Word Meanings Matter: Cultivating | English Vocabulary Knowledge in Fifth- Grade Spanish-Speaking Language | Minority Learners |

Word Meanings Matter: CultivatingEnglish Vocabulary Knowledge in Grade Spanish-Speaking Language Fifth-Minority Learners JEANNETTE MANCILLA-MARTINEZ University of Illinois at Chicago Chi

Word Meanings Matter: Cultivating

English Vocabulary Knowledge in Grade Spanish-Speaking Language

Fifth-Minority Learners

JEANNETTE MANCILLA-MARTINEZ

University of Illinois at Chicago

Chicago, Illinois, United States

This pilot study investigated the effects of a 20-week quasiexperimentalvocabulary intervention aimed at improving Spanish-speaking languageminority students’ English vocabulary and writing outcomes.Participants were two matched samples of fifth graders (N 5 49) in apredominantly Latino, low-income urban school Pre- and posttestanalyses revealed that the treatment group gained knowledge of alarger number of target words than did the contrast group and that thetreatment group students were generally better at determining theirown word knowledge Further, individual growth modeling revealedthe treatment students’ overall writing quality improved over the course

of the 20-week intervention, even though writing instruction was notpart of the intervention, and improvements in students’ writing qualitywere larger during the last 10 weeks of the intervention The need forpurposeful activities that provide students with authentic contexts tolearn and productively use newly taught words is discussed

doi: 10.5054/tq.2010.213782

T here are approximately 98 million ethnic minority group members(or 33% of the country’s population) in the United States, withLatinos comprising the largest and fastest growing segment of thispopulation (U.S Census Bureau, 2005) Over the past 30 years thenumber of school-age children who spoke a language other than English

at home, known as language minority (LM) learners, nearly tripled(3.8 million to 10.6 million; U.S Department of Education, NationalCenter for Education Statistics, NCES, 2007a), with Spanish being themost common home language These statistics have clear and immediateimplications for U.S schools, as the number of Latino children isexpected to continue to increase in the coming decades (Harwood,Leyendecker, Carlson, Asencio, & Miller, 2002) Of concern is that thisrapidly growing population of Spanish-speaking LM learners accounts

for a disproportionate percentage of struggling comprehenders (August

& Shanahan, 2006; NCES, 2007b) Given that Latino families continue

to be overrepresented among America’s poor (Suarez-Orozco & Paez,2002) and that low-income status is a well-known risk factor associatedwith poor academic outcomes (Hart & Risley, 1995), a considerableproportion of Spanish-speaking Latino students is doubly at-risk forschool failure It is beyond the scope of this article to attempt todisentangle the effect of LM status from low socioeconomic status onstudent academic outcomes, but it is clear that these variables tend to beconfounded, creating compounding obstacles to successful academicoutcomes for the growing population of low-income LM students Manyfactors are associated with LM students’ reading comprehensiondifficulties, but low levels of vocabulary knowledge continue to emerge

as key impediments to successful comprehension (Garcia, 1991; Nagy,1997; Stahl & Nagy, 2006; U.S Department of Health and HumanServices, National Institutes of Health, National Institute of Child Healthand Human Development, NICHD, 2000; Verhoeven, 1990) Despite thestrong and well-established link between vocabulary and comprehension(Anderson & Freebody, 1981; Anderson & Nagy, 1991; U.S Department

of Education, Office of Educational Research and Improvement, RANDReading Study Group, 2002), over the past 20 years relatively fewresearchers have conducted experimental studies on the effectiveness ofvocabulary interventions with LM students (August & Shanahan, 2006;Calderon, August, Slavin, Duran, Madden, & Cheung, 2005) Thepresent study strengthens this research domain by assessing theeffectiveness of a pilot vocabulary intervention designed to improvefifth-grade (ages 10–11 years) LM students’ literacy outcomes

