Closing the gap: Addressing the vocabulary needs of English-language learners in bilingual and mainstream classrooms

Reading Research QuarterlyVol. 39, No. 2

April/May/June 2004© 2004 International Reading Association

(pp. 188–215)

C lo sin g th e g a p : A d d re ss in g th ev o ca b u la r y n e e d s o f E n g lish -la n g u a g e le a rn e rs in b ilin g u a la n d m a in stre a m c la ssro o m sMARÍA S. CARLOUniversity of Miami, Miami, Florida, USA

DIANE AUGUSTAugust & Associates, Washington, DC, USA

BARRY MCLAUGHLINUniversity of California, Santa Cruz, California, USA

CATHERINE E. SNOWCHERYL DRESSLERHarvard Graduate School of Education, Cambridge, Massachusetts, USA

DAVID N. LIPPMANTERESA J. LIVELYUniversity of California, Santa Cruz, California, USA

CLAIRE E. WHITEHarvard Graduate School of Education, Cambridge, Massachusetts, USA

he existence of a large and persistent gap between the reading performance ofAnglo and Latino children on national assessments in the United States (Donahue,Finnegan, Lutkus, Allen, & Campbell, 2001) represents both an intellectual and apractical challenge. Practically, gaining access to the information taught in middleand secondary content area classes requires that all children exit the elementary

188

T

189

GAPS IN reading performance between Anglo and Latino children are associated with gaps in vocabulary knowl-edge. An intervention was designed to enhance fifth graders’ academic vocabulary. The meanings of academicallyuseful words were taught together with strategies for using information from context, from morphology, fromknowledge about multiple meanings, and from cognates to infer word meaning. Among the principles underlyingthe intervention were that new words should be encountered in meaningful text, that native Spanish speakers shouldhave access to the text’s meaning through Spanish, that words should be encountered in varying contexts, and thatword knowledge involves spelling, pronunciation, morphology, and syntax as well as depth of meaning. Fifthgraders in the intervention group showed greater growth than the comparison group on knowledge of the wordstaught, on depth of vocabulary knowledge, on understanding multiple meanings, and on reading comprehension.The intervention effects were as large for the English-language learners (ELLs) as for the English-only speakers(EOs), though the ELLs scored lower on all pre- and posttest measures. The results show the feasibility of improv-ing comprehension outcomes for students in mixed ELL–EO classes, by teaching word analysis and vocabularylearning strategies.

Closing the gap:Addressing thevocabulary needsof English-languagelearners inbilingual andmainstreamclassrooms

LAS DIFERENCIAS en el desempeño lector entre niños angloamericanos y latinos se asocian a brechas en elconocimiento del vocabulario. Se diseñó una intervención para mejorar el vocabulario académico de niños de quin-to grado. Se enseñaron los significados de palabras académicamente útiles, junto con estrategias para usar informa-ción del contexto, de la morfología, del conocimiento de significados múltiples y de palabras emparentadas para in-ferir los significados. Entre los principios subyacentes a la intervención se propuso que las palabras nuevas debíanencontrarse en textos significativos, que los hablantes nativos de español debían tener acceso al significado del textoa través de su lengua, que las palabras debían aparecer en contextos variados y que el conocimiento de las palabrasincluye ortografía, pronunciación, morfología y sintaxis, así como significado en profundidad. Los niños de quintogrado del grupo de intervención mostraron un mayor crecimiento que el grupo de comparación en: conocimiento delas palabras enseñadas, profundidad de conocimiento del vocabulario, comprensión de significados múltiples y com-prensión lectora. Los efectos de la intervención fueron tan grandes para los aprendices de inglés (ELL), como para loshablantes nativos de inglés (EO), aunque el grupo ELL obtuvo puntajes inferiores en todas las medidas pre y posttest. Los resultados muestran la factibilidad de mejorar la comprensión de estudiantes en aulas mixtas ELL–EO,mediante la enseñanza de análisis de las palabras y estrategias de aprendizaje del vocabulario.

Cerrando la brecha: Acerca de lasnecesidades devocabulario deaprendices deinglés en aulasbilingües ycomunes

LÜCKEN BEI der Leseleistung zwischen Anglo- und Latino-Kindern werden mit Wissenlücken im Wortschatz as-soziiert. Ein Einwirken wurde darauf ausgerichtet, den so geschulten Wortschatz bei Schülern der fünften Klassezu verbessern. Die Bedeutungen von theoretisch-akademisch dienlichen Wörtern wurden zusammen mit Strategienüber die Anwendung von Informationen aus dem Textinhalt, der Morphologie, dem Wissen um die vielfältigeSinndeutung, und den Kognaten zum Ableiten der Wortbedeutung unterricht. Unter den der Intervention zu-grundeliegenden Prinzipien galt es, daß neue Wörter innerhalb eines sinnverdeutlichenden Textes in Angriffgenommen werden sollten, so daß die als Muttersprache spanisch Sprechenden Zugang zu Sinn und Bedeutung desTextes in spanisch haben sollten, daß Wörter in variierenden Kontexten angewandt werden sollten und daß dieWortbeherrschung, das Buchstabieren, die Aussprache, Morphologie und Syntax, sowie die tieferen Bedeutungenmit einbezogen sind. Fünftklässler in der Interventionensgruppe zeigten größere Fortschritte als ihreVergleichsgruppe in Kenntnissen der unterrichteten Wörter, in der Gründlichkeit ihrer Vokabelbeherrschung, beimErfassen von Vieldeutigkeiten, und ihrem Leseverständnis. Die Interventionsauswirkungen waren ebenso weitre-ichend bei Anfängern in der englischen Sprache (ELL) wie bei originär-englischen Muttersprachlern (EO), wobeijedoch die ELL-Werte bei allen Vor- und Nachtests geringer ausfielen. Die Resultate zeigten deutlich Möglichkeitenim Verbessern der Ergebnisse für Schüler in gemischten ELL–EO-Klassen auf, sowie beim Unterrichten vonWordanalyse und Vokabel-Lernstrategien.

Schließen der Lücke:Ansprechen desWortschatzbedarfsvon Anfängern in derenglischen Sprachein bi-lingualen undallgemeinenKlassenzimmern

ABSTRACTS

190

LES ÉCARTS de réussite en lecture entre élèves anglophones et hispanophones vont de pair avec des écarts dansleur connaissance du vocabulaire. On a planifié une intervention en vue de développer le vocabulaire scolaired’élèves de cinquième année. On a enseigné le sens de mots utiles dans le cadre scolaire en même temps que desstratégies pour utiliser l’information apportée par le contexte, la morphologie, la pluralité des sens et l’origine afind’inférer le sens des mots. Un des principes sous-tendant l’intervention était que les mots nouveaux devaient êtrerencontrés dans un contexte signifiant, que les enfants de langue espagnole devaient avoir accès au sens du texteen passant par l’espagnol, et que la connaissance d’un mot implique l’orthographe, la prononciation, la mor-phologie et la syntaxe tout comme son sens profond. Les élèves de cinquième année du groupe d’intervention ontréalisé plus de progrès que ceux du groupe témoin en ce qui concerne la connaissance des mots enseignés, la con-naissance profonde du vocabulaire, la compréhension de la pluralité des sens et la compréhension de la lecture.Les effets de l’intervention ont été aussi importants pour les élèves ayant l’anglais pour langue 2 que pour ceux dontc’est la langue maternelle, bien que les résultats des premiers aient été plus faibles à tous les tests, avant et aprèsl’intervention. Les résultats montrent qu’il est possible d’améliorer les résultats en lecture des élèves dans des clas-ses mixtes (anglais langue maternelle ou langue 2), en leur enseignant des stratégies d’analyse des mots et d’apprentissage du vocabulaire.

Réduire l’écart :intervenir sur les

besoins envocabulaire

anglaisd’apprenants declasses bilingues

et de classesd’intégration

ABSTRACTS

grades with good reading comprehension capacity.Without this capacity, access to grade-appropriatecontent knowledge, entry into challenging courses insecondary school, success on the tests increasinglybeing required for promotion and graduation, andentry into tertiary education are all unlikely. Thus,closing this gap has high priority if U.S. education isto fulfill its goal of reducing inequities in access toeconomic opportunities that are contingent uponsuccessful school achievement. Yet as recently as2000 García noted that “few researchers have devel-oped programs to improve students’ second-languagereading vocabulary” (p. 826).

