved achieve-ment test before being reclassified.Moreover, SB 477 directed TEA tomonitor local compliance by visitingeach school district at least once everythree years.

The legislature appropriated approxi-mately $18 million to fund the bilin-gual education program for thebiennium (House Study Group, 1981).Under the bill, school districts wouldreceive a special allowance of $50 foreach LEP student enrolled in a bilin-gual program. Districts would also forthe first time receive funding for LEPstudents in ESL or special languageclasses in the amount of $12.50 perstudent. These allotments could beused for program and pupil evaluation,instructional materials, staff develop-ment, and supplemental staff ex-penses.

While the SBOE prepared to changeits rules to implement the new law,legal action continued in the EasternDistrict Court (Leo, 1985). JudgeJustice’s remedial order instructingTEA to phase in mandatory bilingualeducation in Grades K-12 beginningwith the 1981-82 school year was stillin effect. In July, the state appealedthe case to the Fifth U.S. Circuit Courtof Appeals, which stayed the judge’sorder just as schools were set to open.Criticizing Judge Justice for relyingon questionable evidence, the courtoverturned the case on July 12 thefollowing year (United States v. Stateof Texas, 1981/1982). The three-judgeappeals panel declared that the“fundamentally flawed” facts behindJustice’s ruling “form a slender basisindeed for the sweeping statewideorder imposed by the trial court.”

Adopted on an emergency basis inJuly 1981, the SBOE rules on bilin-gual education contained two provi-sions that caused Senator CarlosTruan to accuse the SBOE of “sub-verting legislative intent” (HouseStudy Group, 1981, p. 21; SBOE,1981a). First, the rules continued toallow school districts under certaincircumstances to consider reclassify-ing a LEP student who scored as lowas the 23rd percentile on the readingand language arts sections of anachievement test. Originally intro-duced in the SBOE’s 1978 TexasState Plan for Bilingual Education, theprovision had been denounced forkeeping many LEP students fromneeded services. Senate Bill 477appeared to clarify the question ofproficiency by specifying among theprogram exit criteria achievement testscores at or above the 40th percentile.

The eligibility criteria for students inkindergarten and first grade were asecond source of concern (HouseStudy Group, 1981). Under SBOErules, students in Grades K-1 could beclassified as LEP or non-LEP basedexclusively on a test of oral English



The makeup of the LEP population in Texas publicschools in 1996-97 varies by region. The border regionsof Edinburg (Region 1) and El Paso (Region 19) have thehighest concentrations of LEP students; in 1996-97 overone-third of their students were identified as havinglimited English proficiency. However, with over 125,000LEP students, it was the Houston region (Region 4) thathad the largest number of LEP students. At the other endof the spectrum are the Beaumont (Region 5), Mt. Pleas-ant (Region 8), Wichita Falls (Region 9) and Abilene(Region 14) regions, where LEP students number fewerthan 3,000 and represent only 3 percent of the studentsenrolled in 1996-97.

Spanish is the native language of 91 percent of LEPstudents statewide, and is the native language of over 80percent of students in all regions except Beaumont. Only66 percent of the LEP students in the Beaumont regionspeak Spanish while 27 percent speak Asian languages.

Education Service

6% - 10%

5% and Under

Over 30%

11% - 30% 17

18

19

16

9

13

14

15

20

1

2

3

4

56

7

8

12

1011

19

1

Asian students represent 3 percent of LEP studentsstatewide. It is the Houston, Richardson (Region 10), andFort Worth (Region 11) regions that have the largestnumbers of Asian students; 79 percent of all Asian LEPstudents attend school in those three regions.

The type of special language services LEP studentsreceive vary based on number of LEP students in the areawith the same language, grade-level of students, andavailability of teachers. At least half of the LEP studentsin the Edinburg, Austin (Region 13), San Angelo(Region 15), Midland (Region 18), El Paso, and SanAntonio (Region 20) regions were in bilingual educationprograms in 1996-97, while three-fourths of the LEPstudents in the Mt. Pleasant and Abilene regions are inESL programs. The El Paso region has the largest per-centage of LEP students not in any type of special lan-guage program.

Test administration practices for LEP students also varyby region. In 1996-97 about half (51%) of LEP studentsin grades at which the TAAS is administered took the

reading test in English and 16 percenttook the reading test in Spanish. Theremaining LEP students were exemptedfrom the TAAS by the LPAC (22%) orARD (7%), or were not tested for someother reason such as absence (4%).Students exempted from the TAAS areadministered a locally selected alterna-tive assessment.

Currently only non-special educationEnglish TAAS results for studentsenrolled in the district by October areincluded in the TAAS performanceindicator used to accredit districts andrate campuses. Statewide 13 percent ofLEP students were identified as specialeducation in 1996-97; however, thisvaried from a low of 6 percent in theBeaumont region to a high of 27percent in the Corpus Christi(Region 2) region.

LEP as % ofEnrollment


Center Regions

Source: TEA PEIMS 1996-97

noigeRCSEforebmuN

stcirtsiDforebmuN

stnedutS

PELfo%sa

tnemllornE

hsinapSfo%sa

PEL

naisAfo%sa

PEL

rehtOegaugnaL

fo%saPEL

laugniliBfo%sa

PEL

LSEfo%sa

PEL

margorPoNfo%sa

PEL

grubnidE1 83 075,182 %14 %99 %0 %1 %36 %82 %9

itsirhCsuproC2 24 174,211 %6 %09 %2 %8 %14 %63 %32

airotciV3 04 956,75 %4 %29 %6 %2 %62 %65 %81

notsuoH4 55 974,908 %61 %98 %6 %5 %84 %04 %21

tnomuaeB5 92 655,78 %3 %66 %72 %7 %13 %56 %4

ellivstnuH6 65 280,521 %5 %98 %3 %8 %83 %94 %31

erogliK7 69 626,751 %5 %19 %2 %7 %93 %05 %11

tnasaelP.tM8 84 214,55 %3 %49 %0 %6 %22 %57 %3

sllaFatihciW9 04 406,24 %3 %28 %9 %9 %81 %36 %91

nosdrahciR01 18 823,235 %41 %28 %5 %31 %73 %65 %7

htroWtroF11 77 480,963 %8 %58 %9 %6 %33 %36 %4

ocaW21 87 779,131 %4 %58 %4 %11 %02 %45 %62

nitsuA31 65 134,042 %8 %68 %5 %9 %05 %33 %71

enelibA41 34 297,05 %3 %68 %2 %21 %21 %67 %21

olegnAnaS51 34 976,25 %7 %89 %0 %2 %35 %93 %8

olliramA61 56 969,08 %8 %29 %5 %3 %52 %85 %71

kcobbuL71 06 097,38 %6 %29 %1 %7 %52 %15 %42

dnaldiM81 33 154,48 %31 %89 %0 %2 %25 %52 %32

osaPlE91 21 045,251 %43 %79 %0 %3 %85 %41 %82

oinotnAnaS02 15 674,513 %11 %59 %1 %4 %85 %42 %81

Administration of the TAAS 1996-97 Reading Testto LEP Students

Source: TEA PEIMS 1996-97; TAAS 1996-97 Spring, Year-round, and Spanish Tests

Education Service Center Regions

0%

20%

40%

60%

80%

100%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Not Tested Other

Exempt ARDExempt LPAC

Tested SpanishTested Special EducationTested English

% of'Students

1996-97 LEP Student Enrollment


proficiency; reading and writingproficiency were not considered.Despite “more than 300 letters andtelegrams” of protest, these provisionswere left unchanged when the SBOEfinally adopted the rules in November(House Study Group, 1981, p. 22;SBOE, 1981b).

With new bilingual education policiesin place, TEA met with the Office forCivil Rights to renew collaboration inthe “voluntary enforcement of civilrights laws” (Office for Civil Rights,1982). Two days of discussion inMarch 1982 led to a memorandum ofunderstanding between the agencies.Under the agreement, school districtsdetermined by TEA to be in compli-ance with SB 477 and the new SBOErules would, in effect, also be incompliance with federal requirementsunder Title VI of the Civil Rights Actof 1964.

During a special session in the sum-mer of 1984, the Texas Legislaturepassed sweeping education reforms inHB 72, the Education Reform Act,that included two new initiatives foryoung children with language difficul-ties. A prekindergarten program wasestablished for four-year-olds whohave limited English proficiency orcome from low-income families. Inaddition, the bill directed schooldistricts to offer eight-week summerprograms for LEP children ages fourthrough six who would be eligible toenter kindergarten or first grade thefollowing year. House Bill 72 alsodramatically altered state funding forbilingual education. The previous per-pupil allowance was replaced with aweighted pupil formula that allocatedthe adjusted basic allotment multipliedby 0.1 for every student in a speciallanguage program. As a result, statefunding for bilingual programs wouldincrease from $7 million in 1983-84 to$37 million in 1986-87 (TEA, 1987).

Over the next decade, the legislatureand SBOE supplemented bilingual

education policy with initiatives inareas such as assessment and thecurriculum. Following a pilot study inspring 1986, the SBOE began devel-oping Spanish-language versions ofthe statewide testing instrumentknown as the Texas EducationalAssessment of Minimum Skills(TEAMS) to help measure theprogress of LEP students (SBOE,1986a; TEA, 1987). The SBOE at thesame time adopted rules that allowedschool districts under certain circum-stances to consider exempting LEPstudents from TEAMS (SBOE,1986b).

In 1989, the legislature required TEAto establish and evaluate pilotprekindergarten programs for three-year-old children (Act approved June16, 1989, ch. 813; Act approved June16, 1989, ch. 1179). The SBOEadopted new curriculum essentialelements for bilingual education andESL in 1991 and continued to developalternative teacher certificationprograms to help offset bilingualteacher shortages (SBOE, 1991a;SBOE, 1991b; SBOE, 1991c; SBOE,1991d; SBOE, 1991e). Finally, in1994, the SBOE approved a plan todevelop Spanish-language versions ofthe Texas Assessment of AcademicSkills (TAAS), which had replacedTEAMS during the 1990-91 schoolyear (SBOE, 1994).

Current Texas Policy

Today in Texas, districts are requiredto offer bilingual education programsin the elementary grades if 20 or morestudents with the same language areenrolled in the same grade. English asa second language programs areoffered for LEP students in thesecondary grades, and at the elemen-tary level when there are too fewstudents with the same languageenrolled at the same grade level tooffer a bilingual program.

In July 1997 the SBOE adopted newTexas essential knowledges and skills(TEKS) for Spanish Language Arts,which are to be used in bilingualSpanish instruction, and for ESL (19TAC Chapter 128), replacing essentialelements of the state mandatedcurriculum that had been in placesince 1986. The TEKS are moredetailed and more rigorous than theessential elements they are replacing,and establish learning standards orexpectations for students rather thanmaterial to be presented (TEA,1997g).

Districts that are unable to providerequired bilingual education programsbecause there are not sufficientnumbers of teachers at the schoolfluent in the native languages of thestudents must apply to the commis-sioner for an exception to the pro-gram. In this situation, certifiedpersonnel are assigned to the lowestgrade levels first, beginning withprekindergarten. Districts that do nothave a sufficient number of certifiedteachers to provide required ESLprograms must apply to the commis-sioner for a waiver of certificationrequirements for the teachers who willprovide ESL services to LEP students.

Currently, a LEP student in Grades3-8 may be (1) exempted from theTAAS and administered an alternativeassessment, (2) administered theSpanish version of the TAAS (avail-able for Grades 3-6), or (3) adminis-tered the English TAAS. No combi-nation of options one and two may beused for more than three administra-tions of the TAAS. After that time,the student must be administered theEnglish version of the test. StateBoard of Education rules allow onepostponement of the exit-level test forrecent immigrants (students who haveentered the country within 12 monthsof the date the test is administered).