LITERATURE REVIEW

Notwithstanding the poor academic outcomes of many LM learners atall grade levels, reading research has tended to focus on the readingdevelopment of young monolingual English speakers (e.g., NICHD,2000; Snow, Burns, & Griffin, 1998), and that which has been conductedwith LM learners has similarly focused on young children (for a review,see Lesaux, Koda, Siegel, & Shanahan, 2006) Thus research andpractice have focused more on word reading skills than vocabulary andcomprehension skills However, converging evidence finds that LMlearners tend to develop relatively strong word reading skills, but oftenwithout the necessary language skills to support comprehension (e.g.,Droop & Verhoeven, 2003; Hoover & Gough, 1990; Hutchinson,Whiteley, Smith, & Connors, 2003; Proctor, Carlo, August, & Snow,2005; Proctor, August, Carlo, & Snow, 2006; Verhoeven, 1990, 2000)

Thus a distinction between word reading and knowledge of themeanings of the words being read is especially important in under-standing LM learners’ literacy development It is estimated that studentsneed to know (not simply recognize and/or decode) at least 95% of thewords they encounter in text for successful comprehension (Calderon

et al., 2005; Lipson & Wixson, 2003) Estimates of words learned during

a typical school year range from 1,000 (Goulden, Nation, & Read, 1990)

to 3,000 (Nagy & Herman, 1987); although clearly most of these wordsare learned incidentally, explicit vocabulary instruction has also beenfound to contribute to word learning (e.g., Biemiller, 2003).Considering the early vocabulary knowledge disadvantage that many

LM learners face, a sole reliance on incidental vocabulary learning forthis group of learners is both impractical and negligent However, onlyfive vocabulary intervention studies have focused on LM learners(Calderon et al., 2005; Carlo et al., 2004; Perez, 1981; Ramirez, 1986;Vaughn-Shavuo, 1990), and only one (Carlo et al., 2004) has targetedupper-elementary students

Vaughn-Shavuo’s (1990) doctoral dissertation investigated the effect

of vocabulary instruction by randomly assigning two groups of first-gradeSpanish-dominant children to two groups Over the course of 3 weeks,

31 words were presented to both groups Group one (the experimentalgroup) received vocabulary instruction focused on elaborated meanings.Specifically, they learned the target words in meaningful paragraphs(i.e., the sentences containing the target words formed narratives), theywere provided with picture cards of the target words that illustrated theirmeanings, and they dictated their own sentences using the target words

In contrast, the other group (the control group) received instructionfocused on individual sentence contexts (i.e., the sentences containingthe target words were unconnected) Results showed that the experi-mental group learned more words than the control group In anotherstudy, Perez’s (1981) work with third graders revealed that theexperimental group receiving 20 min of daily oral language instruction

on word meanings, over the course of about 3 months, showedsignificant improvements on word learning over the group receivingregular instruction (i.e., reading text and answering questions) Thethird vocabulary study (Ramirez, 1986) also focused on third-gradeSpanish speakers In this study, a method called suggestopedia (a languagelearning method that uses music to create an atmosphere conducive tolearning) was used Ten words were taught per day in 40-min lessonsover the course of 4 days, and the experimental groups performedsignificantly better than the control Calderon and colleagues (2005)likewise investigated Spanish-speaking third graders’ word learning, butthe major goal of their 22- to 25-week intervention (an adaptation of theSuccess for All reading program) was on facilitating students’ Spanish-to-

English transition from a Spanish reading program However, anadditional goal of the intervention was on building English vocabulary.Thus, in the 90-min lessons, vocabulary was a major focus, as 30 min weredevoted to oral language activities revolving around grade-levelchildren’s literature Their results revealed modest positive effects onstudents’ English vocabulary Finally, and most recently, Carlo andcolleagues (2004) conducted a 15-week vocabulary intervention withfifth-grade Spanish-speaking LM learners They found that, while theimpact of their 30- to 45-min a day intervention was greater onvocabulary than on reading comprehension, there were significantimprovements in both domains.