The intellectual challenge posed by the gap in-volves isolating its root cause. The problem is one oftoo many rather than too few likely explanations.Considerable previous work suggests that one majordeterminant of poor reading comprehension, forLatino children (García, 1991; Nagy, 1997;Verhoeven, 1990) and for other lagging readers(National Institute of Child Health and HumanDevelopment, 2000), is low vocabulary. Lack ofknowledge of the middle- and lower frequency “aca-demic” words encountered in middle and secondaryschool texts impedes comprehension of those texts,which in turn impedes the natural process of learn-ing new word meanings from exposure during read-ing (Stanovich, 1986). It is widely known thatvocabulary relates to reading comprehension scores(Freebody & Anderson, 1983), and the presumptionis that the effect is reciprocal—greater vocabularyknowledge makes comprehension easier, while widerreading generates larger vocabularies. One goal ofthe current study was to test whether improvementsin vocabulary related to improvements in readingcomprehension for English-language learners (ELLs).

Review of literature on vocabularyinstruction and learning

Attempts to address the practical challenge ofimproving reading comprehension by explicitlyteaching vocabulary have met with mixed success(Stahl & Fairbanks, 1986), in part because of thedifficulty of generating a large instructional impacton vocabulary knowledge. Successful vocabulary cur-ricula increase children’s word knowledge by approx-imately 300 words a year (Stahl & Fairbanks). Whilesuch gains are not unimportant, they are hardly suf-ficient to close the gap between the vocabulary skillsof lower socioeconomic status (SES) and middle SESchildren, which is estimated to be as high as 6,000words at school entry (Hart & Risley, 1995).

Addressing the vocabulary deficits of second-language learners, who may arrive in U.S. classroomsin second or third grade with no English vocabularyat all, is even more challenging. While such childrenmay appear to acquire oral English vocabularyquickly, they can remain well behind children whohave been exposed to oral and literate English sincebirth, unless provided with skills and strategies forrapid learning of the words they will encounter intheir reading—words that may be used rarely inspoken language.

For avid readers, much vocabulary acquisitionoccurs incidentally as a result of encountering unfa-miliar words while reading (Sternberg, 1987).Nonetheless, the probability of acquiring an un-known word incidentally through reading is onlyabout 15% (Swanborn & de Glopper, 1999), whichmeans the word would need to be encountered eighttimes to be learned with high probability. The prob-ability of learning any word at a first encounter islower for younger readers, for more difficult texts,and probably for students who have had no trainingin deriving meanings for unknown words (Fukkink& de Glopper, 1998; Kuhn & Stahl, 1998). Thus,incidental vocabulary learning is not a reliable proce-dure for promoting vocabulary growth.

Relying on incidental vocabulary learning iseven more problematic for ELLs than for theirEnglish-only (EO) counterparts. ELLs are less ableto use context to disambiguate the meaning of unfa-miliar words because a higher proportion of words intext is likely to be unknown to them. Furthermore,because they lack full command of the Englishgrammar, they are less able to exploit linguistic cuesto word meaning as an EO speaker could (Stoller &Grabe, 1995). Reading texts in which more than 2%of the words are unfamiliar blocks comprehensionand novel word learning (Carver, 1994). As is sug-gested in many of the chapters in Huckin, Haynes,and Coady (1995), which explore the relationshipbetween reading and vocabulary development in asecond language, vocabulary instruction for ELLswould ideally combine direct teaching of words withincidental learning fostered by multiple opportuni-ties to encounter novel words in authentic and moti-vating texts.

In addition to direct instruction and incidentallearning, evidence suggests the desirability of en-hancing the value of incidental exposure by teachingELLs strategies for inferring the meanings of newlyencountered, unfamiliar words. Many such strategiesexist and are used with ease by high-level readers.They include using contextual cues, morphologicalinformation, and cognate knowledge, as well as

Closing the gap 191

using aids such as dictionaries and glossaries (García& Nagy, 1993; Jiménez, García, & Pearson, 1996;Nagy, García, Durgunoglu, & Hancin-Bhatt, 1993;Nation, 2001). Beck, McKeown, and Omanson(1987) demonstrated that it was possible to teachEO children the use of strategies for inferring wordmeaning, while at the same time enriching vocabu-laries with direct teaching. Nagy et al. as well asGarcía and Nagy have presented results suggestingthe efficacy of teaching native Spanish speakers ex-plicitly about the value of cognates and of morpho-logical relationships between Spanish and English.But no one has, to our knowledge, previously testedthe impact of an English vocabulary enrichment in-tervention that combined direct word instructionwith instruction in word-learning strategies on theword knowledge and reading comprehension abili-ties of ELLs. That was our major goal in the studypresented here.

Theoretical framework for designing a vocabulary intervention

What is involved in learning words? The intervention we tested and the measures

we selected to assess its impact reflect our model ofthe complexity of word meaning. “Knowing a word”implies knowing many things about the word—itsliteral meaning; its various connotations; the sorts ofsyntactic constructions into which it enters; the mor-phological options it offers; and a rich array of se-mantic associates, including synonyms, antonyms,hypernyms, hyponyms, and words with closely relat-ed yet contrasting meanings, as well as its capacity forpolysemy (see Bloom, 2002; Nagy & Scott, 2000, forreviews). Learning a word requires learning (over a se-ries of encounters) these various aspects of its mean-ing, and inferring word meaning from context canalso require being alert to these various aspects; a firstencounter with a word might, for example, provideinformation about syntactic word class and some verygeneral specification of meaning domain, whereassubsequent encounters will expand the semantic spec-ification and may lead to discovery of polysemouspossibilities. Thus, subsequent encounters build depthof word knowledge, which is as important in usingwords as is the more commonly assessed breadth.Second-language speakers have been shown to belacking depth of word knowledge, even for frequentlyoccurring words (Verhallen & Schoonen, 1993).Sensitivity on the part of learners to issues of depth

(e.g., recognition that polysemy exists) may ease theprocess of reading comprehension.

How can one best promote word learning? We present here findings from the implemen-

tation of a 15-week intervention focused on teachinguseful words and word-learning strategies. The de-sign of the intervention was related to practicesshown to be effective in previous work (e.g., Beck etal., 1987; Blachowicz & Fisher, 1996; Graves, 2000;Huckin et al., 1995; Nagy, 1988; Nation, 2001;National Institute of Child Health and HumanDevelopment, 2000; Stahl, 1986) that together dic-tated our answers to key questions about whichwords to teach, how often to present them, what as-pects of word knowledge to focus on, and what in-structional activities to use.

(a) Which words? As recommended by Beck et al. (1987), theintervention focused on general-purpose academic wordslikely to be encountered across a variety of content areas,rather than words specific to a particular subject matter. Inaddition, we followed Beck et al.’s method by choosingwords of middle-tier frequency that had depth potentialand morphological and cognate affordances.

(b) How to introduce the words? We chose words from textsdetermined to be appropriate for and of interest to thelearners, building on the evidence that engaging textspromote reading comprehension (Guthrie & Wigfield,2000) and that encountering words in meaningful con-texts promotes memory for them (Nation, 2001). Wealso selected texts that were available in Spanish as well asEnglish, to facilitate engagement for the ELLs and topromote their comprehension of the context in whichthe words were introduced in English. (See list of wordsand activities taught in Appendix A and B and a sum-mary of a week’s worth of sample lessons in AppendixC.) Additional examples are presented in Lively, August,Snow, and Carlo (2003).

(c) How often? Target words should be encountered severaltimes, in diverse contexts, and with varying tasks requiredof learners (Beck et al., 1987). By designing the inter-vention around texts and themes, we created a situationin which target words could be recycled in later lessonsand would appear naturally with high frequency in thetexts being read.

(d) What aspects of word knowledge to focus on? Our instruc-tion focused on depth of meaning, polysemy, morpho-logical structure, and cross-language relationships, as wellas spelling and pronunciation, of the target words. Theinstruction was designed to provide children withgeneral-purpose strategies for acquiring word meaningwhile at the same time teaching specific word meanings.Following Graves (2000) and others, we chose to pro-mote strategic knowledge at some cost to the time avail-able for building vocabulary size, because previous work(e.g., Stahl & Fairbanks, 1986) had shown the futilityof trying to teach large numbers of words directly.

192 Reading Research Quarterly APRIL/ MAY/JUNE 2004 39/2

Closing the gap 193

(e) What instructional techniques? The intervention relied onexplicit instruction in using context to infer word mean-ing (Graves, 2000; Nation 2001), in depth of word mean-ing, in the possibility of polysemy, in performingmorphological analysis, in glossary use, and in cognateuse. (In all glossaries we included the Spanish translationfor the meaning used in the text as well as the English de-finition. We also alerted students to the presence of poly-semous words by including all definitions for the wordand highlighting the definition used in the text.)

Collaborative work in groups that includedboth EOs and ELLs provided practice in English forELLs and made knowledge about cognates availableto EOs. Students were sensitized to the task of vo-cabulary acquisition using techniques like WordWizard (see Beck, McKeown, & Kucan, 2002, fordescription of Word Wizard) in order to enhancetheir utilization of opportunities to extend wordknowledge in nonlesson contexts. Teachers were pro-vided with information about the principles of vo-cabulary acquisition as well as specific strategies toenhance their use of effective vocabulary teachingprocedures outside the vocabulary lessons.