The local language proficiencyassessment committee (LPAC) has



primary responsibility for determiningthe eligibility of LEP students toparticipate in the statewide assess-ment program and for identifying theappropriate alternative assessment forstudents exempted from the TAAS.Alternative assessments must beselected from a list of commercialinstruments approved by TEA. TheTexas Education Agency receivesaggregate information concerning thenumber of students in each grade whoare administered each alternativeassessment and the number demon-strating improvement in reading,writing, and mathematics. Thisinformation is not published becausethere is not a consistent basis onwhich to compare results of thedifferent tests.

Test results for LEP students who areenrolled in the district by the end ofOctober and take the English TAASare included in the state accountabilityrating system. Results for LEPstudents are included in the baseTAAS indicator used to determinedistrict accreditation status andcampus performance ratings. TheTAAS performance indicator – thepercentage of students passing eachtest (reading, writing, and mathemat-ics) summed across grades – isevaluated for individual studentgroups (African American, Hispanic,White, and economically disadvan-taged), as well as for all studentstested. The English TAAS results arenot disaggregated based on nativelanguage or level of English profi-ciency.

Spanish versions of the TAASreading, mathematics, and writingtests have been developed for Grades3-6. Beginning in 1996-97, SpanishTAAS results are reported on theAcademic Excellence IndicatorsSystem (AEIS) reports. Decisionsregarding use of the Spanish TAAS inthe state accountability rating systemhave not been finalized. The commis-sioner has proposed that Spanish

TAAS results be included in theaccountability rating system and iscurrently seeking input from educatorsregarding when to make this change.He has also initiated development of astandardized reading proficiency testin English (RPTE) that would beadministered to all LEP students whodo not take the English TAAS tomeasure their progress toward achiev-ing English proficiency.

1992-93 Grade 1 Students

Student Characteristics

There were 275,142 students enrolledin Grade 1 for the first time in Texaspublic schools in 1992-93. (Anadditional 24,929 students wererepeating first grade that year.) Thebeginning first graders in 1992-93were 48 percent white, 35 percentHispanic, 14 percent African Ameri-can, and 2 percent Asian. About 94percent were six years old on Septem-ber 1, 1992; most of the remainingstudents were seven years old. Many(83%) attended kindergarten the prioryear, and 23 percent attendedprekindergarten programs two yearsearlier. About half of the first gradestudents (51%) were economicallydisadvantaged and 29 percent wereidentified as being at risk of schoolfailure or dropping out. Students wereidentified as being at risk in Grade 1 ifthey did not perform satisfactorily ona beginning-of-school-year readinessor achievement test, had limitedEnglish proficiency, had been a victimof abuse, or engaged in delinquentconduct (TEA, 1992).

There were 50,352 students, or 18percent of the 1992-93 beginning firstgraders, who had limited Englishproficiency. Spanish was the nativelanguage of 87 percent of the LEPstudents. In fact, almost half (48%)of the Hispanic beginning first gradershad limited English proficiency.Four percent of the LEP studentsspoke Asian languages (Vietnamese,

Laotian, Cambodian, Chinese, Ko-rean, or Japanese), representing half(50%) of all Asian first graders.English was spoken in the homes of 7percent of the LEP students. Most(90%) of the LEP students fromEnglish-speaking homes were His-panic.

Among the 50,352 first gradersconsidered to have limited Englishproficiency in this study were 3,034students who were not identified in1992-93 as having limited Englishproficiency. About one-fourth (26%)of these 3,034 students had beenidentified as having limited Englishproficiency in kindergarten the prioryear; all were identified in latergrades. Most (91%) of the LEPstudents not identified in Grade 1were Hispanic, but almost two-thirds(65%) came from homes whereEnglish was spoken.

As Table 1 on Page 14 shows, theLEP students differed in a number ofways from students who did not havelimited English proficiency. Aswould be expected, most were ethnicminorities, with Hispanic (93%) andAsian (5%) students making up thelargest groups. A much higherpercentage, 87 percent compared to43 percent of non-LEP students, wereeconomically disadvantaged. LimitedEnglish proficiency was one of thecriteria for identifying elementarystudents as being at risk (TEC§29.081, 1996), and 94 percent of theLEP students were identified as beingat risk of school failure or droppingout.

The LEP students started first grade atabout the same age as their non-LEPclassmates. Students with limitedEnglish proficiency and non-LEPstudents also attended public schoolkindergarten the prior year at aboutthe same rate. However, a muchlarger percentage (45%) of LEPstudents attended public schoolprekindergarten programs two years


All Grade 1 Not LEP LEP *Gender: Male 51% 51% 52% Female 49% 49% 48%Race/Ethnicity:** African American 14% 17% 0% Asian 2% 1% 5% Hispanic 35% 22% 93% White 48% 59% 2% Other 0% 0% 0%Age September 1, 1992:** Under 6 0% 0% 0% 6 years old 94% 94% 94% 7 years old 5% 5% 6% Over 7 0% 0% 0%Home Language:** Spanish 87% Asian 4% English 7% Other 3%Economically Disadvantaged 51% 43% 87%At Risk 29% 14% 94%LEP * 18% 0% 100%Bilingual Education 67%ESL 18%Special Education 7% 8% 4%Gifted 4% 4% 2%Migrant 1% 0% 3%Immigrant 2% 0% 9%Kindergarten in 1991-92 83% 84% 82%Prekindergarten in 1990-91 23% 18% 45%Total Students 275,142 224,790 50,352

Table 1Profile of 1992-93 Beginning First Graders

Source: TEA PEIMS 1990-91–1996-97 * Includes students not identified as LEP until later years.** Percentages may not add to 100 due to rounding.

Students with limited English proficiency beginning Grade 1 in1992-93 differed from their non-LEP classmates on a number ofcharacteristics.

earlier, which districts are required tooffer for economically disadvantagedand LEP students.

Fewer LEP students were receivingspecial education services — 4 percentcompared to 8 percent of the non-LEPstudents. Only 2 percent of LEPstudents were served in programs forgifted and talented students, comparedto 4 percent of non-LEP students.

Understandably, LEP students weremore likely to be identified as recentimmigrants; they were also morelikely to be from migrant families.

Campus/District Characteristics

There were 3,268 campuses withstudents beginning Grade 1 in 1992-93; LEP enrollment ranged from noGrade 1 LEP students on 788 cam-

puses to 100 percent of Grade 1students on 10 campuses. Campuseslocated in major urban school districts,the state’s largest metropolitan dis-tricts serving the Houston, Dallas, SanAntonio, Fort Worth, and El Pasoareas, had first grade classes with thehighest average percentage of LEPstudents — 28 percent of their Grade 1students had limited English profi-ciency. Districts located in othercentral cities and their suburbs alsohad first grade classes that were, onaverage, over 20 percent limitedEnglish proficient. The lowest per-centages were in rural districts, wherean average of 8 percent of the begin-ning Grade 1 students in 1992-93 hadlimited English proficiency.

Districts and campuses with highpercentages of economically disadvan-taged students also had higher percent-ages of LEP students among their firstgraders. In their 1995 study of LEPstudents, Moss and Puma identified apoverty factor related to the perfor-mance of students in schools with highconcentrations of economicallydisadvantaged students. This relation-ship is explored further in this study.

Larger percentages of LEP studentswere also found on large campuses,compared to smaller campuses. Onelementary campuses with 800 ormore students, an average of 26percent of the Grade 1 students hadlimited English proficiency in 1992-93, compared to 9 percent on cam-puses with fewer than 200 students.This is due in part to the fact that moreLEP students live in large urbandistricts where elementary campusestend to be larger.

Students with limited English profi-ciency are not distributed uniformlyacross the state. In the border regionsof Edinburg (Region 1) and El Paso(Region 19), the first grade classesconsisted of 65 and 47 percent LEPstudents, respectively. By compari-son, as Figure 1 shows, LEP students


17

18

19

16

9

13

14

15

20

1

2

3

4

56

7

8

12

1011

Education Service Center Regions

LEP Students

% of Grade 1 Enrollment

LEP Students

% of Grade 1 Enrollment

1 Edinburg 12,126 65% 11 Fort Worth 2,475 9%

2 Corpus Christi 942 12% 12 Waco 520 5%

3 Victoria 305 7% 13 Austin 1,721 10%

4 Houston 11,823 20% 14 Abilene 195 6%

5 Beaumont 192 3% 15 San Angelo 581 15%

6 Huntsville 532 6% 16 Amarillo 674 12%

7 Kilgore 523 5% 17 Lubbock 686 11%

8 Mt. Pleasant 144 4% 18 Midland 1,276 20%

9 Wichita Falls 92 3% 19 El Paso 5,223 47%

10 Richardson 6,272 16% 20 San Antonio 4,050 18%

Figure 1Geographic Distribution of Grade 1 LEP Students in 1992-93

1

4

10

19

20

Source: TEA PEIMS 1992-93–1996-97

Large numbers of LEP students live in major urban areas throughout the stateas well as in the border regions.

made up less than 5 percent of thefirst grade classes in the Beaumont(Region 5), Kilgore (Region 7), Mt.Pleasant (Region 8), and WichitaFalls (Region 9) regions. Twentypercent of the first grade students inthe Houston region (Region 4) hadlimited English proficiency in 1992-93; however, this accounted for 24percent of the Grade 1 LEP studentsin the state. Over three-fourths (78%)of the Grade 1 LEP students lived infive regions: Edinburg, Houston,Richardson (Region 10), El Paso, andSan Antonio (Region 20).

One school of research on the effec-tiveness of bilingual education andESL programs emphasizes the studyof program effectiveness within thebroader context of school effective-ness (Carter & Chatfield, 1986;August & Hakuta, 1997). School-wide performance on the TAAS isone measure of the effectiveness ofTexas public schools. Schools withlow overall TAAS performance havelarger percentages of LEP studentsthan schools with high overall TAASperformance. This pattern is alsoevident at the district level. Thismeans bilingual education and ESLprograms are more likely to be foundin districts and schools with lowTAAS performance. This relationshipalso is explored further in this study.

Five Years Later

Student Characteristics

Of the 275,142 students in the 1992-93 first grade class, 237,070 or 86percent were still enrolled in Texaspublic schools in 1996-97. Theremaining 14 percent, or 38,072students, had withdrawn from theTexas public school system and notreturned by 1996-97. Of the 50,352LEP students from the 1992-93 firstgrade class, 44,073 or 88 percent werestill enrolled in Texas public schoolsin 1996-97; half (50%) of those

remaining were receiving speciallanguage services in 1996-97.

The percentage of students identifiedas being at risk of school failure ordropping out had increased from 29percent in 1992-93 to 36 percent in1996-97 for the 1992-93 first gradeclass as a whole. However, thatnumber decreased from 94 percent to67 percent for LEP students. If havinglimited English proficiency was theonly reason a student was identified asbeing at risk, the at-risk designation

would have been dropped when thestudent was no longer considered tohave limited English proficiency.

The percentage of students identifiedas gifted and talented increased forboth the first grade class as a whole(from 4% to 11%) and for the LEPstudents in that class (from 2% to 6%).Growth in students identified as giftedand talented is due primarily toincreased identification of students inthe later elementary grades. However,programs for gifted and talented

Over 4,000 Grade 1LEP Students in 1992-93in ESC Region


students also expanded during thatperiod from representing 7 percent ofstudents statewide in 1992-93 to 8percent in 1996-97.

Special Language Programs

Most of the LEP students who beganGrade 1 in 1992-93 received speciallanguage services at some time duringthe next five years. In Texas, commis-sioner of education rules define twotypes of special language programs –bilingual education and English as asecond language. Bilingual educationprograms are designed to ensure thatstudents master the content of theessential knowledge and skills of thestate-mandated curriculum in theirfirst language, and in English as theirskills progress, while learning En-glish. Students receive content areainstruction in both languages. Englishas a second language programs aredefined as intensive programs ofinstruction designed to developstudent proficiency in English and incontent areas using second languagemethodologies. Students in ESLprograms receive all instruction inEnglish (19 TAC §89.1210).