The scope of vocabulary intervention work with LM learners is sparse,most notably beyond the primary grades, but findings to date point tothe promising role of vocabulary instruction to improve LM learners’vocabulary knowledge Further, the strong and significant correlationbetween vocabulary and reading comprehension among LM learners(Droop & Verhoeven, 2003; Hoover & Gough, 1990; Hutchinson et al.,2003; Proctor et al., 2005, 2006; Verhoeven, 1990, 2000) suggests that

LM learners can benefit from targeted vocabulary instruction Further,even though vocabulary might be expected to also impact writing, whichplays an increasingly prominent role in evaluating students’ academicperformance after the primary grades, a key limitation in the field is thatnone of the vocabulary intervention studies conducted to date haveexamined writing outcomes

Some work with third- and fourth-grade monolingual English ers suggests that limited vocabulary knowledge contributes to depen-dence on repetitive uses of the same words and thus to underelaboration

speak-of thoughts and ideas in writing (Moats, Foorman, & Taylor, 2006).Additionally, Saddler and Graham (2007) note that writers’ familiaritywith the writing topic is related to writing performance, suggesting thatbackground knowledge plays a central role in students’ writing quality.Because LM learners tend to have more limited vocabularies andbackground knowledge (two highly interrelated areas) than theirmonolingual English-speaking counterparts, we can expect LM students’writing to be greatly impeded Indeed, the 2002 National Assessment ofEducation Progress writing assessment revealed that only about one-quarter of 4th, 8th, and 12th graders performed at or above the proficientlevel in writing (NCES, 2004), but even more troubling is the fact thatsubstantial differences emerged when examining the data by ethnicity:

At all grade levels, on average, Whites and Asians scored above the 50%percentile, while Blacks and Latinos scored near the 25% percentile

It seems reasonable to postulate that vocabulary instruction mightresult in improved writing outcomes, but there is a surprisingly limitedresearch base on the effect of vocabulary instruction on students’

writing, even amongst native-English-speaking populations (Duin &Graves, 1987; Graham & Perin, 2007) It is thus difficult to draw evententative conclusions about the impact of vocabulary instruction onstudents’ writing and particularly difficult to ascertain the nature of thisrelationship for Spanish-speaking LM learners To my knowledge, onlyfour studies have directly investigated the writing development ofSpanish-speaking LM learners (Bermudez & Prater, 1994; Davis, Carlisle,

& Beeman, 1999; Ferris & Politzer, 1981; Lanauze & Snow, 1989).However, none of these studies examined the potential relationshipbetween vocabulary and writing development Students need vocabulary

to write, but the only evidence available on the role of vocabulary insecond language writing comes from English-as-a-foreign-language(EFL) college students The educational context of LM learners differs

in substantive ways from that of older EFL students, greatly limiting thegeneralizations that can be drawn from EFL work to the LM learnerschool-age population Notwithstanding, EFL studies with collegestudents indeed find that a key determinant in nonnative Englishspeakers’ overall writing quality is vocabulary (e.g., Leki & Carson, 1994;Raimes, 1985; Walters & Wolf, 1996) The lack of empirical researchinvestigating this potential relationship amongst the large and growingpopulation of school-age LM learners is a major limitation in the literacyresearch field

Because receptive vocabulary knowledge generally precedes tive vocabulary knowledge (Laufer, 1998; Meara, 1996; Nation, 1990;Pearson, Hiebert, & Kamil, 2007), it can be expected that students, and

produc-in particular LM learners, will require ample opportunities to activelyuse newly taught vocabulary in oral language before they are able to usethe new words in their writing To date, very few studies have examinedstudents’ use of newly taught vocabulary in writing (Bravo & Tilson,2006; Lee, 2003; Lee & Muncie, 2006) As part of a larger studyexamining the effects of an integrated science-literacy curriculum, Bravoand Tilson (2006) analyzed second- and third-grade students’ use ofscience vocabulary in writing, finding that students spontaneously usednewly taught science words in their writing The authors suggest that use

of newly taught vocabulary in writing represents growth in scienceknowledge and also indicates that students have productive control overscience vocabulary However, an analysis of whether students’ overallwriting quality showed improvements was not conducted, and themajority of students in this study were White native English speakers,limiting the generalizations that can be made to LM learners