MethodParticipants

The participants were 254 bilingual andmonolingual children from nine fifth-grade class-rooms in four schools in California, Virginia, andMassachusetts. The California site included twoschools serving largely working class MexicanAmerican children, either in bilingual or in main-stream programs. The Massachusetts site servedworking class, mostly Puerto Rican and Dominicanstudents, again in either bilingual or mainstreamclassrooms, within a school where many of theteachers and administrators were bilinguals. TheVirginia site was a magnet, English-medium schoolthat served mainly working class Spanish speakersfrom the Caribbean and from Central America, to-gether with native speakers of many other languagesand middle class English-only speakers attracted byits excellent academic programs. While the variationsacross the three sites created some complexity in im-plementing the intervention and in interpreting thefindings, they also assure us that any intervention ef-fects found are quite robust.

One hundred forty-two of the student partici-pants were ELLs, and 112 were EOs. Ninety-four ofthe ELLs and 75 of EOs were in the intervention

condition. The remaining students (48 ELLs and 37EOs) were in comparison classrooms.

DesignThe study employed a quasi-experimental de-

sign in which classrooms at each site were randomlyassigned to the treatment and comparison condi-tions. This procedure resulted in the assignment of10 classes to the treatment (3 in California, 4 inVirginia, and 3 in Massachusetts) while 6 classroomsserved as comparisons. Students in the comparisonclassrooms did not receive special instruction otherthan that normally included in the school curricu-lum, though their teachers did participate as mem-bers of school teams in professional developmentactivities focused on vocabulary teaching two yearsprior to the introduction of the intervention.

PilotThe intervention activities were piloted in the

same three sites the year before the intervention itselfwas implemented. Over 200 Spanish-speaking ELLsand EO students participated in the pilot along with12 teachers. This pilot was designed to enable us torefine the teaching activities and the student assess-ments. On the basis of the pilot we made decisionsabout complexity, genre, and variety of texts to use,and we decided that teachers implementing the in-tervention would benefit from ongoing professionaldevelopment focused on the intervention.

Testing procedureFifth-grade students in the intervention and

comparison classrooms were tested in the fall and thespring of the academic year on a series of tests de-signed to reflect the skills the curriculum taught(forming deeper representations of word knowledge,understanding polysemy, morphological analysis, in-ferring word meaning from context) as well as read-ing comprehension. Bilingual graduate researchassistants administered all tests (see the Measures sec-tion). Because of various scheduling difficulties, notall students took all tests.

Intervention procedureThe intervention consisted of 15 weeks of in-

struction. Ten to 12 target words were introduced atthe beginning of each week. Instruction was deliv-ered for 30–45 minutes four days a week. Every fifth

week was devoted to review of the previous fourweeks’ target words.

The intervention was organized around thetopic of immigration; the curriculum drew readingsfrom newspaper articles, diaries, firsthand documen-tation of the immigrant experience, as well as histori-cal and fictional accounts, building each week’sfive-day sequence of lessons around a single textunit. The lessons included some homework assign-ments and a weekly review test. The Spanishspeakers were given the text (in both written andaudiotaped versions) to preview in Spanish, onMonday, before its introduction in English onTuesday. On Tuesday, whole-group lessons involvedpresentation of the English text and target words,followed by an activity that involved identifying tar-get words in the text whose meanings could be in-ferred by context. Wednesday lessons involved workin heterogeneous language groups of four to six inwhich the children completed two types of clozetasks with the target words. The first cloze task al-ways involved sentence contexts that were consistentwith the theme of the instructional text. A secondcloze activity involved sentences that employed thetarget words in contexts that were distant in themefrom the instructional text, being designed to helpstudents understand and use related meanings forthe target words, and in the process develop a sensethat most words are polysemous. The Thursday les-son involved activities of the sort recommended byBeck et al. (2002) to promote depth of word knowl-edge (e.g., word association tasks, synonym/antonymtasks, semantic feature analysis). Again children typi-cally completed these activities in small groups. OnFridays the activities varied in focus, as can be seenin Appendix B. During some weeks, Friday activitiespromoted analysis of root words and derivational af-fixes. In other weeks Friday activities were designedto promote awareness of word polysemy and cog-nates. These activities were designed to promoteword analysis capacities in general, not specifically toreinforce learning of the target words.

The curriculum materials provided to theteachers included detailed lesson plans and quasi-scripted lesson guides, as well as overhead trans-parencies, worksheets, homework assignments, andall necessary reading materials. These materials andthe words to be taught were previewed in biweeklyTeacher Learning Community meetings facilitatedby a researcher at each site. At these meetings, prac-tices that had worked well in previous lessons and as-pects of the curriculum that had been problematicwere discussed. These meetings were meant to pro-vide support to the teachers throughout the imple-

mentation of the curriculum and information to theresearchers about aspects of the curriculum that wereworking well or not. The curriculum itself was notmodified as a result of the meetings with the treat-ment teachers.

Fidelity of treatmentThree lessons (one from week 4, one from week

9, and one from week 13 of the curriculum sequence)were filmed in each intervention classroom to obtaindata on the fidelity of implementation of the treat-ment at each site. The tapes were subsequently codedto reflect the degree to which the teacher correctlyimplemented the key elements of the lesson plan; foreach key element, the teacher was rated as havingomitted it, implemented it incompletely or incorrect-ly, implemented it fully, or enhanced it (White,2000). These ratings achieved acceptable levels of reli-ability (Kendall coefficient of concordance = .70).They were summed per lesson per teacher to providea single score indicating fidelity of implementation.Six of the nine teachers implemented over 70% ofkey lesson elements over the three weeklong observa-tions, while one achieved only 35% implementationand the other two 50–60%. The three teachers withthe highest fidelity enhanced the implementationwith additional elements that were consistent with itsdesign, and none of the six high implementers com-mitted any errors of implementation. In fact, thepoorest implementers committed errors in imple-menting only 4% of the elements; omission of ele-ments was their major failing.

MeasuresThe measures used in the study are described

below. Table 1 includes sample items and relevantpsychometric information about each measure.

PPVT–RStudents were tested individually on the L

form (pretest) and M form (posttest) of the PeabodyPicture Vocabulary Test Revised (PPVT–R).Children who had been tested on PPVT–R in thespring of fourth grade were not retested until springof fifth grade. Students new to the study were testedin both fall and spring.

Polysemy productionThe students’ task on the polysemy production

test was to generate as many sentences as possible


Closing the gap 195

conveying the different meanings of the polysemouswords. The words in the fifth-grade polysemy taskwere ring, place, settle, pitch, back, and check. The stu-dents’ correct responses were awarded points on thebasis of the frequency of the response in the responsepool. Common responses were awarded one point(e.g., “Don’t bug me” or “She left a ring in the tub”).Two points were awarded to responses in the inter-mediate range of frequency (e.g., “My mom drives aVolkswagen bug” or “The phone has a funny ring”).Three points were awarded to the correct but mostinfrequent responses in the pool (e.g., “There was abug in the program” or “The boxer left the ring”).

Reading comprehensionReading comprehension was assessed with

multiple-choice cloze passages with content wordsdeleted at random. Students read three stories withsix cloze items per story. Ten of the deleted wordswere taught in the intervention.

Word masteryThe mastery test was designed to determine if

the vocabulary words directly taught were successful-ly learned. It consisted of 36 target word multiple-choice items. Each target word was followed by fourshort definitions. Students were asked to select thedefinition that best described the word.

Word Association taskThis task, developed originally by Schoonen and

Verhallen (1998), measured depth of word knowledgeby tapping into children’s knowledge of paradigmaticand syntagmatic word relations. The task consisted of

20 target words, half of which were included in thecurriculum. Each of the target words appeared in thecenter of a page, with six other words printed aroundthe periphery. Students were asked to draw lines fromthe target word to the three peripheral words mostclosely connected to it. Specifically, students wereasked to pick three of the words that “always go withthe word in the middle.” For example, the word de-bate has immutable associations to the words rival,discussion, and opinion but only circumstantial associa-tions to the words president, television, and fight.

MorphologyChildren’s knowledge of English morphology

was assessed with a paper and pencil adaptation ofCarlisle’s (1988) Extract-the-Base task. Our task con-sisted of 27 items (fewer than a third were interven-tion words) that required that the student providethe base form of a derived word. Students heard thederived word (e.g., discussion) followed by a lean sen-tence context (e.g., What did he want to _____?)and were asked to provide the word that fit in thesentence (i.e., discuss). The items varied in the trans-parency of relationship between the derived and baseforms. Some items involved no phonological or or-thographic change (e.g., remark-remarkable); someinvolved a phonological change but no orthographicchange (e.g., nation-national ); some involved an or-thographic change but no phonological change (e.g.,furious-fury); and some involved both changes (e.g.,migration-migrate). The children’s spelling of the basewords often did not conform to spelling conventionsof English. The following coding scheme informedby Bear, Invernizzi, Templeton, and Johnston’s(1996) spelling rubrics was used to score the stu-dents’ responses.

