The English TheyNeed for the Test

English language learners require instructionin academic English to make the grade on high-stakes tests.

Brian Bielenberg and Lily Wong Fillmore

Since the p;iss:igc of the NoChild Uft Behind Act (NCI.B) in2(101. higli-stakes testing h:iscome to exert an increasinginfluence on education practice

in our schools. Standardized testingnow starts as early as Head Start andcontinues through the high SC1KM)1 exitexams that many states require. Tliiscurrent focus on high-stakes testing inthe United States has troubling conse-quences for the education of Englishlanguage learners.

B\' holding schools accountable forthe academic progress of all categoriesof students, NCLB has the (X)tential tocreate greater education equity.However, in light of the punitivemeasures that can he visited onschools whose students do not farewell on such tests, the heavy emphasison high-stakes examinations isresulting in even greater inequities—especially for English languagelearners. Schools across the UnitedStates have reported low performancelevels on such tests for their linguisticand cultural minority students. 'Lhesestudents include not only Englishlanguage learners but also studentswho speak dialects of English thatdiffer from the kind of Englishpromoted and used in school. It's nosurprise that teachers are wonderinghow they can find time to supportstudents' language development whenthere is so much else to cover.

What the Numbers ShowIn C alifornia. one in four students is anEnglish language learner, for a total of1.6 million such students in the state.Eigure 1 (p. 47) shows the nationalpercentile ranks of average studentscores on the math suhtest of the Cali-foniia Achievement Test in grades 2, 4,6, 8, and 10. 1 he chart compares twogroups: English-only/fluent English-proficient students and English languagelearners. The average scores of Englishlanguage learners on standardized mathtests steadily decline over time, to reachtlieir lowest points in 8th and 10thgrade. Reading and language standard-

ized test scores of English languagelearners in (California have shownsimilar declines as students movetoward lOth grade.

The substantial test score gap in allacademic areas between Englishlanguage learners and other students isworrisome for schools because pertbr-mance on these tests matters. Perhapsmost seriously, high-stakes testing hasthe potential to undermine Englishlanguage learners' opportunities forhigh school graduation and educationbeyond high school. If we look at thetrends in the performance of (CalifornialOth graders who are learning English

A S S O C i l A ' i [ O N F O R S d P C R V I S I O N A N 1 » C t R K I C I L I M l > K V R I . O P M E N T 4 5

on reading, language, and math stan-dardized tests, we can predict that mostEnglish language learners in Californiawill find the California High School ExitExamination difficult if not impossibleto pass because the test covers Englishliteracy and mathematics. At present, 2Sstates require such tests for graduationfrom high school, with other statesplanning to follow suit. These testsinvolve the highest possible stakes inpublic education because they can limitsubsequent economic opportunities forstudents who fail them.

The stakes are equally high forschools. It is a matter of survival forschools, teachers, and curriculums toprovide the necessary instnictionalsupport to improve overall studentacademic performance as measured bysuch tests. So far, schools have imple-mented only relatively low-ievel fixes,such as the adoption of instructionalprograms that emphasize basic skills andphotiics in reading or that focus on factsand procedures in math. Such solutionsmay offer students short-term gains intest scores, especially in the primarygrades. As long as the tests stay close tothe skills and materials covered in basicskills instniction, students performreasonably well, even when their profi-ciency in English is limited. By 4thgrade, however, when tests require ahigher level of proficiency in English anda greater understanding of content, theperformance of English languagelearners on these tests begins to decline.

The Other EnglishWhat English language learners need—and what teachers should provide—istraining in the academic English skillsthat are foundational to literacy, masteryof subject matter, and superior testpertbmiance. The term academicEnglish is based on a distinction thatJim Cummins made when he arguedthat proficiency in language involveslayers of skills and knowledge (1979,1983, 2000). The earliest and most basicacquired language skills consist of thewords, structures, and devices thatfigure in everyday, face-to-face social

Once students reach

upper elementary school,

the linguistic demands of

school quickly change.

interactions, language that is moresocially than cognitively demanding.Children acquire this kind of languageproficiency through interactions withcaregivers, family members, and play-mates, and they usually possess theselinguistic skills and resources by thetinie they first enter school, regardlessof the language spoken at home.

English language learners easilyacquire these basic language skills whenthey come into contact with English atschool, often within a year or two.Althougli these skills suftice in theearliest years, however, they are farfrom all that is needed for the learningthat takes place in the classroom, espe-cially after 3rd or 4th grade. Oncestudents reach upper elementar)'school, the linguistic demands of schoolquickly change. Text materials and thelanguage of instRiction evolve to amarkedly different form so they cancommunicate more advanced andcomplex subject matter. The language

becomes more precise in reference andmore complex in structure. In addition,the vocabulary' begins to include moreLatin- and Greek-based words thanwords with Anglo-Saxon roots (Corson,1997), as well as words expressingabstract concepts.