The other two studies investigated productive vocabulary use inwriting among secondary school multinative language intermediateEnglish-as-a-second-language students in Canada Specifically, Lee(2003) investigated correct usage of target vocabulary in the writing of

65 students, and Lee and Muncie (2006) investigated learners’ (N 5 48)use of target vocabulary and how their target vocabulary use influencedtheir lexical frequency profile (see Laufer & Nation, 1995, for details).Like Bravo and Tilson (2006), Lee (2003) did not investigate the effects

of vocabulary encountered in reading instruction on students’ overallwriting quality Lee and Muncie considered the relationship betweenvocabulary encountered in reading and learners’ use of the vocabulary

in writing, including improvements in writing quality, finding thatteacher elicitation, explicit explanation, discussion and negotiation, andexposure to target vocabulary increase students’ productive vocabularyuse in writing However, like the other two studies, student income levelswere not reported, effectively limiting our understanding of whetherthese findings can be expected to be replicated with low-incomepopulations

Despite the limited number and limited scope of studies exploringthe relationship between vocabulary instruction and writing, there isreason to believe that vocabulary instruction may indeed be a step in theright direction to improve the writing skills of LM learners Drawing onresearch to date, two reasonable hypotheses are that (1) LM learners willuse newly taught words in their writing if given opportunities to do so on

a consistent basis, and (2) vocabulary instruction will strengthen LMlearners’ overall writing quality over time To adequately explore theeffects of vocabulary instruction on students’ literacy outcomes,research-based vocabulary instruction components and strategies must

be attended to

Components of Effective Vocabulary Instruction

The NICHD (2000) concluded that there is no single best based method for vocabulary instruction, noting that a variety ofmethods are needed, including incidental and structured instruction.Although indirect instruction is vital to any program aiming to developstudents’ vocabulary, direct, carefully designed instruction is also anintegral part of the puzzle The goal of explicit instruction is for students

research-to learn the meanings of words across various contexts research-to ultimatelyimprove their literacy outcomes Researchers agree that the specificwords to be taught should be guided by their potential to aid students’understanding of text and/or concepts and that words students arelikely to encounter relatively frequently should be targeted (e.g., Beck,McKeown, & Kucan, 2002; Biemiller & Slonim, 2001; Nation, 2001) Inother words, words to be taught should be functional, cross-disciplinary,and developmentally appropriate (see Beck et al., 2002, for one widelyused system for selecting words)

Considering the more limited vocabulary levels of many LM learners,vocabulary instruction for LM learners must target students’ languageskills not only more intensively, but also more broadly Because exposure

to academic language is largely confined to the regular school day for

LM learners, LM learners arguably need more opportunities, bothincidental and structured, to hear and use academic language thannative English speakers Additionally, effective vocabulary programs for

LM learners should specifically target the development of wordconsciousness and of word learning strategies to help LM learners catch

up with monolingual English speakers Further, the target vocabulary forinstruction should be presented in meaningful, engaging contexts thatare not only relevant to students’ interests (e.g., Carlo et al., 2004) butthat also serve to bolster their overall background knowledge Finally, anarea that is seldom attended to is ensuring that students are providedwith opportunities to actively use newly taught vocabulary, both orallyand in writing Writing becomes a major form of evaluation after theprimary grades for all students and, on average, over one-quarter ofLatino students write at the below basic level (NCES, 2003) It may be that,

by giving LM learners ample opportunities to write on a consistent basis

as part of a vocabulary program, their understanding of the words theyare being taught will be strengthened That is, the sheer use of newlytaught words in writing may foster learning of the words