However, within these broad defini-tions there is wide variation in pro-gram characteristics and goals.August and Hakuta (1997) defineseven types of bilingual education andESL programs. Students in English asa second language programs receiveinstruction in English-language skills,focusing on grammar, vocabulary, andcommunication, rather than academiccontent areas. In content-based ESLprograms, instruction is structuredaround academic content rather thangeneric English language skills.

Students in sheltered instructionarrangements receive subject matterinstruction in English, modified basedon their level of English proficiency.Structured immersion is an approachsimilar to sheltered instruction usedwith groups of students with different

native languages. Although theEnglish instruction is modified basedon level of English proficiency, thereis no native-language support.

In transitional bilingual educationprograms, students receive someinstruction in their native language;however, the goal of the program is totransition to English as quickly aspossible. In maintenance bilingualeducation programs, on the otherhand, the goal is not only to developEnglish proficiency, but also todevelop academic proficiency in thenative language. Therefore, studentsreceive significant amounts of instruc-tion in their native language. Two-way bilingual programs include bothnative speakers of English and stu-dents with limited English profi-ciency. The goal of these programs isto develop proficiency in both lan-guages for both groups of students.

Different approaches are often com-bined at the school level in an effort tomatch student needs with availableteachers. In Texas, districts report

Table 2Years of Special Language Services 1992-93 to 1996-97

for LEP Students Entering Grade 1 in 1992-93

LEP Students

0 - 3 Years

4 - 5 Years

Bilingual Education 22,760 33% 67%

ESL 6,191 53% 47%

Bilingual to ESL 3,688 11% 89%

Combination Bilingual and ESL 1,342 14% 86%

Non-consecutive Services 6,503 58% 42%

No Services 3,589 100% 0%

TOTAL 44,073 43% 57%Source: TEA PEIMS 1992-93–1996-97

Students entering Grade 1 in 1992-93 were placed in one ofsix groups based on (1) the type of special language services(bilingual education or ESL) they received for the next 5 years and(2) whether they received successive years of services beginningin Grade 1 before exiting to the regular classroom. The mostcommon 5 year pattern of services was for students to be in bilingualeducation progams beginning in Grade 1 until they exited to theregular classroom.

student special language programassignments to TEA as either bilingualeducation or ESL, based on thedefinitions in commissioner of educa-tion rules. Undoubtedly, both labelscover programs with a variety ofinstructional approaches and programgoals.

Patterns of special language programparticipation from 1992-93 through1996-97 varied for the LEP students inthe study. Due to the large number ofdifferent program participationpatterns, the LEP students in the studywere placed in one of six groupsshown in Table 2. The groups arebased on the type of special languageservices (bilingual education or ESL)students received over the five yearsfrom 1992-93 to 1996-97 and whetherthey received services continuouslybeginning in Grade 1.

The most common pattern, represent-ing 22,760 students or 52 percent ofthe LEP students who were stillenrolled in 1996-97, was for studentsto be enrolled in bilingual education


programs for one to five years begin-ning in 1992-93. One-fourth (26%) ofthe LEP students in the study whowere still enrolled in 1996-97 hadbeen in bilingual education programsfor all five years from 1992-93 to1996-97. Fourteen percent of the LEPstudents were enrolled in ESL pro-grams for one to five years beginningin1992-93; 5 percent had been in ESLprograms for all five years from 1992-93 to 1996-97.

It was not uncommon for students tobe moved from bilingual education toESL programs at some time during thefive years. Over 8 percent of studentswere moved from bilingual educationprograms to ESL programs after oneor more years and either remained inthe ESL programs through 1996-97 orexited to the regular classroom fromthe ESL program before 1996-97.Commissioner of education rulesdirect districts that are not able toprovide bilingual education programsat all grade levels as required to assigncertified bilingual education teachersto the lowest grade levels first begin-ning with prekindergarten (19 TAC§89.1205). This practice can result instudents moving from bilingualeducation programs in the earlyelementary grades to ESL programs inlater grades. Program changes mayalso be associated with student mobil-ity, which is discussed in the nextsection. Another 3 percent of LEPstudents in the study received fromtwo to five consecutive years ofspecial language services beginning in1992-93, in some other combination ofbilingual education and ESL pro-grams.

Fifteen percent were enrolled inspecial language programs at sometime during the five years but did notreceive services for successive yearsbeginning in Grade 1. This groupincludes students who did not receivespecial language services in 1992-93but were enrolled in bilingual educa-tion or ESL programs in later years. It

also includes students for whom therewas a break in special languageservices, including students who leftthe Texas public school system andlater returned. The remaining 8percent of students did not receivespecial language services at any timethroughout the five years at the requestof their parents.

Over 39 percent of the LEP students inthe study who were still enrolled in1996-97 had been in special languageprograms for all five years from 1992-93 to 1996-97, and 57 percent hadbeen in special language programs atleast four years. The number of yearsstudents received special languageservices varied by type of program, asshown in Table 2. For example, one-third (33%) of students in bilingualeducation programs were in theprograms for three or fewer yearsbefore being mainstreamed into theregular classroom, compared to half(53%) of the students in ESL pro-grams. The students receiving themost years of special language ser-vices were those who moved frombilingual education programs to ESLprograms during the five years, andthose who were in some other combi-nation of bilingual education and ESLprograms beginning in Grade 1.

The types of special language pro-grams in which students were servedvaried by native language of thestudents, or the language spoken in thehome. As Figure 2 on Page 18 shows,Spanish-speaking students were morelikely to be served in bilingual educa-tion programs than in either ESLprograms or combinations of bilingualeducation and ESL programs inGrades 1-5. Students speaking Asianand other non-English languages wereserved predominantly in ESL pro-grams. Students with limited Englishproficiency from homes in whichEnglish is spoken were more likely tohave received no special languageservices between 1992-93 and 1996-97 or to have had a break in services at

Table 3Year First Identified for

Special Education1992-93 Grade 1 Students


Non-LEP students tend to be identi-fied for special education earlierthan LEP students.

Percent of Total Students

LEP Not LEP

Never 84% 79%

1992-93 4% 8%

1993-94 3% 4%

1994-95 3% 3%

1995-96 3% 3%

1996-97 3% 2%

Total Students 44,073 192,997

some time. Districts are required tooffer bilingual education programs forLEP students in elementary grades ifthere are 20 or more students with thesame language in the same grade.Otherwise, they must offer ESLprograms. The smaller numbers ofLEP students with languages otherthan Spanish and the difficulty offinding teachers who speak thelanguages of the students may explainthe large percentages of students withAsian and other languages in ESLprograms.

Special Education

The percentage of students in thestudy receiving special educationservices increased from 7 percent in1992-93 to 15 percent in 1996-97.The gap between LEP and non-LEPstudents had narrowed over the fiveyears, with 13 percent of LEP studentsreceiving special education services in1996-97 compared to 15 percent ofnon-LEP students. Sixteen percent ofLEP students received special educa-tion services at some time between1992-93 and 1996-97, compared to 21percent of non-LEP students. AsTable 3 shows, students who received



Over half of Spanish-speaking students were served in bilingual education programs until they exited to the regularclassroom; students with other languages were more often served in ESL programs. The LEP students from homeswhere English was spoken were enrolled in special language programs at some time during the five years but oftendid not receive services for successive years beginning in Grade 1.

59%

Figure 2Five Year Special Language Program Participation Patterns

for LEP Students Entering Grade 1 in 1992-93by Home Language

English Other LanguagesAsian

5%

70%

9%

2%10%

4%

15%

11%

19%

1%52% 3%

4%

10%

2%

18%

6%

All LEP StudentsBilingual Education

ESL

Bilingual to ESL

Combination Bilingual and ESL

Non-consecutive Services

No Services

Spanish

52%

14%

8%

3%

15%

8%

58%

11%

7%

3%

12%

9%

special education services were mostlikely to have been first identified forspecial education in 1992-93, withsmaller percentages identified in lateryears. This pattern is less distinctivefor the LEP students in the study.Students did not necessarily continuereceiving services from the time theywere first identified for special educa-tion through Grade 5.

Studies conducted in other states havefound that language difficulties aresometimes misdiagnosed as learningdisabilities, resulting in disproportion-ate numbers of LEP students beingassigned to special education pro-grams (TEA, 1994). Analysis of five-year special education programparticipation patterns for LEP studentswho began first grade in 1992-93 does

not reveal any pattern of identificationfor special education services inrelation to participation in, or exitfrom, special language programs. Inaddition, as shown earlier, LEPstudents in the 1992-93 first gradeclass were less likely to have beenidentified as needing special educationservices than their non-LEP class-mates. Because the LEP students are


identified for special educationservices later than non-LEP students,the relationship between participationin special language programs andidentification for special educationservices may be different for olderstudents.

Mobility

Almost three-fourths (73%) of the1992-93 Grade 1 students who werestill enrolled in 1996-97 had beencontinuously enrolled in the sameschool district for the five years, butonly 37 percent had remained on thesame campus. As Figure 3 shows, 35percent of the non-LEP students stillenrolled in 1996-97 had made onecampus move in the five years.

For 14 percent, this was a normativemove. Normative moves are thosethat result from regular studentprogress that all students are expectedto make, for example, moving from anelementary school to a middle schoolbetween fifth and sixth grades. Nor-mative moves are prescribed by theschool system based on the gradeconfigurations of the campuses in eachdistrict. The most common gradeconfigurations for Texas elementaryschools range from the early childhoodgrades through Grade 5 or 6. How-ever, there are also many elementarycampuses that do not include Grade 5.In 1995-96, 18 percent of elementarycampuses that included Grade 1 didnot extend through Grade 5 (TEA,1997c). Consistent with the gradeconfigurations of Texas elementaryschools, normative moves werehighest between the 1995-96 and1996-97 school years, when most1992-93 first graders would be movingfrom Grade 4 to Grade 5 – 14 percentmade normative moves that year.

Non-normative moves are initiated bythe student’s family and includestudents returning to the Texas publicschool system after they have beengone for one or more years.

LEP

Non-LEP

One MoveNo Moves Two or More Moves

Figure 3Campus Moves 1992-93 – 1996-97

1992-93 Grade 1 Students


The LEP students entering Grade 1 in 1992-93 were slightly less mobile thantheir non-LEP classmates over the next five years.

25% 26%40% 10%

NormativeNon-normative

No Moves One Move Two or More Moves

21% 28%

Normative

36% 14%

LEP

Non-LEP

Non-normative

Non-normative moves are fairlyconsistent over the five years with 16to 18 percent of students making non-normative moves each year. By 1996-97, 22 percent of the 1992-93 Grade 1students still enrolled in Texas publicschools had made one non-normativemove. Twenty-eight percent of thestudents had made two or moremoves, including both normative andnon-normative moves.

The mobility patterns of LEP studentswere similar to those of the 1992-93first grade class as a whole. Overall,LEP students were slightly less mobile– 78 percent were continuouslyenrolled in the same district and 40percent on the same campus over thefive years. As Figure 3 shows, thesame percentage of LEP studentschanged campuses once during thefive years as non-LEP students;however, for the LEP students themove was more likely a non-norma-tive move initiated by the family.

There is a relationship between themobility of LEP students and patternsof participation in special languageprograms. Students who received thesame type of services all five years,

whether that was bilingual education,ESL, or no special language program,were more likely to have remained onthe same campus all five years thanother LEP students. Students whomade a non-normative move, andthose who made two or more moves(including both normative and non-normative moves) over the five years,were more likely to have received amix of services (moving betweenbilingual education and ESL programsover the five years) or to have had abreak in services.