WordMastery Association Polysemy Cloze Morphology

Total items 36 20 6 18 27

Possible range 0–36 0–54 0–54 0–18 0–135

Coefficient alpha .83 .94 .64 .73 .94

Sample items

TABLE 1PROPERTIES OF FIFTH-GRADE ASSESSMENTS

Rigid:(a) soft and smooth,(b) approved by some-one in authority, (c)valuable, (d) stiff anddifficult to bend

Policy:Uniform, decisions,plan, action, insurance,congress

Write a sentence foreach meaning youknow for pitch.

With time things gotbetter and many set-tlers became ____.(a) anxious, (b) sick, (c)open, (d) prosperous

Election: How manypeople did they ____?

5 points: Correct base, correctly spelled: guide

4 points: Correct morpheme boundaries and reasonablespelling with all phonemes represented: assist,emty, cantinew

3 points: Correct morpheme boundaries, more deviantspelling with all phonemes represented: viori,oqupie

2 points: Derived form, but base word preserved/visibleand more or less correctly spelled: elected, assisted

1 point: Some of the target sounds represented, but basi-cally the wrong word: sends for sense

0 points: Completely incorrect. This category includes re-producing the stimulus word (discussion) or pro-viding words that are morphologically unrelatedto the stimulus word (hire for employ)

ResultsEffect of the intervention

A multivariate analysis of variance performedon the six dependent measures for which scores wereavailable (Task: Mastery, Word Association,Polysemy, Cloze, Morphology) in both fall andspring (Time), and incorporating as predictor vari-ables site, language status, and condition, revealedoverall between-subjects effects attributable to siteand to language status. Tests of within-subjects ef-fects showed significant gains over time, and a signif-icant interaction between gain over time andcondition, as well as a three-way interaction betweengain over time, site, and condition (see Table 2).

These results confirmed the justifiability ofpursuing analysis with each of the outcome variablesindividually (see Table 3) following normal practice(Myers & Well, 1991).

While a conservative approach to the follow-upanalyses would correct for Type I error rate per familyof contrast, we chose to report uncorrected effectsgiven the unavailability of data about vocabulary in-tervention effects in this population. The readershould note that effects for which the probability lev-el is larger than .01 would not meet the more strin-gent standard based on the Bonferroni adjustment.

PPVT–R, a measure designed to reflect indi-vidual differences but to be insensitive to curricularinfluences, was significantly higher for the EO stu-dents and showed significant gains from fall tospring. It also differed significantly across the threesites, with the highest scores in Virginia and the low-est in Massachusetts. It did not, however, reveal anyinfluence of treatment; the treatment group by gainsinteraction term was not significant, nor were anyother interactions with gains detected.

We analyzed the remaining five measures,Mastery, Polysemy, Word Association, Cloze, andMorphology, covarying PPVT–R to reduce effects as-sociated with differences in initial English proficiencyand with site differences in populations being served.These five measures showed a gradient of effects.Mastery, Word Association, Polysemy, and Cloze allshowed the same general pattern of results demon-strating impact of the intervention: The interventiongroup showed greater gain in the course of the schoolyear than the comparison group. Mastery, WordAssociation, and Polysemy also showed an interactionof gain over time with site, reflecting either differen-tial effectiveness of the implementation across thevarious sites or differential susceptibility to the im-pact in the various sites (see Figures 1 through 5).

Morphology showed only marginally signifi-cantly differential gains as a function of condition orsite. Morphology, Mastery, and Word Association allshowed a three-way interaction, indicating largergains for the intervention than the comparisongroup in some sites (see Figures 4a and 4b). Theseinteractions with site could reflect either populationor implementation differences.

Main effects of language group were found forWord Association and Polysemy, reflecting the gen-erally higher performance of native English speakers.Main effects for site were found for Morphology,Mastery, and Word Association. The site differencesmay reflect the distinctive demographics of the threesites, differing recruitment policies for the schoolsinvolved, or quality of education delivered.


Effect F df MSE p

Time of test 104.99 1,82 47.39 ***

Time 3 condition 7.92 1,82 47.39 **

Time 3 condition 3 school 3.85 2,82 47.39 **

Task 434.53 1,82 351.15 ***

Task 3 language 30.94 1,82 351.15 ***

Task 3 school 4.03 3,82 351.15 *

Task 3 language 3 school 3.76 3,82 351.15 *

Time 3 task 44.55 1,82 37.35 ***

School 4.65 3,82 727.07 **

Language 56.74 1,82 727.07 ***

School 3 language 3.89 3,82 727.07 *

*p < .05; **p < .01 ; ***p < .001.

TABLE 2RESULTS OF MANOVA ON FIFTH-GRADEOUTCOMES

Closing the gap 197

WordMastery Association Polysemy Cloze Morphology

TABLE 3RESULTS OF ANOVAS ON EACH OUTCOME MEASURE ADMINISTERED IN FIFTH GRADE

Time of test Fdf

MSEh2

**7.64(1,218)

9.09.03

ns ns ns **11.46(1,217)94.83

.05

Time by PPVT–RFdf

MSEh2

ns *4.71(1,217)24.82

.02

ns *4.73(1,213)

3.72.02

ns

Time 3 condition Fdf

MSEh2

***113.28(1,218)

9.09.34

**11.24(1,217)24.82

.05

**11.23(1,212)

3.23.05

***17.84(1,213)

3.71.08

ns

Time 3 schoolFdf

MSEh2

886.38(2,218)

9.09.06

***5.19(2,217)24.82

.05

*3.80(2,212)

3.23.04

ns ns

Time 3 condition 3 school F

dfMSE

h2

***11.46(2,218)

9.09.10

*3.48(2,217)24.82

.03

ns ns ***9.37(2,217)94.84

.08

Time 3 language 3 school F

dfMSE

h2

ns ns ns ns *4.44(2,217)94.84

.04

PPVT–RFdf

MSEh2

***198.49(1,218)24.87

.48

***78.62(1,217)37.85

.27

***49.42(1,212)11.49.19

***137.12(1,213)

8.80.39

***119.37(1,217)849.53

.36

ConditionFdf

MSEh2

***52.22(1,218)24.87

.19

ns ns *4.91(1,213)

8.80.02

ns

LanguageFdf

MSEh2

ns **10.24(1,217)37.85

.05

***13.05(1,212)11.49.06

ns ns

SchoolFdf

MSEh2

***8.59(2,218)24.87

.07

ns ns ***12.56(2,213)

8.80.11

**7.01(2,217)849.53

.06

Language 3 conditionFdf

MSEh2

ns ns **7.36(1,212)11.49.03

ns ns

School 3 conditionFdf

MSEh2

**5.83(2,218)24.87

.05

**5.05(2,217)37.85

.04

ns ns **4.90(2,217)849.53

.04

*p < .05; **p < .01; ***p < .001.


FIGURE 1AAVERAGE ELL FIFTH-GRADE FALL AND SPRING PERFORMANCE ON THE MASTERY TASKAS A FUNCTION OF SITE AND CONDITION

Num

ber

corr

ect

30

25

20

15

10

5

0

Fall SpringTime of test

Site 1 Treatment ELL Site 2 Treatment ELL Site 3 Treatment ELLSite 1 Comparison ELL Site 2 Comparison ELL Site 3 Comparison ELL

FIGURE 1BAVERAGE EO FIFTH-GRADE FALL AND SPRING PERFORMANCE ON THE MASTERY TASKAS A FUNCTION OF SITE AND CONDITION

Num

ber

corr

ect

35

30

25

20

15

10

5

0


Site 1 Treatment EO Site 2 Treatment EO Site 3 Treatment EOSite 1 Comparison EO Site 2 Comparison EO Site 3 Comparison EO

Closing the gap 199

FIGURE 2AAVERAGE ELL FIFTH-GRADE FALL AND SPRING PERFORMANCE ON THE CLOZE TASK AS A FUNCTION OF SITE AND CONDITION

Num

ber

corr

ect

14

12

10

8

6

4

2

0


FIGURE 2BAVERAGE EO FIFTH-GRADE FALL AND SPRING PERFORMANCE ON THE CLOZE TASK AS A FUNCTION OF SITE AND CONDITION

Num

ber

corr

ect

18

16

14

12

10

8

6

4

2

0





FIGURE 3AAVERAGE ELL FIFTH-GRADE FALL AND SPRING PERFORMANCE ON THE WORDASSOCIATION TASK AS A FUNCTION OF SITE AND CONDITION

Num

ber

corr

ect

45

40

35

30

25

20

15

10

5

0


FIGURE 3BAVERAGE EO FIFTH-GRADE FALL AND SPRING PERFORMANCE ON THE WORDASSOCIATION TASK AS A FUNCTION OF SITE AND CONDITION