PRESENT STUDY

Given that LM students tend to have both less breadth and less depth

of vocabulary knowledge and knowing that vocabulary is strongly related

to students’ overall school success, LM students with limited vocabulariesare very disadvantaged academically The vocabulary program piloted inthis study draws from the research base on effective vocabularyinstruction and rests on the premise that vocabulary instruction canreasonably be expected to improve fifth-grade (ages 10–11 years) LMlearners’ literacy outcomes Fifth grade is an optimal school year tointervene, before students enter the more academically demandingmiddle school grades when struggling comprehenders rapidly fallfurther behind in all content areas The study addressed the followingresearch questions:

(1) Following the 20-week vocabulary intervention, do the treatment orcontrast group students gain knowledge of a greater number oftargeted words?

(2) In the treatment group’s weekly student essays, what is the extent oftarget vocabulary word use over the course of the 20-weekintervention? Specifically, do students use the target vocabulary

words cumulatively, or is their use confined to words taught in thecurrent week? Further, is there any change in the quality (e.g.,coherence and academic language use) of the essays over time?

LM (Spanish), with 46% designated limited English proficient and 8%designated special education The primary concern of the Mysticteachers centered on students’ low levels of vocabulary knowledge andpoor reading comprehension outcomes

Design and Participants

This quasiexperimental pilot study employed a matched-controldesign As a pilot study exploring the preliminary effects of a vocabularyintervention, piloting of the intervention was limited to one grade level(fifth grade) in the same predominantly Latino, low-income, urban K-8school in the Northeast to carefully monitor the implementation of theintervention There were only three fifth-grade classrooms in this school;two were mainstream English classrooms, and the other was a self-contained classroom for recent immigrants and thus did not participate.One classroom served as the treatment group and the other as thecontrast group

All 53 students in the two mainstream classrooms participated, buttwo students from the treatment and two students from the contrastclassroom transferred to other schools during the intervention; thus thefinal sample for statistical analyses consisted of 24 students from thetreatment and 25 from the contrast classroom Except for the number of

students formerly identified as limited English proficient, with more inthe treatment classroom,1 there were no significant differences ingender, race, and first language characteristics across the classrooms(see Table 1) The treatment group received the 20-week intervention,starting in January 2007 and ending in May 2007, while the contrastgroup continued with the regular, district-wide literacy instruction.Regular literacy instruction in this school district centered on at least

80 min of the reading and writing workshop model in which a

‘‘balanced’’ approach to literacy instruction is followed

Intervention Overview and Implementation

Word Generation (WG), developed through a SERP collaborativeeffort under the leadership of Dr Catherine Snow, is a research-based20-week vocabulary intervention designed to build students’ academicvocabulary across the content areas Academic vocabulary refers to wordsthat students are likely to encounter in textbooks and on tests (e.g., inferand element), but not in spoken language Without explicit instruction onthese types of words, students, and especially LM learners, are likely toexperience difficulty with comprehension The goal of WG is to increasestudents’ academic vocabulary, in an effort to improve literacy outcomes.The following components are emphasized: (1) building vocabularyknowledge through repeated exposure to frequently occurring academicwords in various contexts, (2) cultivating general word and world

1 Students at the Mystic School are randomly assigned to classrooms and thus the greater number of limited-English-proficient students in the treatment group compared with the contrast group happened by chance.

TABLE 1

Background Characteristics of Treatment and Contrast Group Students (N 5 49)