Grade-level Retention

Grade-level retention, having astudent repeat a grade he or she wasunable to successfully complete, hastraditionally been the chief remedy foracademic failure and is one indicatorthat a student is not making sufficientacademic progress. The LEP studentswho entered Grade 1 in 1992-93 weremore likely to be retained beforereaching Grade 5 than their non-LEPclassmates. However, student demo-graphic characteristics such asethnicity and socioeconomic status,mobility, and participation inprekindergarten programs were


Table 4Five-Year Retention Rates1992-93 Grade 1 Students


Higher retention rates for LEPstudents can be explained in part byfactors other than English proficiency.

Retention Rates

LEPNot LEP

Year First Retained

Not Retained 82% 89%

Grade 1 12% 7%

Grade 2 3% 2%

Grade 3 2% 1%

Grade 4 1% 1%

Ethnicity

Asian 8% 3%

Hispanic 19% 13%

White 15% 9%

SocioeconomicStatus Economically Disadvantaged 19% 16% Not Economically Disadvantaged 7% 6%

Prekindergarten

Prekindergarten 16% 16%

Did Not Attend 21% 17%

Not Eligible — 6%

Mobility

No Moves 16% 9%

1 Normative 12% 6%

1 Non-normative 21% 14%

2 or More Moves 20% 15%

related to retentions for LEP students.Hispanic students were retained athigher rates than Asian or Whitestudents. Also, LEP students who wereeconomically disadvantaged, mobile,or did not attend prekindergarten wereretained at higher rates than theirclassmates. The gap in retention ratesbetween LEP and non-LEP studentswas highest on campuses on whichschool-wide performance on theTAAS was high.

Eighty-eight percent of the 1992-93Grade 1 students who were still

enrolled in 1996-97 had advanced toGrade 5; most of the remaining 12percent were in Grade 4. Over 8percent were retained in Grade 1 in1993-94. The longitudinal retentionpatterns for the 1992-93 Grade 1students are consistent with annualretention rates reported for Grades 1through 4 from 1992-93 to 1994-95(TEA, 1996).

Until 1996, Texas State Board ofEducation rules restricted the numberof times students could be retained ingrade. Except under special circum-stances, students could not be retainedmore than one time in prekindergartenthrough Grade 4, nor more than onetime in Grades 5-8 (19 TAC §75.195,1996). Only 381 of the 1992-93Grade 1 students who were stillenrolled in 1996-97 had been retainedmore than once between 1992-93 and1996-97.

As Table 4 shows, the retention rateamong the LEP students in the studywas higher than for their non-LEPclassmates. Only 82 percent of LEPstudents were in Grade 5 in 1996-97.Twelve percent of the LEP students inthe study were retained in Grade 1;higher percentages of students werealso retained for the first time inGrades 2 and 3. The pattern of higherretention rates for LEP students holdsfor the three major ethnic groupsrepresented by LEP students – His-panic, Asian, and White, also shownon Table 4.

Schools have traditionally had diffi-culty meeting the academic needs ofeconomically disadvantaged studentsand the higher retention rates amongLEP students can be explained in largepart by the higher percentages of LEPstudents who are economically disad-vantaged. Retention rates are muchhigher for economically disadvantagedstudents among both the LEP and non-LEP groups. Economically disadvan-taged students include those studentswho were reported as economically

disadvantaged at any time from1992-93 to 1996-97.

Among both LEP and non-LEPstudents who began Grade 1 in 1992-93, retentions were related to mobilityover the next five years. Students whohad progressed to Grade 5 by 1996-97were more likely to have remained onthe same campus all five years, ormade one normative move. Studentswho were retained in grade one ormore times during the five years weremore likely to have made a non-normative move or to have changedcampuses two or more times includingboth normative and non-normativemoves. A study of student mobility inTexas public schools (TEA, 1997b)found that mobility rates are higher forethnic minorities and economicallydisadvantaged students, groups whofrequently demonstrate relativelylower academic performance regard-less of their mobility. However, evenwhen controlling for these characteris-tics, the academic performance ofmobile students was found to be worsethan that of stable students.

First graders in the study who attendedprekindergarten programs two yearsearlier were less likely to be retainedin the next five years than theirclassmates who were eligible forprekindergarten programs but did notattend. As Table 4 shows, the differ-ence is greater for LEP students thannon-LEP students. Of those LEPstudents who did not attendprekindergarten, 21 percent wereretained before they reached Grade 5,compared to 16 percent of the LEPstudents who did attend prekindergar-ten.

As might be expected, the LEPstudents in the study who were re-tained were slightly more likely tohave received special languageservices for all five years from 1992-93 to 1996-97 than their classmateswho were not retained. They werealso more likely to have received a



Retention rates for both LEP andnon-LEP students vary by campuscharacteristics.

Table 5Five-Year Retention Rates1992-93 Grade 1 Students

by Campus Characteristics

Retention Rates

LEPNot LEP

Campus Size

Under 200 19% 5%

200 to 400 19% 10%

400 to 600 15% 10%

600 to 800 15% 8%

Over 800 18% 9%

% Economically Disadvantaged

Under 25% 8% 5%

25% to 50% 18% 9%

50% to 75% 18% 11%

Over 75% 16% 12%

Campus % LEP 1996-97

No LEP Students — 9%

Under 5% 16% 8%

5% to 20% 17% 9%

Over 20% 16% 11%

Campus % Passing TAAS

Under 65% 15% 12%

65% to 75% 17% 11%

75% to 85% 18% 9%

Over 85% 20% 6%

mix of bilingual education and ESLservices over the five years or to havehad a break in services.

Students who were retained had loweraverage attendance rates over the fiveyears from 1992-93 to 1996-97 thanstudents who were not retained. Thedifference was greater for non-LEPstudents than for LEP students. BothLEP and non-LEP students who hadnot been retained missed an average offive days of school each year from1992-93 to 1996-97. Retained LEPstudents missed an average of sevendays each year, compared to nine daysfor non-LEP students who wereretained. A study of 1995-96 Grade 1students in Texas public schools foundthat students with higher attendancerates were more likely to be promotedthan students with poorer attendancerates, where other factors were takeninto account (TEA, 1997a). Higherstudent attendance was also related tostudents being judged by their teachersas making more overall progress inschool and as having higher masterylevels of the essential elements.

Campus Characteristics. Fourcampus characteristics were examinedin relation to five-year retention ratesof students who entered first grade in1992-93. The campus size, percentageof students who are economicallydisadvantaged, and percentage ofstudents who have limited Englishproficiency are based on campus-wideenrollment in 1996-97. The percent-age of students passing all TAAS teststaken is based on 1996-97 TAASperformance of students on thecampus who were enrolled in thedistrict by October 25, 1996. Becausemany students changed campuses overthe five years and the retentions couldhave taken place at any grade, reten-tion rates by campus characteristicswere computed for only those 1992-93Grade 1 students who remained on thesame campus for the five years from1992-93 to 1996-97.

Campuses with higher 1996-97 TAASpassing rates retained smaller percent-ages of non-LEP students over the fiveyears from 1992-93 to 1995-96 thanpoorer performing campuses, butslightly larger percentages of LEPstudents. As a result, the gap inretention rates between LEP and non-LEP students is largest on highperforming campuses (those with over85 percent of students passing theTAAS), as shown on Table 5. Basedon grade-level retentions alone, highperforming campuses do not appear tohave more effective programs for LEPstudents than campuses with loweroverall TAAS performance.

There is no relationship betweenpercent of LEP students on the campusand five-year retention rates of 1992-93 Grade 1 students for either LEP ornon-LEP students. However, there isa relationship between percent eco-nomically disadvantaged students on acampus and five-year retention rates.Campuses with fewer than 25 percenteconomically disadvantaged studentsretain fewer LEP and non-LEPstudents than campuses with moreeconomically disadvantaged students.As discussed earlier, this relationshipcan be explained in part by the highretention rates for economicallydisadvantaged students.

There is no clear relationship betweenretention rates and campus size.However, campuses with fewer than200 students have the largest gap inretention rates between LEP studentsand non-LEP students – 19 percent ofthe 1992-93 Grade 1 LEP studentswho remained on the same campuswere retained, compared to 5 percentof the non-LEP students, a 14-pointdifference in five-year retention rates.

Grade 5 TAAS Participation

The Texas Assessment of AcademicSkills (TAAS), a statewide testingprogram that was first implemented in1990-91, focuses on students' higher

order thinking and problem-solvingskills. Each student in Grades 3-8 and10 is required to take the TAASreading and mathematics tests, exceptunder the following circumstances:the local admission, review, anddismissal (ARD) committee mayexempt a special education studentfrom testing; the local LPAC mayexempt a LEP student from both theEnglish and Spanish versions of thetest; or a student may not participatefor other reasons, such as absence onthe day of testing. Eighty-one percentof the LEP students in the study whohad been promoted to Grade 5 took


the Grade 5 TAAS reading test and 83percent took the mathematics test,compared to 94 percent of the non-LEP students. The LEP students whohad been mainstreamed into theregular classroom took the EnglishTAAS at the same rate as their non-LEP classmates. However, economi-cally disadvantaged students were lesslikely to take the Grade 5 TAAS thanstudents who were not economicallydisadvantaged, regardless of Englishproficiency. The LEP students whoattended prekindergarten programswere more likely to be tested andmore likely to be tested in Englishthan those who did not attend.

At the time of TAAS testing, cam-puses submit an answer document foreach student enrolled in the gradestested, whether or not that studenttakes the test. Answer documents forthe 1996-97 Grade 5 TAAS werefound for 83 percent of the 1992-93first graders who were still enrolled inTexas public schools in 1996-97.Students for whom answer documentscould not be found include studentswho were retained in grade at some

Table 6Participation in 1996-97 Grade 5 TAAS Reading Test by

LEP Students Entering Grade 1 in 1992-93

Source: TEA PEIMS 1992-93–1996-97; TAAS 1996-97 Spring, Year-round, andSpanish Tests

TAAS answer documents for the Grade 5 reading test were submittedfor 33,676 LEP students from the 1992-93 first grade class. About54 percent, or 18,267 students, were no longer receiving speciallanguage services and 46 percent, or 15,409, were still identifiedas LEP and receiving services.

Former LEP Still Receiving ServicesType of

Language ProgramEnglish TAAS

English TAAS

Spanish TAAS

LPAC Exempt

Bilingual Education 95% 54% 21% 16%

ESL 94% 55% 0% 35%

Bilingual to ESL 92% 45% 0% 47%

Combination Bil/ESL 92% 49% 12% 30%

Non-consecutive Services 90% 55% 11% 24%

No Services 92% — — —

time between 1993-94 and 1996-97and were, therefore, not eligible totake the Grade 5 test. In addition,students who were enrolled in a Texaspublic school in October 1996 but leftthe system before the spring TAAStesting date would not have a TAASanswer document. Finally, correctionsto student identification numbers on asmall number of TAAS answerdocuments would prevent matchingthose documents to earlier studentrecords.

Of those students in the study withanswer documents, 92 percent tookeither the English or Spanish versionof the 1996-97 Grade 5 TAAS readingtest. The remaining students werereceiving special education servicesand exempted from the TAAS by theadmission, review, and dismissalcommittee (5%), LEP students whowere exempted by the local languageproficiency assessment committee(2%), and students who missed the testdue to absence or other reasons (1%).The percentages are the same for theGrade 5 mathematics test. Becauseparticipation rates on the reading and

mathematics tests are so similar, thefollowing analysis focuses on partici-pation in the reading test unless thereare differences to be noted.

About 81 percent of all LEP studentstook the Grade 5 reading test and 83percent took the mathematics test ineither English or Spanish, compared to94 percent of their non-LEP class-mates. As would be expected, stu-dents who exited special languageprograms before 1996-97 were muchmore likely to have been tested inEnglish than students still receivingbilingual education or ESL services.As Table 6 shows, just about half ofthe LEP students still in speciallanguage programs in 1996-97 weretested in English. For former LEPstudents (those students who hadexited the special language programsby 1996-97), however, the rate ofparticipation in the English TAAS wasover 90 percent.