Num

ber

corr

ect

60

50

40

30

20

10

0




Closing the gap 201

FIGURE 4AAVERAGE ELL FIFTH-GRADE FALL AND SPRING PERFORMANCE ON THE MORPHOLOGYTASK AS A FUNCTION OF SITE AND CONDITION

Num

ber

corr

ect

90

80

70

60

50

40

30

20

10

0


FIGURE 4BAVERAGE EO FIFTH-GRADE FALL AND SPRING PERFORMANCE ON THE MORPHOLOGYTASK AS A FUNCTION OF SITE AND CONDITION

Num

ber

corr

ect

140

120

100

80

60

40

20

0





FIGURE 5AAVERAGE ELL FIFTH-GRADE FALL AND SPRING PERFORMANCE ON THE POLYSEMY TASKAS A FUNCTION OF SITE AND CONDITION

Num

ber

corr

ect

14

12

10

8

6

4

2

0


FIGURE 5BAVERAGE EO FIFTH-GRADE FALL AND SPRING PERFORMANCE ON THE POLYSEMY TASKAS A FUNCTION OF SITE AND CONDITION

Num

ber

corr

ect

16

14

12

10

8

6

4

2

0




Closing the gap 203

The fact that these language group and site dif-ferences emerged despite controls for PPVT–R sug-gests that differences across the sites in level ofperformance and in impact of the intervention canbe explained only partly by differences in demo-graphic characteristics among the children served.The striking conclusion must be that the effective-ness of the vocabulary instruction was quite resistantto disruption from other influences.

Fidelity of implementationWhy were there persistent differences among

the sites in so many outcomes? The three sites dif-fered, as noted above, in their demographics, in theorganization of schooling for ELLs, and in aspects ofteacher functioning. Informal observations suggesteddifferences within and across the schools in the en-thusiasm with which the intervention was embraced,in the reflectiveness and thoroughness with which itwas implemented, as well as in the quality of instruc-tion that was occurring during the rest of the schoolday. To see whether fidelity and quality of imple-mentation were accounting for site differences,fidelity ratings were introduced into regressionanalyses, using spring scores on each of the fiveintervention-sensitive measures as outcomes; fallscore on the same measure, language group, andPPVT–R were introduced first to see if fidelity rat-ings accounted for differences in spring scores. Fornone of the measures did fidelity significantly in-crease the amount of variation explained, nor weredifferences between children in the classrooms ofhigh- versus low-fidelity implementers significant.Thus, though there was considerable variation acrosssite and across teacher in fidelity, those differences donot help explain gains.

DiscussionThe major goal of this study was to test the im-

pact of an English vocabulary enrichment interven-tion that combined direct word instruction withinstruction in word-learning strategies on outcomesfor ELLs. In addition, we wanted to test the impactof that same curriculum on EOs in the same class-rooms. We found that a challenging curriculum thatfocused on teaching academic words, awareness ofpolysemy, strategies for inferring word meaning fromcontext, and tools for analyzing morphological andcross-linguistic aspects of word meaning did improvethe performance of both ELL and EO fifth graders,to equal degrees. The children in the intervention

classrooms gained knowledge of the words that wereexplicitly taught as well as knowledge that shouldsupport the efficiency of their incidental learning ofnovel words (i.e., vocabulary depth as well as knowl-edge about morphological structure, about cognates,and about polysemy). Thus, we confirmed the effec-tiveness for ELLs of approaches to vocabulary teach-ing such as that pioneered by Beck et al. (1987) thathad previously been shown effective only with EOs;we demonstrated that procedures previously showneffective only with ELLs, such as those pioneered byNagy et al. (1993), could be used in mixed-languagegroups in mainstream classrooms; and we developeda single curricular approach that incorporated manyspecific techniques and strategies, such as thosedeveloped by Graves (2000), Nagy (1988), Stahl(1986), and others, previously shown to be effectivein supporting vocabulary learning.

A second goal of this study was to see whetherimproved vocabulary and word analysis skills wouldbe associated with improved reading comprehensionoutcomes. We found that, for both ELLs and EOs,the intervention was effective in improving readingcomprehension outcomes. The effects for readingcomprehension were less dramatic than for wordknowledge, as shown by the effect size of .08 (etasquared) on the reading comprehension measure ascompared with .34 for mastery of the words taught.Nonetheless, it is striking that any improvement inreading comprehension was measurable after a rela-tively brief curricular intervention that did not focusspecifically on teaching comprehension.

Thus, we conclude that direct vocabulary in-struction is effective, with both ELL and EO learn-ers, if it incorporates the various principles gleanedfrom previous work on monolingual English speak-ers and ELLs (Beck et al., 1987, 2002; Graves, 2000;Nagy, 1997; National Institute of Child Health andHuman Development, 2000; Stahl & Fairbanks,1986). For example, teachers should introduce novelwords in the context of engaging texts, design manyactivities such as Charades that allow learners to ma-nipulate and analyze word meaning, heighten atten-tion to words in general with techniques like WordWizard, ensure that learners write and spell the tar-get words several times, ensure repeated exposures tothe novel words, and help children note how theword meaning varies as a function of context. Wealso conclude that teaching children strategies for in-ferring the meaning of unknown words is effective,with both ELL and EO learners, if it builds on well-verified procedures (García & Nagy, 1993; Graves;Jiménez et al., 1996; Nagy et al.,1993) such as teach-

ing explicitly how to use context cues, teaching mor-phological analysis, and teaching about cognates.

Challenges to curricular changeAnother conclusion to be drawn from this

work is the complexity of implementing a challeng-ing curriculum in urban schools. The teachers in ourstudy were volunteers, and most were experiencedclassroom practitioners. They were, however, work-ing in schools that were adopting a variety of curric-ular initiatives, which in some cases diverted timeand attention from this particular intervention.Furthermore, though we made every attempt to pro-vide curricular materials that were self-explanatoryand easy to use, those materials presupposed knowl-edge about vocabulary and about lexical analysis thatthe teachers in some cases did not possess. Despiteefforts to design the curriculum so it could be usedwith little additional preparation time from theteachers, we found it necessary to provide consider-able professional support, and even so, differences inquality of implementation were measurable.

Additional measures of successThe intervention was, overall, successful in its

specific aim of enhancing reading skills and wordknowledge. It was successful by other measures aswell. The teachers reported quite a high level of satis-faction with the materials, and the students enjoyedthe lessons and displayed heightened sensitivity toword meanings and increased awareness of Spanish–English relationships (see Dressler, 2000). It is disap-pointing, on the other hand, that the interventionwas not observed to change classroom practice out-side the structured lessons and has not survived as anintact instructional program in any of the classroomswhere it was introduced, though some teachersreport continuing to use some of the techniques theylearned during the intervention.

The challenge of optimizing learning forELL and EO children simultaneously

The intervention assessed here incorporated anumber of specific instructional components (e.g.,the specific texts used, the small-group activities, thecrossword puzzles and other homework assignments)that we believe contributed to its success, though wecould not assess their independent impact. Some ofthese components were designed to support the par-ticipation of Spanish speakers: providing the key

texts in Spanish so they could be previewed, provid-ing translation equivalents of the target words in theteaching materials, selection of immigration as theorganizing topic, inclusion among the texts to beread of several that were particularly relevant to theexperiences of Mexican and Dominican immigrantfamilies, and incorporating instructional activitiesthat relied on the Spanish speakers as resources (e.g.,in identifying cognates). Our analyses cannot help usdecide whether these supports for Spanish speakers’participation contributed importantly to the effec-tiveness of intervention for them, or indeed whetherit contributed to or detracted from its effectivenesswith their English-only classmates. Future researchmight well explore the impact of these various com-ponents independently.

LimitationsThis study was, of course, subject to many

limitations. Several of the measures we used wereexperimenter designed, though their psychometricproperties were satisfactory. We did not employ astandardized measure of reading comprehension; werelied on cloze procedures because they have oftenbeen used to study the interaction between compre-hension and word knowledge. However, concernshave been raised in the research literature about thevalue of cloze tests as measures of comprehension(Shanahan, Kamil, & Tobin, 1982). Testing the ef-fects of the intervention on comprehension perfor-mance using a wider variety of reading measureswould be valuable. Because we had no general meas-ure of English-language proficiency for the ELL stu-dents, we were unable to test the interaction betweenEnglish proficiency and intervention effects. Giventhe enormous variability in English proficiencyamong ELLs, it would be important to determinewhether the effects of the intervention vary as afunction of English proficiency.

In addition, we were unable to test long-termeffects of the intervention, either on students or onteachers. A particular issue to think about in evaluat-ing a curriculum-based intervention like this is thetrade-off between the value of a predesigned chunkof curriculum with some built-in professional devel-opment versus a more significant investment in pro-fessional development itself. Giving teachers moreaccess to information about and practice with thevocabulary teaching techniques incorporated intothe intervention might have enabled them to usethose techniques more broadly across a variety ofsubject areas, possibly with greater impact than the45-minute vocabulary-focused lessons we designed.