To make sense of the informationtaught in school, students must becomeproficient in the academic English usedby these texts and instruction—and byhigh-stakes tests.

Tackling Math—In EnglishThe difficulties that English languagelearners are likely to encounter inlanguage tests are obvious. Problemswith language also present themselvesin math tests, however, and these areperhaps more troublesome becausethey can easily go unnoticed. Even thesimplest word problems call for the testtaker to interpret the text and recognizethat ordinary words may have quitespecialized meanings in the math realm,as is evident in the following 6th grademath test item from the MassachusettsComprehensive Assessment System (www.doe. mass. edu/mcas/2003/reiease):

Students in Mr. Jacob s English classwere giving speeches. Each student sspeech was 7 to 10 minutes long.Which of the following is the bestestimate for the total number ofstudent speeches that could be givenin a 2-hour class?

A. 4 speeches

B. 8 speeches

C. 13 speeches

D. 19 speeches

The only technical math term used inthis item is the expression best esti-mate, which tells the test taker that thecorrect answer will be more likely ormore believable than the other answers,given the statement of the problem. Atfirst glance, this problem is simplystated, and both the vocabulary and thecalculations called for are fairly straight-forward: The estimate will fall betweenthe number of 7-minute speeches thatcan be given in 120 minutes (17) andthe number of 10-minute speeches that

4 6 E o r c A T i O N A L L R A D E R S H I I ' / O F C I E M B E R 2 O O 4 / J A N I : A U V 2 O ( ) S

c;m be given in the same time period(12). Therefore. C—13 speeches—is theeorrect answer.

But is it that simple? Not for manyEnglish language learners, Tliey mustunderstand that the time given for eachspeech is a range, meaning that eachspeech is between 7 and 1(1 minuteslong. The test taker must also success-fully interpret the question posed. Todo this, he or she will need to unpackseveral grammatical features typical ofacademic English structures. The first isa complex noun phrase stnicture: Ihebest estimate for the total number ofstudent speeches that could be given ina 2-hour class. Tliis in turn contains acomplex prepositional phrase^"or thetotal number of student speeches thatcould be gii'en in a 2-hour class—thatitself contains a relative clause construc-tion; that could be given in a 2-hourclass. In addition, the complex nounphrase contains the passive construc-tion could be given—a type ofconstruction frequently encountered inacademic linglish. As long a.s the testtaker is familiar with the expression togive a speech and tlie passive construc-tion in l-nglish, this in itself is not diffi-cult to interpret.

The following problem, taken frompreparation materials for the C alifbmiaHigh School Exit Examination (www. cde. ca. gov/ta/tg/Iis/reso urees. asp),provides another illustration:

A submarine is 285 feet under thesurtaee of the oeean. A helicopter isflying at •i,5()() feet above sea level.Given that the heliet>pter is directlyabove the submarine, how far apartarc they?

A. 285 feet

B. 4,215 feet

C. 4,785 feet

D. 4.500 feet

Mathematically, this is a simple arith-metic problem: 285 + 4.500 = 4,785,Yet many students struggle with theproblem. What makes it difficult? Eirst.test takers must be able to envision thescene described in the problem. It callsfor considerable background knowl-

F I G U R E 1 California Achievement Test Results—Math Subtest

National Percentile Ranks/

Average Student Score

All Students

English-Only/

Fluent Engiish-Proficient

English Language Learners

The complete score report

Grade 2

58

65

44

Grade 4

51

59

34

Grade 6

51

59

25

Grade 8

47

54

22

is available at http://star.cde.ca.gov/star2004.

Grade 10

50

55

24

edge, triggered by specific vocabularyand expressions, such as helicopter,submarine, surface of the ocean, andsea level. The test taker must know thatsea and ocean are used as syncjnymshere and that sea lei'el and surface ofthe ocean refer to the same baseline.Further, it would help to know that theterm sea level is an abstraction and thatsea level can apply to places miles awayfrom an ocean—Denver, for example, isa mile above sea level—and that placescan be below sea level but not literallyunder the surface of the ocean—DeathValley, for example, is 282 feet belowsea level

There are other difficulties. Theproblem asks how far apart the heli-copter and the submarine are. It doesnot specify whether the distance is tobe thought of as horizontal or vertical.The student must construe this informa-tion from the positions given for thesubmarine and the helicopter: The firstis 285 feet under the surface of theocean\ the second is 4,500 feet abovesea level. The most important informa-tion is contained in a conditionalconstruction (given that) that Englishlanguage learners are quite unlikely tohave encountered before. Test takersmost likely will have difficulty realizitigthat the information introduced by thisexpression is crucial to interpreting theproblem. Prepositional phrases are alsonotoriously difficult for Englishlanguage learners, and this short mathproblem contains five. Moreover,English language learners may have diffi-

culty interpreting pronoun references.Tlie word they—as in hoivfar apartare they—tends to be especially prob-lematic because it can refer to people orobjects, so students may not understandthat the pronoun in the test item refersto the helicopter and the submarine.