Treatment (n 5 24) Contrast (n 5 25) x 2

p value Gender

Formerly limited English proficient

knowledge, as well as word study strategies, and (3) engaging students inweekly persuasive writing The core program centers on the weeklypresentation of five high-utility target words to be learned in the context

of brief passages outlining controversies currently under debate in theUnited States (ranging from the abilities of women in math and science

to global warming).2 The 100 target words are therefore relevant to arange of settings and subject areas and were selected from the AcademicWord List (Coxhead, 2000), which was originally developed as a supportfor instruction to college-level nonnative English speakers and does notinclude words that are in the most frequent 2,000 words of English Themain criteria in selecting the 100 target words were that they be high-utility, high-functional, and cross-disciplinary (see Appendix A for thefull list of WG words) The cross-content focus on a small number ofwords each week aims to enable students to understand the variety ofways in which words are related, and the multiple exposures to words indifferent contexts (e.g., math and history) seek to provide students withample opportunities for deeper understanding of the words The WGmaterials include a teacher’s guide that explains the structure of thevocabulary program and rationale behind it; a set of 20 engagingparagraphs written about current topics in journalistic style, whichconnect to real word issues and to students’ lives; brief instructionalactivities associated with weekly topics and target words; and references

to support teachers in implementing WG activities (see http://www.wordgeneration.org/index.html for more detailed information about

WG, including access to the Teacher’s Guide and a sample weeklylesson)

Another central component of every WG lesson to build students’academic vocabulary is classroom talk Aside from improving students’vocabulary knowledge, the promotion of classroom discussion and talkalso aims to support the development of students’ reasoning and theirability to express their reasoning Thus the following are key features ofthe WG intervention: revoicing by the teacher (i.e., repeating a student’sutterance with the purpose of checking back with them for clearerinterpretation of their statement or position), student repetition (i.e.,having other students repeat or paraphrase another student’s position inorder to check on their interpretation of the statement), asking students

to debate (i.e., giving students opportunities to agree and disagree andhaving them state and make clear their reasoning), and partner talk(i.e., giving students who are less inclined to join whole groupdiscussions the opportunity to talk with a partner to ensure that allstudents are on the same page) Finally, the end-of-week writing activity

2 The WG paragraphs are written at a 6th grade readability level, because the WG materials were specifically developed for use with middle school students (grades 6–8).

is essential to WG, because the goal is to make writing an integral part ofthe vocabulary program, such that students have the opportunity tocompose a short piece based on the controversial topic they havediscussed all week This provides yet another means of allowing students

to express their thoughts and opinions For LM learners, in particular,writing serves as a nonthreatening way to express their views.Additionally, the incorporation of a weekly writing component aims toprovide students the opportunity to explore the use of newly taughtvocabulary in their writing, which is critical to cultivate deeper learning

On Monday, the teacher began the weekly WG lesson by introducingthe controversial topic of the week For example, the teacher would tellthe class that they would be discussing their thoughts on whether rapmusic has a negative impact on kids, and she would explicitly tell herstudents that they would be learning five words, which appeared in boldtype in the passage Before reading the weekly passage with the class,the teacher showed the five weekly target words on large index cards tothe class, one at a time, asking if they knew what the words meant; thiswas a way for the teacher to gauge her students’ knowledge of the fivetarget words Most of the time, the students were asked to complete aword chart for homework, during which they were asked to do one ormore of the following: provide the part of speech and meanings of thewords, provide inflectional forms of the words, identify prefixes andsuffixes, and list any related words On Tuesday, Wednesday, andThursday, the teacher selected from an array of math, science, andsocial studies related activities For math, there was typically a problem

of the day, in which at least two of the five target words wereincorporated into a word problem that resembled the type of wordproblem students would likely encounter on state standardized tests.The inclusion of the math problem of the day was not on teachingmath skills per se Rather, the focus was on exposing students to theweekly target vocabulary in a math context to the extent possible Thescience activities tended to consist of cloze paragraphs, in which many

of the five target words needed to be filled in to complete the sentencesdealing with a science topic Finally, the most typical social studiesactivity was the weekly whole class debate During the debate, studentswere asked to take a stance on the controversial topic they had been

learning about and to present their view to the group holding anopposing view When time permitted, students were allowed to rebutthe opposing group’s argument Finally, on Friday, students wererequired to ‘‘take a stand’’ by writing a short persuasive essay about thecontroversial topic they had been discussing Students were encouraged

to use the five weekly target WG words, as well as previously learnedtarget words, in their essays However, this was not required Onaverage, students tended to finish writing their persuasive essay inabout 10 min Importantly, the teacher did not provide any writinginstruction related to the WG intervention or give students feedback onthe weekly essays they produced