Closing the gap 205

ConclusionThe key distinction of the vocabulary training

offered here was that teaching new words was subor-dinated to the goal of teaching about words—various kinds of information about words that couldhelp children figure out word meanings on theirown. Thus, the curriculum introduced only 12 to 14new words a week, sacrificing the opportunity toteach an additional 10 to 15 words in order to focusinstruction on strategies for narrowing candidateword meanings by using context, noticing words innew contexts, checking the likelihood that the wordhas a Spanish cognate, and analyzing morphologicalstructure for cues to meaning. Such strategies couldhave ongoing value to children who encounter un-known words in semantically rich contexts, who un-derstand enough of the context to use contextualinformation in analyzing word meaning, and whoremember to use them. Their value at least in theshort run was, in fact, confirmed by our finding of asignificant impact on reading comprehension.

MARÍA S. CARLO is an assistant professor in the Teaching andLearning Program at the School of Education of the University ofMiami. She is a psychologist studying bilingualism in children andadults. Her research focuses on the cognitive processes that underliereading in a second language and on understanding the differences inthe reading processes of bilinguals and monolinguals. She has a Ph.D.in psychology from the University of Massachusetts at Amherst. Shecan be contacted at the University of Miami, PO Box 248256, CoralGables, FL 33124-3310, USA, or by e-mail at [email protected].

DIANE AUGUST is a senior research scientist at the Center forApplied Linguistics. Currently she is the principal investigator of aNational Institute for Child Health and Human Development programproject grant focused on the development of literacy in English-language learners; she is also staff director for a Department ofEducation National Literacy Panel on Language Minority Children andYouth. Previously she was a senior program officer at the NationalResearch Council and study director for the Committee on Developinga Research Agenda on the Education of Limited English Proficient andBilingual Students. August has also worked as a public school teacherand school administrator in California, a legislative assistant in thearea of education for a U.S. Congressman from California, a grantsofficer for the Carnegie Corporation of New York, director of educationfor Children’s Defense Fund, and an educational consultant inevaluation and testing, program improvement, and federal and stateeducation policy. She has a Ph.D. in education from StanfordUniversity. She can be contacted by e-mail at [email protected].

BARRY MCLAUGHLIN is professor emeritus in the Department ofPsychology at the University of California, Santa Cruz. He was formerlydirector of the National Center for Research on Cultural Diversity andSecond Language Learning. His research interests are in the areas ofsecond-language acquisition and bilingual development. He can becontacted at Psychology Department, Program in CognitivePsychology, University of California, Santa Cruz, CA 95064, USA, or bye-mail at [email protected].

CATHERINE E. SNOW is the Henry Lee Shattuck professor at theHarvard Graduate School of Education. She chaired the RAND ReadingStudy Group that produced the report Reading for Understanding andis currently involved with others in thinking about implications ofresearch on reading for improvement of teacher education andprofessional development. She can be contacted at Harvard GraduateSchool of Education, Larsen 313, Cambridge, MA 02138, USA, or bye-mail at [email protected].

CHERYL DRESSLER is an assistant professor at the Graduate Collegeof Education, University of Massachusetts, Boston. She teachescourses on literacy education and has a particular interest in theliteracy development of English-language learners. She can becontacted by e-mail at [email protected].

DAVID N. LIPPMAN has worked in education and psychologyresearch since 1995. He is currently working for Human Developmentand attending graduate school in Oakland, California, USA. He may becontacted at the Institute of Human Development, 1211 Tolman Hall,University of California, Berkeley, CA 94720-1690, USA, or by e-mailat [email protected].

TERESA J. LIVELY enjoyed a 14-year career as a bilingual teacher.She first became aware of the struggles that accompany confrontationwith a new language when she lived in Mexico for two years as achild. This childhood experience, in combination with recognition thatstudents’ academic success is greatly influenced by their ability tounderstand and use words, prompted her to emphasize acquisition ofvocabulary in her classroom. Lively is currently completing apostdoctoral internship in clinical psychology. She may be contacted at940 E. Meadow Drive, Palo Alto, CA 94303, USA, or by e-mail [email protected].

CLAIRE E. WHITE is a doctoral student in the Language and Literacyprogram at the Harvard Graduate School of Education. She can becontacted at Massachusetts Department of Education, 350 Main Street,Malden, MA 02148, USA, or by e-mail at [email protected].

R E F E R E N C E SBEAR, R., INVERNIZZI, M., TEMPLETON, S.R., & JOHNSTON,

F. (1996). Words their way: Word study for phonics, vocabulary, and spelling in-struction. Upper Saddle River, NJ: Prentice Hall.

BECK, I., MCKEOWN, M.G., & KUCAN, L. (2002). Bringingwords to life: Robust vocabulary instruction. New York: Guilford.

BECK, I., MCKEOWN, M.G., & OMANSON, R.C. (1987). Theeffects and uses of diverse vocabulary instructional techniques. In M.G.McKeown & M.E. Curtis (Eds.), The nature of vocabulary acquisition(pp. 147–163). Hillsdale, NJ: Erlbaum.

BLACHOWICZ, C., & FISHER, P. (1996). Teaching vocabulary inall classrooms. Columbus, OH: Prentice Hall.

BLOOM, P. (2002). How children learn the meaning of words.Cambridge, MA: MIT Press.

CARLISLE, J.F. (1988). Knowledge of derivational morphology andspelling ability in fourth, sixth and eighth graders. Applied Psycholinguistics,9, 247–266.

CARVER, R.P. (1994). Percentage of unknown vocabulary words intext as a function of the relative difficulty of the text: Implications for in-struction. Journal of Reading Behavior, 26, 413–437.

DONAHUE, P.L., FINNEGAN, R.J., LUTKUS, A.D., ALLEN,N.L., & CAMPBELL, J.R. (2001). The nation’s report card: Fourth-gradereading 2000. Washington, DC: National Center for Education Statistics.

DRESSLER, C. (2000). The word-inferencing strategies of bilingual andmonolingual fifth graders: A case study approach. Unpublished qualifyingpaper, Harvard Graduate School of Education, Cambridge, MA.

FREEBODY, P., & ANDERSON, R. (1983). Effects of vocabularydifficulty, text cohesion, and schema availability on reading comprehen-

sion. Reading Research Quarterly, 18, 277–294.FUKKINK, R.G., & DE GLOPPER, K. (1998). Effects of instruc-

tion in deriving word meaning from context: A meta-analysis. Review ofEducational Research, 68, 450–468.

GARCÍA, G.E. (1991). Factors influencing the English reading testperformance of Spanish-speaking Hispanic students. Reading ResearchQuarterly, 26, 371–392.

GARCÍA, G.E. (2000). Bilingual children’s reading. In M.L. Kamil,P.B. Mosenthal, P.D. Pearson, & R. Barr (Eds.), Handbook of reading re-search (Vol. III, pp. 813–834). Mahwah, NJ: Erlbaum.

GARCÍA, G.E., & NAGY, W.E. (1993). Latino students’ conceptof cognates. In D.J. Leu & C.K. Kinzer (Eds.), Examining central issuesin literacy research, theory, and practice. Chicago: National ReadingConference.

GRAVES, M.F. (2000). A vocabulary program to complement andbolster a middle grade comprehension program. In B.M. Taylor, M.F.Graves, & P. van den Broek (Eds.), Reading for meaning: Fostering com-prehension in the middle grades (pp. 116–135). New York: TeachersCollege Press.

GUTHRIE, J., & WIGFIELD, A. (2000). Engagement and motiva-tion in reading. In M.L. Kamil, P.B. Mosenthal, P.D. Pearson, & R. Barr(Eds.), Handbook of reading research (Volume III, pp. 403–422). Mahwah,NJ: Erlbaum.

HART, B., & RISLEY, T.R. (1995). Meaningful differences in theeveryday experience of young American children. Baltimore: Paul H.Brookes.

HUCKIN, T., HAYNES, M., & COADY, J. (1995). Second languagereading and vocabulary learning. Norwood, NJ: Ablex.

JIMÉNEZ, R., GARCÍA, G.E., & PEARSON, P.D. (1996). Thereading strategies of bilingual Latina/o students who are successful Englishreaders: Opportunities and obstacles. Reading Research Quarterly, 31,90–112.

KUHN, M.R., & STAHL, S.A. (1998). Teaching children to learnword meanings from context: A synthesis and some questions. Journal ofLiteracy Research, 30(1), 19–38.

LIVELY, T., AUGUST, D., SNOW, C.E., & CARLO, M.S. (2003).Vocabulary improvement program for English language learners and theirclassmates. Baltimore: Paul H. Brookes.