English TAAS participation of LEPstudents who were still in speciallanguage programs in 1996-97 did notvary significantly by type of program.Just over half of the students enrolledin either a fifth year of bilingualeducation or a fifth year of ESL tookthe Grade 5 TAAS reading test inEnglish in1996-97, for example.However, 35 percent of the ESLstudents were exempted from theTAAS by their language proficiencyassessment committees, compared toonly 16 percent of the bilingualeducation students. The primaryreason is availability of the Spanishversion of the test, which was taken by21 percent of the students still inbilingual education programs.

There is a relationship betweenprekindergarten attendance andparticipation in the Grade 5 TAAS forLEP students. As Table 7 shows, 79percent of the LEP students in thestudy with Grade 5 answer documentswho attended prekindergarten pro-grams took the reading test in English.


LEP

Not LEP

English TAAS

Spanish TAAS

Total LEP

Socioeconomic Status Economically Disadvantaged 74% 7% 81% 91% Not Economically Disadvantaged 89% 0% 89% 97%Prekindergarten Prekindergarten 79% 5% 84% 92% Did Not Attend 72% 8% 80% 91% Not Eligible – – – 97%Mobility No Moves 77% 7% 84% 95% 1 Normative Move 75% 4% 79% 95% 1 Non-normative Move 74% 7% 81% 94% 2 or More Moves 73% 6% 79% 92%

Table 7Participation in 1996-97 Grade 5 TAAS Reading Test

by Students Entering Grade 1 in 1992-93

Source: TEA PEIMS 1992-93–1996-97; TAAS 1996-97 Spring, Year-round, andSpanish Tests

Socioeconomic status, prekindergarten attendance, and mobility are relatedto Grade 5 TAAS participation of LEP students.

Eighty percent took the mathematicstest in English. Another 5 percenttook the tests in Spanish. Of thosewho did not attend prekindergarten, 72percent took the reading and math-ematics tests in English and 8 percenttook the tests in Spanish.

The percentage of LEP students testedvaried by native language of thestudent. About 89 percent of LEPstudents speaking Asian languagestook the 1996-97 Grade 5 TAASreading test in English, compared to74 percent of Spanish-speakingstudents. An additional 7 percent ofSpanish-speaking students took theSpanish reading test. There are similarvariations in participation in themathematics test. One reason largerpercentages of students with nativelanguages other than Spanish took theEnglish tests may be that largerpercentages of those students receivedESL services, which are designed totransition the student to English morequickly than bilingual educationprograms.

Economically disadvantaged studentswere less likely to take the Grade 5TAAS than their classmates who werenot economically disadvantaged. AsTable 7 shows, this is true for bothLEP and non-LEP students. Althougha higher percentage of LEP studentswho were economically disadvantagedtook the Spanish TAAS than studentswho were not economically disadvan-taged, the percentage tested on bothtests combined was still substantiallylower than for non-economicallydisadvantaged students. Six percent ofthe economically disadvantaged LEPstudents had admission, review, anddismissal committee exemptions and12 percent had LPAC exemptions.Eight percent of the economicallydisadvantaged non-LEP students hadARD exemptions.

There is a slight relationship betweenstudent mobility and Grade 5 TAASparticipation for both LEP and non-

LEP students. There is no relationshipbetween Grade 5 TAAS participationand attendance rates for LEP students.On average, students who take eitherthe English or Spanish TAAS andthose who have LPAC exemptions allhave high attendance rates.

Campus Characteristics. The fourcampus characteristics describedearlier – campus size, percentage ofstudents who are economicallydisadvantaged, percentage of studentswho have limited English proficiency,and percentage of students passing allTAAS tests taken – were examined inrelation to 1996-97 Grade 5 TAAS.Analyses are based on all students forwhom Grade 5 answer documentswere submitted.

Although campuses in all size groupstested about the same percentages ofLEP students, smaller campuses weremore likely to test in English. Eightypercent of LEP students on campuseswith fewer than 200 students took theEnglish TAAS, compared to 71percent on campuses with over 800students. Students on smaller cam-

puses were more likely to have beenin ESL programs rather than bilingualeducation programs. There was noclear relationship between school-wide performance for individualcampuses on the 1996-97 TAAS andpercent of the 1992-93 Grade 1 LEPstudents who took the Grade 5English TAAS in 1996-97.

Campuses with 5 percent or fewerLEP students school-wide were morelikely to test their Grade 5 LEPstudents in English than campuseswith higher percentages of LEPstudents. This can probably beattributed to the type of speciallanguage services students are receiv-ing – campuses with smaller percent-ages of LEP students tend to providemore ESL than bilingual educationprogramming. There is no relation-ship between percent economicallydisadvantaged students on a campusand participation of LEP students inthe TAAS, although campuses withfewer economically disadvantagedstudents are more likely to take thetests in English.


Table 81996-97 Grade 5 English TAAS Performance

for Students Entering Grade 1 in 1992-93

Source: TEA PEIMS 1992-93–1996-97; TAAS 1996-97 Spring and Year-round Tests

TAAS performance of both LEP and non-LEP students varied basedon student ethnicity, family socioeconomic status and mobility, andprekindergarten attendance.

Percent Passing

Reading Mathematics

LEP Not LEP LEP Not LEP

All Students 76% 88% 83% 89%

Ethnicity Asian 92% 96% 97% 98% Hispanic 74% 83% 82% 85% White 88% 93% 90% 94%

Socioeconomic Status Economically Disadvantaged 75% 81% 83% 82% Not Economically Disadvantaged 91% 93% 94% 95%

Prekindergarten Prekindergarten 77% 79% 84% 81% Did Not Attend 75% 81% 82% 82% Not Eligible — 94% — 94%

Mobility Stable 78% 90% 85% 91% 1 Normative Move 74% 88% 83% 89% 1 Non-normative Move 77% 88% 84% 88% 2 or More Moves 72% 84% 80% 85%

Grade 5 TAAS Performance

TAAS performance is one of threebase indicators in the AcademicExcellence Indicators System (AEIS)used to determine district accreditationstatus and campus performanceratings. For this study, performance onthe 1996-97 Grade 5 English TAASwas analyzed for all non-specialeducation students who entered Grade1 in 1992-93. Percent passing eachtest, reading and mathematics, arepresented in relation to student charac-teristics, program participation, andcampus characteristics. Economicallydisadvantaged students had lowerTAAS passing rates than their class-mates, as did mobile students, regard-less of English proficiency. Studentspassing the tests also had slightlybetter attendance than those who

failed. LEP students performed betteron the mathematics test than thereading test, and those who hadpreviously taken the test in Englishperformed better than first-timetesters. Also, LEP students who hadattended prekindergarten outper-formed those who had not attended.Two campus characteristics — per-centage of students who are economi-cally disadvantaged and campusTAAS performance — were alsorelated to TAAS performance.

Overall, 86 percent of the 1992-93first graders tested passed the readingtest and 88 percent passed the math-ematics test. As Table 8 shows, LEPstudents did not perform as well on theGrade 5 English TAAS as their non-LEP classmates. The performance gapbetween LEP and non-LEP students

was greater on the reading test, which76 percent of LEP students passedcompared to 88 percent of non-LEPstudents. For the non-LEP students,the passing rate on the mathematicstest was very similar to that of thereading test. The LEP students, onthe other hand, performed consider-ably better on the mathematics test,which is less language-dependent thanthe reading test.

Average TAAS performance variedby ethnicity for both LEP and non-LEP students. Table 8 shows passingrates on the Grade 5 English TAASreading and mathematics tests for thethree ethnic groups with more than100 LEP students tested. Passingrates were lower for LEP studentsthan their non-LEP classmates for allethnic groups shown, with the excep-tion of mathematics passing rates forAsian students, where the differenceis not large enough to be statisticallysignificant. Hispanic students hadlower passing rates than non-Hispanicstudents on both tests among both theLEP and non-LEP groups. Thelargest performance gap between LEPand non-LEP students was for His-panic students on the reading test.

Economically disadvantaged studentshad lower TAAS passing rates thantheir non-economically disadvantagedclassmates regardless of Englishproficiency. There is only a onepercentage point difference in passingrates between LEP and non-LEPstudents on the mathematics test aftercontrolling for socioeconomic status,and for economically disadvantagedstudents the gap is in favor of the LEPstudents. However, there was still asix percentage point difference in thepassing rates on the reading test foreconomically disadvantaged LEP andnon-LEP students.

The Grade 5 TAAS passing rates forLEP students who attendedprekindergarten in 1990-91 are twopercentage points higher than for




Percent Passing

LEP Students

Reading Mathematics

First English TAAS 61% 74%

Former English TAAS Taker 81% 87%

Last Year in Special Language Program

1996-97 55% 70%

1995-96 81% 88%

1994-95 87% 90%

1993-94 88% 91%

1992-93 88% 90%

Never 82% 87%Source: TEA PEIMS 1992-93–1996-97; TAAS 1996-97 Spring andYear-round Tests

There is evidence of high growth in performance for LEPstudents between the first and second years of testing inEnglish, and to a lesser degree between the second andthird years of testing.

those who did not attend. This isconsistent with findings of a five-yearstudy of prekindergarten programs inTexas public schools completed in1994 (TEA, 1995). For non-LEPstudents the relationship is reversed –students who attended prekindergartenhad TAAS passing rates one or twopercentage points lower than thosewho were eligible but did not attend.This can be attributed to the perfor-mance of economically disadvantagedwhite students. There is no differencein passing rates for non-LEP Hispanic,Asian, and African American studentswho attended prekindergarten andthose who were eligible but did notattend. Students who were noteligible for prekindergarten hadsubstantially higher passing rates.

As Table 8 shows, LEP students whoattended prekindergarten had highermathematics passing rates than non-LEP students who attendedprekindergarten. The performancegap on the reading test between LEPand non-LEP students who attendedprekindergarten is only two percent-age points, with non-LEP studentsout-performing LEP students.

The non-LEP students who remainedon the same campus from 1992-93 to1996-97 had higher TAAS passingrates than their mobile counterparts, asshown in Table 8. For LEP studentsthe relationship is less clear; however,stable students had higher passingrates on both the reading and math-ematics tests than students whomoved two or more times. Both LEPand non-LEP students who movedtwo or more times had passing ratesfive to six percentage points lowerthan those of stable students.

About 27 percent of the LEP studentstaking the Grade 5 English TAAS didnot take the Grade 4 English test theprior year. The large performance gapbetween LEP students taking theEnglish TAAS for the first time andthose who took the test the prior year

is shown on Table 9. There is a 20-point difference in percent passing onthe reading test for first-time TAAStakers and more experienced TAAStakers, and a 13-point difference onthe mathematics test. After control-ling for ethnicity, economicallydisadvantaged LEP students who hadpreviously taken the TAAS passed thereading test at the same rate as non-LEP students who were economicallydisadvantaged and passed the math-ematics test at the same or a higherrate.

As Table 9 shows, LEP students whoare still receiving special languageservices in 1996-97 have much lowerEnglish TAAS passing rates on boththe reading and mathematics teststhan former LEP students who haveexited the special language programs.The LEP students who are stillreceiving special language servicesmay be in programs designed todevelop English proficiency over alonger period of time, or they may bestudents who remained in programs

longer because they were slower indeveloping English proficiency.Students who only exited the speciallanguage programs the prior year alsohad lower passing rates on the readingtest than students who exited thoseprograms earlier. Performance growthover time on the English TAAS forLEP students as they make the transi-tion from special language programsto the regular classroom is discussedlater in the next section.