The treatment group teacher completed a weekly feedback form toreport on the implementation of the intervention (see Appendix B) Shecompleted the form each of the 20 weeks Her responses indicated thatshe had implemented WG every week Additionally, 38 observations inthe treatment classroom and 12 in the contrast classroom wereconducted, to ensure that the treatment teacher was implementing

WG as expected and that the contrast teacher was continuing with theregular instruction and not implementing WG or some other instruc-tional program targeting academic language specifically The directclassroom observations revealed that, on average, the teacher imple-mented WG for about 20 min a day In accordance with the teacherreports on the weekly form, the direct observations confirmed that theintervention was being implemented faithfully, even during the periods

of district-mandated testing

Measures

Students in the treatment and contrast groups were administeredpretests in the fall of 2006 and posttests in the spring of 2007 Pretestmeasures included researcher-developed and standardized (for match-ing purposes) group and individually administered tests of languagecomprehension, word reading, reading comprehension, and writing.Posttest measures included the researcher-developed tests only Allmeasures were administered during the school day

Researcher-Developed Measures

Two measures of students’ knowledge of the target vocabulary wordswere administered The multiple-choice (MC) test and the vocabularyself-check (VSC) were both group administered The MC test wascomprised of a randomly selected sample of 30 of the 100 target WGwords Following standard format, students silently read a phrase orshort sentence in which the target word was printed in bold type, and

they selected the appropriate meaning of the target word from a list offour choices For example, students read the following sentence: Thebuilding collapsed after the earthquake After reading the sentence, studentswere required to choose the word or group of words that mean the same

or almost the same as the underlined word In this example, theirchoices were: (a) exploded, (b) fell apart, (c) stayed standing, and (d)collated Based on work by Dale (1965) and Wesche and Paribakht(1996), the VSC assesses students’ familiarity with the same 30 words3and 10 pseudowords Students rated their own knowledge of the wordsusing the following scale: (1) I do not know it, (2) I have heard it, (3) Iknow something about it, and (4) I know it well and can use it.Responses on the VSC were averaged for analysis, such that an overallaverage score close to four indicated students reported knowing thewords well enough to use them, whereas an average score close to oneindicated that students reported knowing few word meanings On thefall pretest, only 4 of the 30 WG words were reportedly known by 80% ormore of all students, indicating that the WG target words meritedexplicit instruction Further, the 10 pseudowords were reported asunknown by 49 (88%) of all students, suggesting that students weregenerally able to assess their own word knowledge Both the MC andVSC pretests significantly correlated with the standardized literacymeasures (see Appendix C)

Additionally, students in the treatment group wrote persuasive essays

on a weekly basis All essays were transcribed by three trained researchassistants and scored using a researcher-developed rubric that consisted

of three sections: ideas, overall cohesion/structure, and academiclanguage (see Appendix D) Interrater agreement was 81 (Kendall’sCoefficient of Concordance), and raters were blind to the time (i.e.,week) in which the essays were produced Analyses of the essaysprovided information on WG target word use, as well as on overallwriting quality

Standardized Measures

Vocabulary, listening comprehension, and reading comprehensionwere assessed using the Group Reading Assessment and DiagnosticEvaluation (GRADE) (Williams, 2002) The GRADE is an untimedgroup-administered reading test that includes vocabulary, listeningcomprehension, sentence comprehension, and passage comprehensionsubtests The Reading Vocabulary subtest measures students’ vocabularyknowledge without the benefit of contextual clues Students silently read

3 There was one exception to this set of 30 randomly selected WG target words for the pretest: reluctant appears on the self-check and not on the multiple-choice test; culture appears on the multiple-choice test and not on the self-check This was corrected for the posttest.