MYERS, J., & WELL, A. (1991). Research design and statistical analy-sis. New York: HarperCollins.

NAGY, W.E. (1988). Teaching vocabulary to improve reading compre-hension. Newark, DE: International Reading Association.

NAGY, W.E. (1997). On the role of context in first- and second-language vocabulary learning. In N. Schmitt & M. McCarthy (Eds.),Vocabulary: Description, acquisition and pedagogy (pp. 64–83). Cambridge,UK: Cambridge University Press.

NAGY, W.E., GARCIA, G.E., DURGUNOGLU, A., & HANCIN-BHATT, B. (1993). Spanish-English bilingual students’ use of cognates inEnglish reading. Journal of Reading Behavior, 25, 241–259.

NAGY, W.E., & SCOTT, J.A. (2000). Vocabulary processes. In M.L.Kamil, P.B. Mosenthal, P.D. Pearson, & R. Barr (Eds.), Handbook ofreading research (Vol. III, pp. 269–284). Mahwah, NJ: Erlbaum.

NATION, I.S.P. (2001). Learning vocabulary in another language.Cambridge, UK: Cambridge University Press.

NATIONAL INSTITUTE OF CHILD HEALTH AND HUMANDEVELOPMENT. (2000). The report of the National Reading Panel.

Teaching children to read: An evidence-based assessment of the scientific lit-erature on reading and its implications for reading instruction. Washington,DC: U.S. Government Printing Office.

SCHOONEN, R., & VERHALLEN, M. (1998). Kennis van woor-den: De toetsing van diepe woordkennis (Knowledge of words: Testingdeep word knowledge). Pedagogische Studiën, 75, 153–168.

SHANAHAN, T., KAMIL, M.L., & TOBIN, A.W. (1982). Clozeas a measure of intersentential comprehension. Reading Research Quarterly,17, 229–255.

STAHL, S.A. (1986). Three principles of effective vocabulary in-struction. Journal of Reading, 29, 662–668.

STAHL, S.A., & FAIRBANKS, M. (1986). The effects of vocabu-lary instruction: A model-based meta-analysis. Review of EducationalResearch, 56, 72–110.

STANOVICH, K.E. (1986). Matthew effects in reading: Some con-sequences of individual differences in the acquisition of literacy. ReadingResearch Quarterly, 21, 360–407.

STERNBERG, R. (1987). Most vocabulary is learned from context.In M.G. McKeown & M.E. Curtis (Eds.), The nature of vocabulary ac-quisition (pp. 89–105). Hillsdale, NJ: Erlbaum.

STOLLER, F., & GRABE, W. (1995). Implications for L2 vocabu-lary acquisition and instruction from L1 vocabulary research. In T.Huckin, M. Haynes, & J. Coady (Eds.), Second language reading and vo-cabulary learning (pp. 24–45). Norwood, NJ: Ablex.

SWANBORN, M.S.L., & DE GLOPPER, K. (1999). Incidentalword learning while reading: A meta-analysis. Review of EducationalResearch, 69, 261–286.

VERHALLEN, M., & SCHOONEN, R. (1993). Vocabulary knowl-edge of monolingual and bilingual children. Applied Linguistics, 14,344–363.

VERHOEVEN, L.T. (1990). Acquisition of reading in a second lan-guage. Reading Research Quarterly, 25, 90–114.

WHITE, C.E. (2000). Implementation of a vocabulary curriculum de-signed for second-language-learners: Knowledge bases and strategies used bymonolingual and bilingual teachers. Unpublished qualifying paper, HarvardGraduate School of Education, Cambridge, MA.

Received October 3, 2001Final revision received September 5, 2003

Accepted September 19, 2003

A U T H O R S ’ N O T ETeresa Lively, Diane August, María Carlo, and Catherine Snow have

a significant financial interest in and affiliation with a product that hasbeen developed and published as a result of the research described in thisarticle.

E D I T O R S ’ N O T EThis manuscript was accepted prior to the finalization of the RRQ

guidelines pertaining to conflict of interest (see Editorial in RRQ, Vol.39, Issue 1).


207

TARGET WORDS (NONCOGNATES ARE IN ITALICS) TAUGHT DURING INTERVENTION BY WEEK OF INSTRUCTION

APPENDIX A

Week Title of reading material Source of selection Target words

(1)

(2)

(3)

(4)

(5)

(6)

(7)

A journey to the newworld—Part 1

Lasky, K. (1996). DearAmerica: A journey tothe new world. Thediary of RememberPatience Whipple. NewYork: Scholastic.

Ambition, catastrophe,determination, dictator,discriminate, diverse,epidemic, famine, flee,immigrant, motive,optimism, prospect,settle


Same as above Authority, condemn,corrupt, establish,faction, native,persecution, pilgrim,political, puritan,reformer, reign, scorn,worship


Same as above Accumulate, charter,coalition, colony,economy, essentially,financial, grueling,idealist, indenture,livelihood, merchant,profit, sponsor

Immigrant kids Freedman, R. (1980).Immigrant kids. NewYork: Puffin.

Ally, contact,determine, document,draft, exert, fledgling,jurisdiction, official,ominous, ravage,surplus, trade, treaty

Review week No new words

Immigrant kids athome—Part 1

Same as above Absorb, century,congested, custom,dense, dialect, district,ethnic, fervent,impoverished, occupy,prosperous, social,tenement

Immigrant kids athome—Part 2

Same as above Arouse, common,congregate, dank,elevated, humanity,monotonous, pitchedbattle, relief, rival,stifling, torment,ultimatum, unfamiliar

(continued)

TARGET WORDS (NONCOGNATES ARE IN ITALICS) TAUGHT DURING INTERVENTION BY WEEK OF INSTRUCTION (continued)

APPENDIX A

Week Title of reading material Source of selection Target words

(8)

(9)

(10)

(11)

(12)

(13)

(14)

(15)

Immigrant kids atschool—Part 1

Same as above Allegiance, facility,humiliate, laborious,meticulous, monitor,nonexistent,penetrating, pledge,promote, represent,rigid, script, strive

Immigrant kids atschool—Part 2

Same as above Agency, anxiety,fundamental, heritage,obtain, periodic,reflect, reject, shame,stenography, tradition,transform, values,vocational


New kids in town Bode, J. (1989). Newkids in town: Oralhistories of immigrantteens. New York:Scholastic.

Amend, collective,debate, demographics,extend, impression,inaccurate, issue, midst,opportunity, resident,shift, stem,unprecedented

The new immigranttide: A shuttle betweenworlds

The New York Times.(1998, July 19–21).

Campaign, civic,concentration,contemplate, degree,dual, ebullient,forsaken, fracture, inutero, renown, shuttle,straddle, transnational

A Mexican town thattranscends all borders

Same as above Assimilate, bestride,communal, flourish,hybrid, identity,juncture, novel, psyche,redefine, saga, span,transcontinental,transcend

The new immigranttide—Part II

Same as above Balk, conscious,hyperdeveloped,immediate, inevitable,maintain, overwhelming,profound, revolutionize,status, technology,tentative, underclass, vital


208

209

SUMMARY OF VOCABULARY INSTRUCTION ACTIVITIES BY DAY AND WEEK OF INTERVENTION

APPENDIX B

Week Day 1 Day 2 Day 3 Day 4 Day 5

(1) Preview forELLs;studentslisten toSpanishsummary ofreadingpassage andpreview listof targetwords

Introduction;predict storyline; readpassage;circlevocabulary;extractdefinitions;assignhomework

Using wordsin context;completeclozesentencesworking ingroups

Expandingmeaning:Word roots

Tools todevelopvocabulary:Cognates

(2) Same asabove

Same asabove

Same asabove

Expandingmeaning:Deepprocessing

Tools todevelopvocabulary:Affixes

(3) Same asabove

Same asabove

Same asabove


Tools todevelopvocabulary:Idioms

(4) Same asabove

Same asabove

Same asabove

Expandingmeaning:Multiplemeanings

Tools todevelopvocabulary:Root words

(5) Word bee Wordmeaninganalysis

Charades Word guess Posttest

(6) Preview forELLs

Introduction Using wordsin context

Expandingmeaning:Antonyms/synonyms

Tools todevelopvocabulary:Inferencing

(7) Same asabove

Same asabove

Same asabove



(8) Same asabove

Same asabove

Same asabove

Expandingmeaning:Wordsubstitution

Tools todevelopvocabulary:Affixes

(continued)

210

SUMMARY OF VOCABULARY INSTRUCTION ACTIVITIES BY DAY AND WEEK OF INTERVENTION (continued)