There is a small difference in atten-dance rates of students who passed the1996-97 Grade 5 English TAAS andstudents who failed. Students whofailed the reading test missed aboutone more day of school each year overthe five years from 1992-93 to 1996-97 than students who passed. Stu-dents who failed the mathematics testmissed about one and a half more daysa year than students who passed. Thisgeneral pattern is true for both LEPand non-LEP students in the study.


Percent PassingReading Mathematics

LEP Not LEP LEP Not LEP

Campus Size

Under 200 70% 88% 76% 89%

200 to 400 72% 87% 81% 89%

400 to 600 75% 88% 82% 88%

600 to 800 77% 88% 85% 89%

Over 800 77% 88% 84% 88%

% Economically Disadvantaged

Under 25% 89% 94% 91% 94%

25% to 50% 80% 90% 86% 91%

50% to 75% 76% 86% 83% 87%

Over 75% 74% 79% 82% 79%

Campus % LEP 1996-97

No LEP Students — 91% — 92%

Under 5% 78% 90% 84% 90%

6% to 20% 76% 87% 82% 87%

Over 20% 75% 84% 83% 84%

Campus % Passing TAAS

Under 65% 66% 74% 75% 74%

65% to 75% 76% 84% 84% 86%

75% to 85% 82% 90% 89% 91%

Over 85% 91% 96% 95% 96%




TAAS passing rates for both LEP and non-LEP students vary by campuscharacteristics.

Grade 5 academic performance wasnot examined in relation to speciallanguage program participationpatterns for several reasons. Althoughthe primary goal of all special lan-guage programs is to develop studentproficiency in English, secondaryprogram goals vary. Consequently,different types of programs havedifferent expectations regarding thenumber of years students will receivespecial language services beforetransitioning to English-only instruc-tion. Districts report only whetherstudents are in bilingual education orESL programs. It is impossible todetermine from these data whetherstudents who continue to be served in

special language programs into thelater elementary grades are in pro-grams designed to serve them throughGrade 5 or simply failed to developsufficient English proficiency inprograms intended to exit studentsearlier. Also, since districts areallowed to exempt LEP students fromthe English TAAS for three adminis-trations of the test, LEP studentsentering Texas public schools inGrade 1 are not required to take theEnglish TAAS until Grade 6, andexemption rates vary by type ofspecial language program. Finally,LEP students show high growth inacademic achievement between theirfirst and second TAAS tests. For

these reasons statewide analyses ofperformance of LEP students on theEnglish TAAS in relation to the sixspecial language program participationpatterns was postponed until thosestudents have exited the programs andresults from their second EnglishTAAS are available. For students inthis study, these analyses will be basedon results from the Grade 7 EnglishTAAS administered in 1998-99.

Campus Characteristics. The 1996-97 Grade 5 English TAAS perfor-mance of students who entered Grade1 in 1992-93 was examined in relationto the four campus characteristics —campus size, percentage of studentswho are economically disadvantaged,percentage of students with limitedEnglish proficiency, and percentage ofstudents passing all TAAS tests taken.The LEP students on campuses with600 or more students performed betteron both the reading and mathematicstests than students on smaller cam-puses, as shown in Table 10. For non-LEP students there was no relationshipbetween campus size and TAASperformance.

For LEP and non-LEP students, TAASpassing rates on the reading testdecrease as a percentage of studentson the campus who are economicallydisadvantaged increases. Passing rateson campuses with the fewest economi-cally disadvantaged students (under25%) were 15 points higher thanpassing rates on campuses with themost economically disadvantagedstudents (over 75%). For non-LEPstudents the same pattern exists on themathematics test. For LEP students,the pattern is less pronounced on themathematics test. This pattern re-mains when controlling for studentsocioeconomic status and ethnicity.For example, passing rates on thereading test for LEP and non-LEPeconomically disadvantaged Hispanicstudents are six to eight points higheron campuses with the fewest economi-cally disadvantaged students than on


campuses with the most economicallydisadvantaged students. Thesefindings suggest that a campus povertyfactor may influence student academicperformance independently of theinfluence on individual students ofbeing economically disadvantaged.

There is no relationship betweenpercent of LEP students on the campusin 1996-97 and TAAS performance foreither LEP or non-LEP students aftercontrolling for student ethnicity andsocioeconomic status.

Because the Grade 5 TAAS perfor-mance of students entering Grade 1 in1992-93 contributes to the overallcampus performance, it is not surpris-ing that average performance ofstudents in the study increases ascampus performance increases. Itshould also be noted in Table 10,however, that the performance gapbetween LEP and non-LEP studentson the reading test is smaller oncampuses with the highest TAASperformance (over 85% passing). Thepattern is even more pronounced whencomparing performance of LEP andnon-LEP economically disadvantagedHispanic students, where the perfor-mance gap of eight percentage pointson campuses with the lowest TAASperformance is eliminated on cam-puses with the highest TAAS perfor-mance. This suggests that effectivecampuses, as measured by overallTAAS performance, may have moreeffective programs for LEP studentsthan less effective campuses, based onthis one indicator. To confirm thishypothesis, it would be necessary toexamine the characteristics of effec-tive campuses more closely, includingvariables such as special languageprogram participation and TAASparticipation of LEP students, as wellas other indicators of effectivenesssuch as retention rates.

Performance Improvement

The Texas Learning Index (TLI) wasdeveloped to assess student progressacross grades on the TAAS readingand mathematics tests. A TLI score of70 corresponds to the passing standardat each grade level. Table 11 showsthree years of TLI scores on theEnglish TAAS reading and mathemat-ics tests for students who took the testsall three years, and for LEP studentswho first took the tests in 1995-96 atGrade 4 or in 1996-97 at Grade 5.

Students entering Grade 1 in 1992-93had average TLI scores on the readingtests of 80.3 at Grade 3, 81.7 at Grade4, and 85.7 at Grade 5. As Table 11shows, LEP students did not performas well on the TAAS reading tests astheir non-LEP classmates. This isconsistent with the analysis of percentpassing the test in Grade 5. The LEPstudents first tested at Grade 4 in1995-96 did not perform as well onthe Grade 4 reading test as either LEPor non-LEP students first tested atGrade 3. The LEP students first


LEP students show high TLI growth the first three years they are tested onthe English TAAS.

Table 11TLI Growth Between Grade 3 (1994-95)

and Grade 5 (1996-97)

Reading

Grade 31994-95 Growth


Grade 51996-97

All Students 80.3 1.4 81.7 4.0 85.7

Not LEP 80.8 1.3 82.1 4.0 86.1

LEPFirst Tested 1994-95 75.8 1.9 77.7 4.1 81.8

First Tested 1995-96 74.9 4.6 79.5

First Tested 1996-97Not Receiving Services 73.2First Tested 1996-97Receiving Services 69.4

Mathematics



Grade 51996-97

All Students 75.5 3.5 79.0 3.0 82.0

Not LEP 75.8 3.4 79.1 3.0 82.1

LEPFirst Tested 1994-95 72.7 4.6 77.3 3.8 81.1

First Tested 1995-96 75.6 4.6 80.1

First Tested 1996-97Not Receiving Services 79.8First Tested 1996-97Receiving Services 73.5


tested at Grade 5 who were no longerin special language programs per-formed worse on their first EnglishTAAS than LEP students first tested inearlier grades. The average TLI forthose still in special language pro-grams was below the passing standardof 70, explaining the low percentpassing for this group discussed in thelast section.

Performance improvement on thereading test averaged 1.4 pointsbetween Grades 3 and 4, and 4 pointsbetween Grades 4 and 5. The highergrowth between Grades 4 and 5 is truefor both LEP and non-LEP students.However, the LEP students first testedat Grade 3 showed greater growthbetween their first two TAAS teststhan the non-LEP students. Thehigher TLI growth between the firstand second TAAS test was alsoevident for LEP students first takingthe reading test in Grade 4, although itwas not enough to close the perfor-mance gap between them and LEPstudents first tested earlier.

The TLI performance on the math-ematics test over the three yearsfollows a slightly different pattern, asshown on Table 11. The 1992-93 firstgrade class as a whole scored 4.8 TLIpoints lower on the Grade 3 math-ematics test in 1994-95 than thereading test. Mathematics perfor-mance was still 3.7 points lower thanreading performance on the Grade 5test in 1996-97. As with the readingtest, LEP students scored lower thannon-LEP students on the mathematicstest. However, the performance gapwas smaller, and disappears aftercontrolling for socioeconomic statusand ethnicity of students tested.Among economically disadvantagedHispanic students, for example, theLEP students who first took themathematics test in English at Grade 3in 1994-95 performed slightly betterthan their non-LEP classmates.

The LEP students first tested at Grade4 in 1995-96 did not perform as wellon the Grade 4 mathematics test aseither LEP or non-LEP students firsttested at Grade 3. However, LEPstudents not receiving services whowere first tested at Grade 5 performedas well on the Grade 5 tests as LEPstudents first tested at Grade 4.Furthermore, the later LEP studentstook their first mathematics test inEnglish, the higher their first testscores, with the exception of studentsstill receiving special languageservices in 1996-97. Althoughstudents first taking the mathematicstest in English in 1996-97 who werestill in special language programs didnot perform as well as former LEPstudents who had already exited theprograms, on average their mathemat-ics performance was better than theirperformance on the reading test.

Unlike the reading test, TLI growthwas slightly higher between Grades 3and 4 than between Grades 4 and 5 onthe mathematics test, and cumulativegrowth across the three years washigher. The TLI growth betweenGrades 3 and 4 and Grades 4 and 5was higher for LEP students than non-LEP students. The LEP students firsttested at Grade 4 in 1995-96 alsoshowed high TLI growth betweentheir first and second tests.

Evidence of high growth between thefirst and second years of testing inEnglish supports reports from educa-tors that the transition to testing inEnglish is difficult for LEP students.This has led to suggestions for ex-panding the types of test accommoda-tions allowed for LEP students,especially those taking the EnglishTAAS for the first time (TEA, 1997f).

Spanish TAAS Performance

In 1996-97, the statewide benchmarkadministration of the Grades 5 and 6Spanish TAAS reading and mathemat-ics assessments was conducted. Based

on results from that test administra-tion, the SBOE set a passing standardof 70 for each test. Of the LEPstudents entering Grade 1 in 1992-93,there were 2,237 who took the Grade5 Spanish reading test and 2,223 whotook the Grade 5 Spanish mathematicstest in 1996-97. At the 70 percentpassing standard, 29 percent of thosestudents would have passed thereading test and 31 percent wouldhave passed the mathematics test.

Virtually all of the students taking theSpanish tests were economicallydisadvantaged Hispanic students.Students tested in Spanish wereslightly more likely than other LEPstudents to be enrolled on largercampuses, campuses with higherpercentages of economically disad-vantaged students, and campuses withhigher percentages of LEP students.Almost all (98%) were receivingspecial language services in 1996-97,and about 80 percent had been inbilingual education programs for thefive years from 1992-93 to 1996-97.An analysis of the Spanish test resultsdid not reveal any statistically signifi-cant differences in percent passing bystudent and campus characteristics.This is due in part to the relativelysmall number of students taking theSpanish test and similar characteristicsof the students tested.

Grade 4 TAAS Performance

Answer documents for the 1996-97Grade 4 English TAAS were submit-ted for 42,428 students who enteredGrade 1 in 1992-93 but were retainedin grade before reaching Grade 5. Theprimary goal of retention is to give astudent a year to grow and master theacademic tasks of the current gradelevel before advancing to the nextlevel (TEA, 1997d). Statewide, 1996-97 Grade 4 students did slightly betterthan 1995-96 Grade 4 students on theEnglish TAAS, scoring one TLI pointhigher on the reading test and two TLIpoints higher on the mathematics test.