a phrase or short sentence in which one of the words is printed in boldtype, and they select the appropriate meaning of the word from a list offour or five choices The Listening Comprehension subtest measuresstudents’ linguistic comprehension without printed cues Students listen

to a sentence or pair of sentences that are read aloud by the testadministrator, and they then select one of four pictures that bestmatches what was read aloud to them Finally, the Sentence and PassageComprehension subtests yield the ‘‘comprehension composite.’’Sentence Comprehension measures students’ comprehension of asentence as a whole thought or unit Students silently read shortsentences in which one of the words is missing (as indicated by a blank),and they then select the appropriate word from a list of choices PassageComprehension measures students’ comprehension skills with anextended passage After silently reading a passage with one or moreparagraphs, students answer multiple-choice questions about thepassage Level 5, Form A was used at pretest Internal consistencyreliability was reported as 95

Word reading accuracy and fluency were assessed individually usingthe Test of Word Reading Efficiency (TOWRE; Torgesen, Wagner, &Rashotte, 1999) The Sight Word Efficiency subtest assesses the number

of real words a student can read in 45 s; the Phonemic DecodingEfficiency subtest assesses the number of pronounceable nonsensewords a student can read in 45 s The TOWRE subtests were combinedfor analysis (r 5 8219, p 5 , 0.0001) Internal consistency reliability wasreported as 95 and scorer reliability as 99

RESULTS

Research question 1: Following the 20-week vocabulary intervention,

do the treatment or contrast group students gain knowledge of a greaternumber of targeted words?

Analytic Approach

First, the treatment and contrast groups were compared on allmeasures (researcher-developed and standardized) prior to thebeginning of the intervention Next, posttest performance on theresearcher-developed measures was examined by conducting Bonferonit-tests on the gain scores (change from pretest to posttest) Finally,standardized effect sizes (Cohen’s d; Cohen, 1988) were computedusing differences in mean performance divided by the pooled standarddeviation

Pretest Performance

Except for the TOWRE, on which the treatment group scored slightlyhigher, the treatment and contrast groups were very well matched atpretest (see Table 2)

Posttest Performance

Table 3 presents the results of posttest performance, including means,standard deviations, gain scores, significance tests, and effect sizes on theresearcher-developed measures

Word Generation Multiple-Choice Test

The treatment group students gained knowledge of approximately sixtarget WG vocabulary words, whereas the contrast group students gainedknowledge of about one Only 30 of the 100 taught WG vocabulary wordswere randomly included in the MC test, and thus a gain of 6 target words

on the MC test translates into gained knowledge of approximately 20target words and a gain ratio of 1 into approximately 3 Put differently,students in the treatment group went from knowing about 65% of thetarget words at pretest to knowing about 83% at posttest, whereasstudents in the contrast group went from knowing 65% of the targetwords at pretest to knowing about 68% at posttest The difference in thegain was statistically significant, with an impressive effect size (d 5 1.24)

TABLE 2

Fall Pretest Scores on All Literacy Measures for the Treatment and Contrast Group Students (N

5 49)

Treatment (n 5 24)

Contrast (n 5 25)

Bonferroni

p values

Effect size

WG multiple-choice testa Mean 19.71 19.96 1.00 0.06

hensiond

Mean 15.08 13.84 1.00 0.28

SD 4.35 4.63 GRADE listening compre-

hensione

Mean 13.50 13.36 1.00 0.07

SD 1.77 2.18 TOWRE word reading flu-

ency composite

Mean 56.06 48.50 0.04 0.86

SD 8.23 9.39 GRADE comprehension

compositef

Mean 28.67 26.32 1.00 0.25

SD 8.83 9.91 Note SD 5 standard deviation.aMaximum 5 30 b Maximum 5 4 c Maximum 5 4 (a high average is not desirable, as it indicates that students reported knowing nonsense words well enough to use them) d Maximum 5 35 e Maximum 5 17 f Maximum 5 39.