APPENDIX B

Week Day 1 Day 2 Day 3 Day 4 Day 5

(9) Same asabove

Same asabove

Same asabove

Expandingmeaning:Relatedwords


(10) Homework Polysemy Charades Word sort Posttest

(11) Preview forELLs

Introduction Using wordsin context

Expandingmeaning:Synonyms/antonyms

Tools todevelopvocabulary:Dictionaries

(12) Same asabove

Same asabove

Same asabove

Expandingmeaning:Synonyms/antonyms


(13) Same asabove

Same asabove

Same asabove

Expandingmeaning:Wordsubstitution


(14) Same asabove

Same asabove

Same asabove


Tools todevelopvocabulary:Multiplemeanings

(15) Homework Word guess Charades Word bee Posttest

211

OVERVIEW OF WEEK 7 ACTIVITIES: IMMIGRANT KIDS AT HOME

APPENDIX C

Lesson Instructions to teacher Instructions to student in teacher handbook in student workbook

q Instruct ELL students to go tostations where they will listen toaudiotaped recordings of the Spanishversion of the English text to be readthe next day.

q Students will also be given briefdefinitions for Spanish translations ofthe English target words.

q Listen to audiotape.Day 1: Preview forELLs

q Engage students in a brief storyprediction activity using illustrationsand review of prior week’s story.

q Read the passage aloud to thestudents.

q Read posted target words.q Reread the passage aloud to the

students.q Call on students who appear to know

the meaning of the target word.q Guide students in discussion of those

words whose meaning can be inferredfrom the context.

q Guide students in discussion ofmeaning of compound words in thepassage and add to the compoundword wall.

q Assign homework: definitions.

q Three- to five-minutestudent discussion ofpredictions

q Listen to passage and followalong in your workbook.

q Read posted target words.q As teacher rereads the

passage circle each of thetarget words as they come upin the text. Raise your handif you know the meaning ofthe target word withouthaving to look it up in yourglossary for this week.

q For homework write thecorrect target vocabularyword next to the definitionprovided.

Day 2: Introductionof text and vocabularyinferringstrategies

(continued)

212

OVERVIEW OF WEEK 7 ACTIVITIES: IMMIGRANT KIDS AT HOME(continued)

APPENDIX C

Lesson Instructions to teacher Instructions to student in teacher handbook in student workbook

q Correct your homeworkresponses if necessary.

q As a group discuss each ofthe items in the Contextingactivity in your workbook.Write the answer in yourworkbook.

q Complete the crosswordpuzzle on your own.

Day 3: Using wordsin context

q Review crossword puzzle homeworkusing the homework transparencythat is provided. Write the correctresponses on the transparency.

q Group students in heterogeneouslanguage groups of four or fivestudents.

q Guide students through DeepProcessing activity using thetransparency that is provided (seeexample below).

q Correct your homeworkresponses if necessary.

q As a group discuss each ofthe items in the DeepProcessing activity in yourworkbook. Write theanswer in your workbook.

Day 4: Expandingmeaning

q Review homework using thehomework transparency that isprovided. Write the correct responseson the transparency.

q Group students in heterogeneouslanguage groups of four or fivestudents.

q Show students the transparency forthe Contexting activity (see examplebelow). Have students discuss theanswers and raise their hands whenevery member of the group knows theanswer and agrees with the group’sanswer. Assign points to groups forcorrect answers.

q Ask one member of the group toprovide the answer and to explainwhy it is the correct one.

q Read aloud three or four sentencesfrom previous week’s Word Wizardactivity. Send the sentences to theproject webmaster so that they maybe posted on the project website.

q Assign homework: crossword puzzle,Word Wizard List for posting at home.

(continued)

213


APPENDIX C

Vocabulary inferring strategies (example of directions to teacher)q Before you reread the passage from the transparency, say, Follow along in your Student

Word Book as I read the passage again. Give me a “thumbs up” when I read a targetword and then circle the word. If you think you know what the word means, withoutreading the definition, raise your hand and I will call on you.

q After students have suggested meanings, say, Let’s check the definition to see how closeyou were.

q Meaning can be inferred for humanity. When you reach humanity, say, Rememberthat sometimes you can figure out what a word means by skipping over it and finishingthe sentence. Or you can reread the sentence while thinking about what the word mightmean. Let me remind you how this works by reading the sentence with humanity in it.“The sunlight and fresh air of our mountain home...were replaced by four walls andpeople over and under and on all sides of us until it seemed that humanity from allcorners of the world had congregated in this corner of New York City.” Let’s see, earlier inthe sentence it talks about people on all sides of them. Do you think humanity hassomething to do with groups of people? Let’s look up the definition to see if we’re close.

Word Wizard activity (example of directions to teacher)q Encourage each group of students to read 3–4 sentences from the previous week.

Choose one from each group to be posted on the Web. q Motivate students to find sentences with this week’s vocabulary. Say, Each time you

hear or read one of this week’s words used in a sentence at home, school, or even on TV, Iwant you to write the sentence on a sentence strip. Please write down where you heard orread it.

q Group students in heterogeneouslanguage groups of four or five studentsfor the Cognates activity. Be aware thatELL students will be the experts in thistask. Instruct them to help English-speaking group members complete theactivity.

q Guide students through Cognatesactivity using the transparency that isprovided (see example below).

q Instruct students to complete theVocabulary Review activity individually.

q As a group discuss eachof the items in theCognates activity in yourworkbook. Write theanswer in yourworkbook.

q Complete the VocabularyReview by writing thecorrect target vocabularyword next to thedefinition provided.

Day 5: Tools todevelopvocabularyandvocabularyreview

(continued)

214


APPENDIX C

Contexting activity (example of directions to teacher)q Using the transparency (in Teacher’s Materials), read the first cloze sentence aloud to

the class.q Say, Your job is to figure out which word fits in the blank using the clues that are in the

sentence. When everyone in the group knows the correct word and why it fits, raise yourhands. I’ll call on one of the first groups ready. You will get a point if you get the correctanswer. Remember, everyone in your group must know the answer and why it is correct.

q Ask one child at a table for the correct answer and to explain why it is correct.q Find ways to encourage English-language learners to participate.q Continue until the lesson is completed, giving each group 1 point for each correct

answer.q Note that first set of 14 cloze sentences consists of “near contexts.” These are

sentence contexts that are similar in topic to the target passage (e.g., Manyimmigrants from the same country gather together for religious holidays. They oftencongregate at the local church.) The second set of 14 cloze sentences are “farcontexts” indicating the use of the word in a different thematic context (e.g., Atsome parties, the guests like to congregate in a person’s kitchen).

Deep processing activity (example of directions to teacher)q Say, You have already learned many of the definitions for this week’s vocabulary words.

Remember that definitions alone don’t teach you everything you need to know to reallyunderstand what a word means. In today’s activity you will be asked to think about howone word’s meaning relates to another word’s meaning. Does anyone remember what thisis called? That’s right, it is deep processing.

q Say, For example, remember when I asked you to think about a reformer. Which of thesethings would a reformer be likely to do? (a) Go to the park for a picnic; (b) work tochange rules that are not fair to a group of people; (c) notice that something isn’t fair andsay, “Oh well, I can’t do anything.”

q Say, When you think about each of these possibilities, ask yourself whether a reformerwould or would not be likely to do each thing. This makes you think more aboutreformers and gives you a richer and deeper picture of what reformer means.

q Say, Now you will work in your groups with questions similar to the one we just talkedabout. Your job is to read each question and talk with your group to decide which answeris correct and why. Remember, you must be able to explain why you chose your answer.

q Examples of items: Would you rather congregate with your friends in a dank placeor a stifling place? Explain why. What has caused the most torment in your life?What gave you relief from the torment? Name 3 things that commonly arouse ateacher’s anger.

(continued)

215


APPENDIX C

Cognate activity (example of directions to teacher)q Say, Remember that when you are learning another language, it sometimes helps to look

for similarities in how words sound or are spelled. If you were in a country orneighborhood where all the signs were in Spanish, which word would you look for if youwanted to find a police station? Teléfono, policia, or parque? Who remembers whatwords are called that have similar spellings in English and Spanish and are related inmeaning? Remember, there are also false cognates. False cognates are words that havesimilar spellings in English and Spanish, but are NOT related in meaning. Does anyoneremember a false cognate? What does the Spanish word red mean? Red means net inSpanish!

q If this appears to be too difficult for your students, you may need to work as a wholegroup and look for the cognates on the text transparency. You might emphasize thecognates as you read.

q Say, For this activity I will give each team a passage to read. Your job is to look for thewords that have Spanish cognates. When you find a word that you think is a cognate,write the word and the Spanish cognate on the worksheet. Discuss the meanings of theSpanish cognates you find to make sure that they do have the same meaning as theEnglish word in the fable.

q When all teams have completed their worksheets, collect the worksheets and writethe words and their cognates on the board.

q Ask students for meaning of both words to decide if they are true or false cognates.q Prompt for additional clues if students have not found them all.

Documents

Closing the gap: Addressing the vocabulary needs of English-language learners in bilingual and mainstream classrooms