In response to a 1994-95 survey conducted by the Councilof Chief State School Officers (CCSSO), Texas was oneof 32 states indicating that they have licensing require-ments for teachers of LEP students (CCSSO, 1997). Moststates with licensing requirements, including Texas,require a teaching certificate and bilingual education orESL endorsements. Texas is also one of only 11 stateswith alternative certification programs for individualsseeking ESL certification. Yet a shortage of certifiedbilingual education and ESL teachers continues to be aproblem in Texas. Of the 16,559 bilingual education andESL teachers in Texas public schools in the 1996-97school year, 1,038 bilingual education and 1,348 ESLteachers were teaching on a permit, meaning they werenot certified to teach the classes to which they wereassigned. Bilingual education and ESL teachers ac-counted for 7 percent of all teachers but 22 percent ofteachers on permit. In addition, 74 districts applied forexceptions that year to the requirement that they providebilingual education programs in the elementary grades.These districts reported needing over 2,000 additionalbilingual education teachers to provide the requiredprograms.

In contrast, 1,878 prospective teachers received bilingualeducation certificates in 1996-97. This represents anincrease of 57 percent from 1992-93, when 1,194 bilin-gual education teachers were certified. Although thenumber of prospective bilingual education teacherscertified through the state’s university-based teacherpreparation programs increased by 20 percent from1992-93 to 1996-97, most of the increase was from othersources such as alternative certification programs.

The inability of the state’s teacher preparation programsto provide a sufficient number of qualified bilingualeducation and ESL teachers is due in part to the growth ofthe LEP population. The number of LEP students in thestate increased by 158 percent in the 15 years from 1981-82 to 1996-97, compared to a 30 percent increase in sizeof the total student body. In fact, LEP students accountedfor over one-third of the growth over that period. Growthof the LEP population continues – between 1995-96 and1996-97 the LEP population increased by 7 percentcompared to a 2 percent growth rate for the student bodyas a whole.

Teacher Preparation Issues

Bilingual Education Teachers Certified

Other SourcesUniversity-based Programs

Source: State Board for Educator Certification

ProspectiveTeachers

0

500

1,000

1,500

2,000

1992-93 1993-94 1994-95 1995-96 1996-97

LEP Student Enrollment Trends

Source: TEA PEIMS 1996-97

Total Enrollment

LEPStudents

PercentLEP

1996-97 3,828,975 514,139 13%

1995-96 3,740,260 479,390 13%

1994-95 3,670,196 455,224 12%

1993-94 3,601,840 425,940 12%

1992-93 3,535,876 398,633 11%

1991-92 3,460,378 361,127 10%

1981-82 2,935,547 198,872 7%


However, for students who began firstgrade in 1992-93, retained studentswho took the Grade 4 English TAASin 1996-97 did not perform as well asstudents who took the test on gradelevel the prior year, even though theretained students had received anadditional year of instruction. AsTable 12 shows, this was true forstudents regardless of English profi-ciency, socioeconomic status, orethnicity.

Summary and Conclusions

Eighteen percent of the studentsentering Grade 1 in 1992-93 hadlimited English proficiency. The LEPstudents were predominantly eco-nomically disadvantaged Hispanicstudents and represented almost halfof the Hispanic first graders. Fewerthan 10 percent were identified asrecent immigrants to the UnitedStates.

Socioeconomic Status. The fact thatmost LEP students are economicallydisadvantaged is important becauseeconomically disadvantaged studentshave not performed as well as theirnon-economically disadvantagedclassmates, regardless of Englishproficiency. Both LEP and non-LEPstudents who are economicallydisadvantaged experienced higherretention rates than non-economicallydisadvantaged students. Thoseeconomically disadvantaged studentswho were promoted to Grade 5 in1996-97 were less likely to take theTAAS, and those who took theEnglish TAAS had lower passingrates. This relationship remainedafter controlling for student ethnicity.Statewide academic performance ofeconomically disadvantaged studentson the TAAS accountability indicator(percent passing summed acrossgrades for students enrolled in thedistrict by the fourth week in October)improved dramatically between 1993-94 and 1996-97. A continued focus

on this student group will be impor-tant as their numbers increase. The1996-97 first grade class had a higherpercentage of economically disadvan-taged students than the 1992-93 class– 56 percent compared to 51 percent.

Prekindergarten. Almost one-fourthof the students entering Grade 1 in1992-93 attended prekindergartenprograms two years earlier; 45 percentof the LEP students attendedprekindergarten. Students whoattended prekindergarten were lesslikely to be retained before Grade 5than their classmates who wereeligible for prekindergarten but didnot attend; this was especially true forLEP students. Those LEP studentswho were promoted to Grade 5 in1996-97 were more likely to take theTAAS and more likely to take the testin English if they had attendedprekindergarten. The LEP studentswho attended prekindergarten also hadhigher passing rates on the Grade 5English TAAS than their counterpartswho did not attend.

Prekindergarten attendance increasedin the state from 1992-93 to 1996-97.Almost one-third of 1996-97 begin-ning first graders attendedprekindergarten two years earlier.Prekindergarten attendance of non-LEP students increased from 18percent for 1992-93 first graders to 27percent for 1996-97 first graders.About 48 percent of 1996-97 non-LEPfirst graders were eligible forprekindergarten based on socioeco-nomic status. Prekindergarten atten-dance of LEP students increased from45 percent of 1992-93 first graders to54 percent of 1996-97 first graders.All LEP students are eligible forprekindergarten.

Campus Characteristics. Not onlydistricts located on the Texas/Mexicoborder, but also large urban areas suchas Houston and Dallas have largepopulations of LEP students. Studentswith limited English proficiency are

Reading MathematicsNot

Retained RetainedNot

Retained Retained1995-96 TAAS

1996-97TAAS

1995-96 TAAS

1996-97TAAS

All Students 81 72 78 68

LEP 76 71 76 67

Not LEP 81 73 78 68

Economically Disadvantaged 77 71 75 68

Not Economically Disadvantaged 86 77 82 70

Hispanic 77 71 76 67

Asian 86 75 84 70

White 85 76 81 71Source: TEA PEIMS 1992-93–1996-97; TAAS 1996-97 Spring, Year-round, andSpanish Tests

Retained students who took the Grade 4 TAAS in 1996-97 did notperform as well as their classmates who took the test on grade levelthe prior year.

Table 12Grade 4 English TAAS Average TLI Scoresfor Students Entering Grade 1 in 1992-93



found in disproportionately largenumbers on large campuses that havehigh percentages of economicallydisadvantaged students and poorTAAS performance. Since LEPstudents often attend campuses withpoor TAAS performance school-wide,data for high performing campuseswere examined to see if campuseffectiveness extended to moreeffective programs for LEP students.High performing campuses had lowerfive-year retention rates for non-LEPstudents than campuses with poorTAAS performance, but higherretention rates for LEP students.There was no relationship betweencampus effectiveness and Grade 5TAAS participation for those studentswho were promoted to Grade 5 in1996-97. The performance gapbetween LEP and non-LEP studentson the TAAS reading test is smalleron campuses with the highest TAASperformance, and disappears whencomparing LEP and non-LEP eco-nomically disadvantaged Hispanicstudents. It is difficult to drawconclusions regarding campus effec-tiveness and quality of programs forLEP students from these findings.

Campuses with high percentages ofeconomically disadvantaged studentswere also examined to determine ifcampus poverty influences academicperformance independently of theinfluence on individual students ofbeing economically disadvantaged.For LEP and non-LEP students,TAAS passing rates on the Grade 5English reading and mathematics testsdecrease as campus percent economi-cally disadvantaged students in-creases. This pattern remains whencontrolling for student socioeconomicstatus and ethnicity, which suggeststhat there may be factors associatedwith campus poverty that are relatedto student performance.

Special Language Programs. Texaslaw requires districts to offer bilingualeducation programs to LEP students inthe elementary grades if there are 20students with the same language in thesame grade. Otherwise they mustoffer ESL programs. Just over half(52%) of the 1992-93 first gradersreceived only bilingual educationservices over the next five years, 14percent received only ESL services,and 8 percent were moved frombilingual education programs to ESLprograms in later grades. The remain-der of the LEP students received someother combination of bilingual educa-tion and ESL services (3%), had abreak in services (15%), or receivedno special language services at theparents’ request (8%). Over half(57%) of the LEP students receivedspecial language services for at leastfour years from 1992-93 to 1996-97.

The LEP students who received thesame type of special language servicesall five years, whether that wasbilingual education, ESL, or noservices, were more likely to haveremained on the same campus all fiveyears than other LEP students. Stu-dents who made a non-normativemove and those who made two ormore moves were more likely to havebeen in a mix of bilingual educationand ESL programs or to have had abreak in services during the five years.Students who received a mix ofservices or had a break in serviceswere more likely to be retained.

Grade 5 LEP students who had exitedthe special language programs before1996-97 had very high English TAASparticipation rates. They also hadmuch higher TAAS passing rates onboth the reading and mathematics teststhan students who were not yetclassified as English proficient andwere still receiving special languageservices.

The LEP students who were firsttested in English at Grade 3 had higherTLI scores on the Grade 5 reading andmathematics tests than those firsttested at Grade 4, and students firsttested at Grade 4 had higher TLIscores than those first tested at Grade5. Students first tested in 1996-97who were still receiving speciallanguage services had the lowestaverage TLI scores. Whether theywere first tested at Grade 3 or Grade 4,LEP students showed high TLI growthbetween their first and second readingand mathematics tests. This highgrowth may suggest a need to expandthe types of test accommodationsallowed for LEP students during theirfirst year of testing in English.

Performance on the Grade 5 EnglishTAAS was not examined in relation tothe six special language programparticipation patterns because differenttypes of programs have differentexpectations regarding the number ofyears students will receive servicesbefore transitioning to English. As aresult, exemption rates for the Grade 5English TAAS vary. Also, LEPstudent performance on the EnglishTAAS improves significantly betweenthe first and second test. Therefore,analysis of performance in relation totype of special language services waspostponed until Grade 7 TAAS resultsfor the students in the study areavailable in 1998-99.

Special Education. Fewer LEPstudents are identified as needingspecial education services in Grade 1than their non-LEP classmates – 4percent compared to 8 percent for non-LEP students. Over the next fiveyears, the percentage of studentsreceiving special education servicesincreases, and the gap between LEPand non-LEP students narrows. In1996-97, 13 percent of the LEPstudents and 15 percent of the non-LEP students who entered Grade 1 in1992-93 were receiving specialeducation services. Analysis of five-


year special education programparticipation patterns for LEP studentsdoes not reveal any pattern of identifi-cation for special education services inrelation to participation in, or exitfrom, special language programs.

Retention. Students with limitedEnglish proficiency entering Grade 1in 1992-93 were more likely to beretained before reaching Grade 5 thantheir non-LEP classmates; 18 percentof LEP students were retained.Highest retentions for both LEP andnon-LEP students were at Grade 1.Retained students who took the Grade4 TAAS in 1996-97 did not performas well as students who took theGrade 4 test on grade level the prioryear. These data are consistent withextensive research on the effects ofretention that consistently indicatesthat the practice produces no long-term academic benefits (TEA, 1997d).

Attendance and Mobility. Studentsentering Grade 1 in 1992-93 who wereretained before Grade 5 had lowerattendance rates than students whowere promoted to Grade 5 in 1996-97.Although there was no relationshipbetween attendance and TAASparticipation, students who passed theGrade 5 English TAAS had slightlyhigher average attendance rates thanstudents who failed. Mobile students,especially those who had made two ormore moves over the five years, werealso more likely to be retained, lesslikely both to take the Grade 5 TAASand to take it in English, and morelikely to fail the test than stablestudents. These findings reinforceearlier research concerning theeducational progress of students inTexas public schools (TEA, 1997a;TEA, 1997b).