Surviving High-Stakes TestsEnglish language learners must acquirethe language skills needed for everydaycommunication—skills that nativespeakers of English usually bring toschool—as well as those needed forsubject-matter learning. Children do notlearn this kind of language on their ownor through immersion in an English-speaking environment. Masteringacademic English—and thus survivinghigh-stakes te.sts—requires instructionalactivities that actively promote languagedevelopment in the context of learningintellectually challenging content.

Project ChallengeBoston University' researchers andeducators Suzanne C^hapin, (.atherineO'Connor, and Nancy Anderson <2OO3)have been collaborating with classroomteachers in Chelsea. Massachusetts,since 1998 to demonstrate that Englishlanguage learners and other under-achieving students in low-income,urban scht)ols can develop the languageand disct)urse skills typical of higher-performing students. Project Chal-lenge—a math program funded by agrant from the U.S. Department ofEducation and designed to meet anti

A S S O C I A T r O N FOK S l P H K V I S I O N A N D C U R R I C U L U M D E V E L O P M E N T 4 7

surpass the state of Massachusetts'smath standards—sought to pn)videurban students with opportunities thatwould enable them to successfullypursue higher-level math in high schooland college. The demanding mathcurriculum was aligned to the standardsadopted by the National Council ofTeachers of Mathematics.

The researchers and educatorsrecniitcd at-risk students who seemedto have potential talent in mathematicson the basis of a set of ncjnstandardmeasures. Most of the studentsappeared average on these measuresand had significant gaps in their knowl-edge base, lacking knowledge of suchtopics as fractions, for example, or howmath is used in sports. Two-thirds of thestudents participating in the programwere English language learners, mostlyLatinos: the remaining students weremembers of other typically under-achieving minorit)' groups.

The researchers and teachers workedwith four cohorts of 10(1 4th graders. Anew 4th grade cohort was added eachyear as the previous one advanced agrade. Between the ith and 7th grades,the students tackled a challenging mathcurriculum designed to devek)p theirunderstanding of content and improvetheir ability to solve complex mathproblems. Each day, students receivedan hour of scaffolded instmctionalsupport from their teachers that empha-sized integrating their knowledge ofvarious tt)pics and taking responsibilityfor their own learning.

English in the Math ClassProject Challenge teachers encouragedstudents to talk about their reasoningand evolving understanding of the prob-lems that they were working on. Usinga technique called revoiciug. a ProjectChallenge teacher would repeat someor all of what the student had just said,thereby clarifying and modeling a morefnlly realized version of the student'sstatement. The teacher would also askstudents to restate the contributions ofclassmates: "Can you repeat what shejust said in )()iir own words?" Students

Project Challenge

teachers stressed that

understanding counted

more than getting

the right answer.

in the program come to realize that theymust not only listen to one another, butthey must also ivy to understand whathas been said so they can paraphrase orrepeat it for the class.

Project C!;hallenge teachers also askedstudents to comment on their class-mates' understanding of a problem andthe reasoning behind it: "Do you agreeor disagree with him, and why?" In thisway, students become more aware oftheir own and their classmates' thinkingprocesses and of the various strategiestised in problem solving. Project Chal-lenge teachers stressed that math had tomake sense to the students, that under-standing counted more than getting theright answer. Over time, the students inthe program became increasingly adeptat explaining their thinking, and theygained a greater understanding of howmathematical thinking works.

By 2002, the Project Challenge teamhad worked with 400 middle schoolstudents from the <^helsea, Massachusetts,schools. Team members had outstandingresults to report. Project (Challenge usedthe Massachusetts (ComprehensiveAssessment System (MCAS) as well asthe California Achievement Test (CAT),wliich served as a check on MCASscores. In both assessments. ProjectChallenge students achieved meanscores in the 8~th percentile, outper-forming students in the state ofMassachusetts as a whole and eventhose in higli-pertbrming districts. Afterjust one year in the program. 57 percentof Project Challenge students werescoring at the advanced or proficientlevel on the MCAS mathematics test—significantly higher than Massachusetts s

students as a whole, whose averagescore was 38 percent (Chapin &O'Connor, 2004).

Making the GradeWhen schools provide English languagelearners with these kinds of instruc-tional supports, students can developthe language and academic skills theyneed to succeed in school. But it takeseducation leadership: Tlie interventionin Project Challenge schools began withan intensive program of professionaldevelopment for the teachers involved(C:hapin et al., 2003). This trainingcould not have happened withoutadministrative support.