There were 295,950 students enrolledin Texas public schools in Grade 1 forthe first time in 1996-97, an 8 percentincrease over 1992-93. Twentypercent of the 1996-97 first graderswere identified as having limited

English proficiency. About 6 percentof the 1992-93 Grade 1 LEP studentswere not identified as having limitedEnglish proficiency in Grade 1. If thissame pattern is true for 1996-97 firstgraders, that class could be as high as22 percent limited English proficient,compared to 18 percent in 1992-93.Increased attention will be focused onthis population of students over thenext few years as educators implementthe new Texas Essential Knowledgeand Skills for bilingual and ESLeducation, curriculum and instructionare aligned with the new Spanishversions of the TAAS tests, andSpanish TAAS results are included inthe state accountability rating systemused to accredit school districts andrate campuses.


Selected References

Legal Documents

Act approved June 1, 1959, ch. 481, 1959 Tex. Gen. Laws 1052.


Act approved May 22, 1969, ch. 289, 1969 Tex. Gen. Laws 871.




Bilingual Education Act of 1968, Pub. L. No. 90-247, Title VII, §701 et seq., 81 Stat. 816 (1968).

Bilingual Education Act of 1974, Pub. L. No. 93-380, Title VII, §701 et seq., 88 Stat. 503 (1974).

Bilingual Education and Training Act of 1973, ch. 392, 1973 Tex. Gen. Laws 860.

Civil Rights Act of 1964, Pub. L. No. 88-352, Title VI, §401 et seq., 78 Stat. 252 (1964).

Education Reform Act of 1984, ch.28, 1984 Tex. Gen. Laws 117.

Emergency School Aid Act of 1972, Pub. L. No. 92-318, Title VII, §701 et seq., 86 Stat. 354 (1972).

English Only Act of 1918, ch. 80, 1918 Tex. Gen. Laws 170.

Equal Educational Opportunities Act of 1974, Pub. L. No. 93-380, Title II, §201 et seq., 88 Stat. 514(1974).

Lau v. Nichols, 414 U.S. 563 (1974).

State Board of Education. (1978a). Emergency rules regarding bilingual education, adopted to be effectiveJune 16, 1978, 3 TexReg 2165.

State Board of Education. (1978b). Rules regarding bilingual education, adopted to be effectiveDecember 8, 1978, 3 TexReg 4145.

State Board of Education. (1981a). Emergency rules regarding bilingual education and special languageprograms, adopted to be effective July 17, 1981, 6 TexReg 2614.

State Board of Education. (1981b). Rules regarding bilingual education and special language programs,adopted to be effective December 11, 1981, 6 TexReg 4456.

State Board of Education. (1986b). Rules regarding student assessment of basic skills, adopted to beeffective August 8, 1986, 11 TexReg 3400.

State Board of Education. (1991b). Rules regarding essential elements for English as a second language inGrades 7-8, adopted to be effective December 23, 1991, 16 TexReg 7015 (1991).

State Board of Education. (1991c). Rules regarding essential elements for Grades 7-8, adopted to beeffective December 23, 1991, 16 TexReg 7011.

State Board of Education. (1991d). Rules regarding essential elements for Grades 9-12, adopted to beeffective December 27, 1991, 16 TexReg 7175.

State Board of Education. (1991e). Rules regarding essential elements for prekindergarten throughGrade 6, adopted to be effective December 23, 1991, 16 TexReg 70 [serialized at 16 TexReg 7089(1991)].


Texas Administrative Code. (1997). Austin, TX: Texas Education Agency.

Texas Administrative Code. (1996). Austin, TX: West Publishing Co.

United States v. State of Texas, 342 F. Supp. 24 (E.D. Tex. 1971), aff’d per curiam, 466 F.2d 518 (5thCir. 1972).

United States v. State of Texas, 506 F. Supp. 405 (E.D. Tex. 1981), rev’d on other grounds, 680 F.2d 356(5th Cir. 1982).

Nonlegal Documents

Arizona Department of Education. [1977]. Task force findings specifying remedies available for eliminat-ing past educational practices ruled unlawful under “Lau versus Nichols.” Phoenix, AZ: Author.(ERIC Document Reproduction Service No. ED 158 358)

August, Diane & Hakuta, Kenji, editors. (1997). Improving Schooling for Language-Minority Children:A Research Agenda. Committee on Developing a Research Agenda on the Education of Limited-English-Proficient and Bilingual Students. Board on Children, Youth, and Families. Commission onBehavioral and Social Sciences and Education. National Research Council. Institute of Medicine.Washington, DC: National Academy Press.

Baker, Colin. (1990). “The Effectiveness of Bilingual Education.” Journal of Multilingual andMulticultural Development. 11(4), pp. 269-277.

Carter, T.P. & Chatfield, M.L. (1986). “Effective Bilingual Schools: Implications for Policy andPractice.” American Journal of Education. November 1986: 201-232.

Council of Chief State School Officers. (1996). Systemic Reform and Limited English ProficientStudents. Washington, DC: Author.

Hardgrave, R., Jr., & Hinojosa, S. (1975). The politics of bilingual education: A study of four SouthwestTexas communities. Manchaca, TX: Sterling Swift.

Harrington, C. (1980, August). Bilingual education in the United States: A view from 1980(Urban Diversity Series No. 68). New York: ERIC Clearinghouse on Urban Education.

House Study Group. (1981, December). Bilingual education law in Texas (Special Legislative Rep. No.78). Austin, TX: Texas House of Representatives.

Leo, D.G. (1985). A historical and political analysis of bilingual education in Texas, 1974-1983: Aneducator’s perspective. Doctoral dissertation, University of Houston, Houston.

Moss, M. & Puma, M. (1995). Prospects: The Congressionally Mandated Study of Educational Growthand Opportunity. First Year Report on Language Minority and Limited English Proficient Students.Prepared for Office of the Under Secretary, U.S. Department of Education by Abt Associates, Inc.Cambridge, MA.

National Center for Education Statistics. (1997). A Profile of Policies and Practices for Limited EnglishProficient Students: Screening Methods, Program Support, and Teacher Training. SASS 1993-94.Washington, DC: U.S. Department of Education.

National Center for Education Statistics. (1995). State Comparisons of Education Statistics: 1969-70 to1993-94. Washington, DC: U.S. Department of Education.

Office for Civil Rights. (1970). Memorandum of May 25, 1970, from J. Stanley Pottinger, Director,Office for Civil Rights, to school districts with more than five percent national origin-minority groupchildren. Washington, DC: U.S. Department of Health, Education, and Welfare.


Office for Civil Rights. [1975]. Task force findings specifying remedies available for eliminating pasteducational practices ruled unlawful under Lau v. Nichols. Reprinted in Arizona Department ofEducation, Task force findings specifying remedies available for eliminating past educationalpractices ruled unlawful under “Lau versus Nichols” (pp. 39-50). Phoenix, AZ: Author. (ERICDocument Reproduction Service No. ED 158 358)

Office for Civil Rights. [1982]. Correspondence of [March 15, 1982], from Taylor D. August, Director,Office for Civil Rights, Region VI, to Raymon L. Bynum, Commissioner, Texas Education Agency.Dallas, TX: U.S. Department of Education.

Olson, J.F. & Goldstein, A.A. (1996). “Increasing the Inclusion of Students with Disabilities andLimited English Proficient Students in NAEP.” Focus on NAEP, 2(1). Washington, DC: NationalCenter for Education Statistics.

San Miguel, G., Jr. (1987). ‘Let all of them take heed:’ Mexican Americans and the campaign foreducational equality in Texas, 1910-1981. Austin, TX: University of Texas Press.

State Board for Educator Certification. (1997). Texas Teachers, Reality vs. Myth. Austin, TX: Author.

State Board of Education. (1971, June). Statewide Design for Bilingual Education. Agenda: June 5,1971. Austin, TX: Texas Education Agency.

State Board of Education. (1974a, June). Minutes of State Board of Education, May 11, 1974: Report ofthe Committee on Priorities. Agenda: June 18, 1974. Austin, TX: Texas Education Agency.

State Board of Education. (1974b). Recommendations for legislative consideration on public educationin Texas. Austin, TX: Texas Education Agency.

State Board of Education. (1986a, June). Request for the adoption of tests and approval of passingstandards for Grades 1, 3, 5, 7, 9, and 11 Texas Educational Assessment of Minimum Skills(TEAMS). Minutes of State Board of Education, June 14, 1986. Austin, TX: Texas EducationAgency.

State Board of Education. (1991a, March). An executive summary of alternative certification programsin Texas during school year 1990-1991. Agenda: March 8, 1991. Austin, TX: Texas EducationAgency.

State Board of Education. (1994, September). Adoption of five-year plan for assessment and next stepsfor the curriculum. Minutes of State Board of Education, July 8, 1994. Austin, TX: Texas EducationAgency.

State Board of Education. (1997, September). “Update on Implementation of State Waivers.”Minutes: September 11, 1997. Austin, TX: Texas Education Agency.

Task Force on Bilingual Education. (1981, April). Executive Order WPC-20: Report and recommenda-tions. Austin, TX: Author.

Texas Education Agency. (1962). A report on the Preschool Instructional Program for Non-English-Speaking Children during the first three summers of operation in Texas schools. Austin, TX: Author.

Texas Education Agency. [1972]. A resource manual for implementing bilingual education programs.Austin, TX: Author.

Texas Education Agency. [1974]. 48th biennial report, 1972-1974. Austin, TX: Author.

Texas Education Agency. (1987). Implementing education reform: A report to the Texas Legislaturefrom the State Board of Education and the Texas Education Agency, 1984-1986 (GE7 741 01).Austin, TX: Author.


Texas Education Agency. (1990). Success Stories: A Case Study of Bilingual/ESL Education. FinalProgram Evaluation Report: Bilingual/ESL Education. Austin, TX: Author.

Texas Education Agency. (1990-1996). Public Education Information Management System DataStandards. Annual. Austin, TX: Author.

Texas Education Agency. (1994). Texas Teacher Diversity and Recruitment. Teacher Supply, Demand,and Quality Policy Research Project. Policy Research Report Number 4. Austin, TX: Author.

Texas Education Agency. (1995). Texas Evaluation Study of Prekindergarten Programs: Final Report.Austin, TX: Author.

Texas Education Agency. (1996). 1994-95 Report on Grade Level Retention of Texas Students. Austin,TX: Author.

Texas Education Agency. (1997a). Systemwide Elementary Reform (SER) Grade 2 Interim Report.Statewide Texas Educational Progress Study. Austin, TX: Author.

Texas Education Agency. (1997b). A Study of Student Mobility in Texas Public Schools. Austin, TX:Author.

Texas Education Agency. (1997c). Snapshot ’96: 1995-96 School District Profiles. Austin, TX:Author.

Texas Education Agency. (1997d). 1995-96 Report on Grade Level Retention of Texas Students.Austin, TX: Author.

Texas Education Agency. (1997e). Interpreting Assessment Reports: Texas Assessment of AcademicSkills and End-of-Course Examinations. Austin, TX: Author.

Texas Education Agency. (1997f). Expanding the Scope of the Texas Public School AccountabilitySystem. Policy Research Report Number 9. Austin, TX: Author.

Texas Education Agency. (1997g). Texas Esssential Knowledge and Skills: An Overview for AgencyStaff. Austin, TX: Author.

U.S. Commission on Civil Rights. (1974). Toward quality education for Mexican Americans (MexicanAmerican Education Study Rep. No. 6). Washington, DC: U.S. Government Printing Office.

U.S. Commission on Civil Rights. (1975). A better chance to learn: Bilingual-bicultural education(Clearinghouse Publication No. 51). Washington, DC: Author. (ERIC Document ReproductionService No. ED 107 417)

Vega, J.E. (1983). Education, politics, and bilingualism in Texas. Washington, DC: University Press ofAmerica.

Yarborough, R. (1992). Introducing the Bilingual Education Act. Reprinted in J. Crawford (Ed.),Language loyalties: A source book on the official English controversy (pp. 322-325). Chicago:University of Chicago Press.