Fortunately, there are a number ofthings that leaders in our schools can doto help English language learnerssucceed in this era of high-stakestesting. Foremost, school leaders needto acknowledge the role that academicEnglish plays in test performance and,of course, in learning in the classroom;they then need to raise awareness of theissue in their schools.

At present, few educators know whatacademic English is or how to distin-guish it from other forms of English.One way of raising awareness is to orga-nize study groups for teachers so theycan look at the language used in text-books for various subjects across thegrades. Beginning with a concisedescription of academic language andits grammatical features (see Biber,Conrad, & Reppen, 1998) should helpteachers familiarize themselves with theways in which the language of academictexts differs from other forms ofEnglish. A number of reference bookscan also help teachers talk about thelanguage that they find in textbooks(Celce-Murcia & Ijirsen-Ereenian. 1998:Huddlestt)n & Pullam, 2001).

Schools should encourage andprovide opportunities for all teachers toparticipate in professional developmentactivities that focus on the role thatlanguage plays in specific grade levelsand in content classrooms (Fillmore &Snow, 2002). Professional developmentmay involve focused workshops on

48 2()O4/J

language development and academicEnglish or foaims of exchange—eitherformal or informal—among teachers indifferent disciplines.

Any professional developmentprogram that targets academic Englishshould encourage teachers to includelanguage objectives alongside contentobjectives in daily lesson plans. Theselanguage objectives remind teachersand students alike to pay attention tofeatures of academic English, such asthose illustrated here. Language objec-tives may focus on academic Englishvocabulary, common academic Englishstructures, or such language fianctionsas explaining, defending, anddiscussing. Highlighting academiclanguage—however briefly—as anobjective in every lesson enhancesstudent awareness of academic Englishand promotes student achievement,

English language learners can makethe grade, even on the toughest tests,

provided they get the instructionalsupport they need. IS

ReferencesBiticr. D,. Conrad, S.. & Rcppcn, R. (1998).

Corpus linguistics: Investigatinglanguage structure and use. (Cambridge,VK: (;:ambridge University Press.

(;eice-Murcia, M,. & Larscn-Ereeman. D,(1998). The grammar book: AnESL/EfL teacher's course. Boston:Heinle,

Chapin. S, H,. & O'Connor, C, (2(X)4).Report Oil Project Challenge: Identi-fying and developing talent in mathe-matics within low income urbanschools. Research Report No. I. fioston:Boston tiniversit>' School of Education.

Chapin. S, H., () Connor, (;., & Anderson,N. C. (2(M)3). Classroom discussions:Using math talk to help studentslearn. Sansalito, CA; Matb Solutions,

Corson, D, (1997). The learning and useof academic English words. LanguageLearning. •J7(4). 6"?1-^IH,

Cummins, J. (1979). Linguistic interdepen-dence and the educational develop-ment of bilingual children. Review of

Educational Research, 49. 221-2'5I.Cummins. J, (1983). Language proficiency

and academic achievement. In J, W,Oiler Jr, (Ed,), Issues in languagetesting research (pp. 108-129),Rowley, MA: Newbury House,

Cummins, J. (2000), Language, poiver,and pedagogy: Bilingual children inthe crossfire. CIcvedon, UK: Multilin-gual Matters,

Fillmore. L, W,, & Snow, C, E, (2002),What teachers need to know aboutlanguage. In C T, Adger, C, E, Snow. &D, Chri.stian (Eds,), What teachers needto know about language. Washington,DC: Center for Applied Linguistics.

Huddleston, R., & Pullani. (i, (2001), TheCambridge grammar of the Englishlanguage. Cambridge, LJK: CambridgeUniversity Press,

Brian Bielenberg is a professional devel-opment consultant at the United ArabEmirates University; bbielenberg@uaeu,ac,ae, Lily Wong Fillmore is a Professor inthe Graduate School of Education at theUniversity of California, Berkeley;wongfill@berkeley,edu.

Research based. Field-tested. Preven. Practical.

These are terms we den't use loosely.Sopris West Educational Services has been creating proven and practical resources for educationprofessionals for more than 25 years. No other company better understands the challenges of meetingthe needs of students, and how critical sound products are to success.

From tools for assessment to ititervention curricula, and more, you can count on us.

Sopris West—While yeu're iooking after your students, we re iookiug aner you.

#SOPRISWESTE D U C A I K J N A L ShRVICES

A Cambium Lcarniiit" Company

(800) 457-6747www.sopriswest.com

A S S O c : l A T I O N F O K S u P B R V I S l O N A M > R K K^i: lU M DEV K t.OPM ENT 49