USING PAIRED READING TO HELP ESL STUDENTSBECOME FLUENT AND ACCURATE READERS

DAQi LiSUNY-Oneonta

SANDRA NES

Texas Tech University

Many ESL students with limited English proficiency (LEP) strugglewith reading fluency and accuracy. For these students to experienceschool success, educators must find ways to help them master such read-ing skills. This study investigated the effects of paired reading onreading fluency and reading accuracy of four ESL students with LEP.Using a single-subject research design, the study paired the studentswith a skilled reader and examined students' reading performances underdifferent controlled conditions. The results of the study showed that allfour students benefited from the paired reading intervention and demon-strated steady improvement in reading fluency and accuracy. Thisfinding indicates that paired reading can serve as a useful instructionalalternative to facilitate ESL students in learning to read in English.

Over the past years, schools in theUnited States have witnessed a bigincrease in the number of students whospeak English as a second language(ESL). Many such students have lim-ited English proficiency. Languagebarriers can hinder academic progress,as well as social development. To insurethe high quality of education inAAmer-ican schools, education programs musttake into account the special needs ofstudents from diverse racial and lin-guistic backgrounds.

One of the central difficulties fac-ing ESL students is reading.Development of proficient readingskills is crucial to school learning forall students. Those who experience dif-ficulties in the development of suchskills do not do as well as other stu-dents in content area classes, have lowerself-esteem, are more likely to pose dis-cipline problems in school, and are lesslikely to graduate from high school(Shanahan & Barr, 1995). Thus, forevery ESL student to succeed in school,it is essential that educators provide

appropriate instruction in reading andcreate more reading opportunities forthese students.

The development of reading fluen-cy is an important aspect of learning toread. Fluent readers read faster and arem-ore likely to absorb more informa-tion from the reading content. Researchin the psychology of reading suggeststhat fluent word recognition may be aprerequisite for good comprehensionand enjoyable reading experiences(Nathan & Stanovich, 1991). For manyESL students, however, limited Eng-lish vocabulary slows down their speedof word recognition. This prevents themfrom reading fluently and enjoyingthese experiences. To help thembecome fluent readers, it is importantfor educators to examine instructionalalternatives that consider the uniquelearning characteristics of studentsalong with an understanding of thereading process. One way to do this isto investigate reading instruction andprogress by deeply exploring individ-ual readers within one-to-one reading

50

Using Paired Reading.../ 51

instruction. the other methods.In recent years, one-to-one instruc- Paired reading requires the reading

tion has received renewed attention and partners to read aloud. Reading aloudconsideration in many schools for early to elementary school students can haveprevention of reading failure and reme- many beneficial effects: a) it improvesdiation of reading problems. their)language skills and motivates themOne-to-one instruction can be imple- to read on their own; b) it makes stu-mented in a number of ways. One such dents familiar with books and theirwav is known as "paired reading." language; and c) it keeps students'Paired reading is an instructional attention to the context and lends itselfmethod that involves the pairing of a to vocabulary learning (Saban, 1994).skilled reader with a less-skilled read- In light of these successful inter-er. The skilled reader demonstrates ventions, it is reasonable to ask aboutappropriate reading rate, inflection, and the effect of this intervention strategypausing for the less-skilled reader. In on ESL students who are strugglingpaired reading, the skilled reader in with reading fluency. If paired readingeach pair reads the connected text first. also works effectiveiv for ESL studentsThen the less-skilled reader reads the who struggle daily with reading, thesame text. Thus, the less-skilled read- strategy will surelv have even greaterer has a role model of fluent reading, instructional value. The purpose. of thisas well as repeated exposure to text study was to investigate the effects of(Mathes, Fuchs, Fuchs, Henley, & paired reading on reading fluency andSanders, 1994). In this instructional accuracy of four ESL students whomethod the pairing configuration may were less-skilled readers. Specifically,consist of parent/child, teacher/student. the studv aimed to find out whether theor student/student. paired reading intervention would

Numerous studies have demon- increase ESL students' reading fluen-strated; th:eeffectiveness of paired cy and improve their reading accuracvreading for improving students' read- at the same time.ing performance. Rasinski andFredericks (1991) reported on a paired Methodread ing project launched by the Akron,Ohio Public School Systern. Results of Participantsthe project suggest that paired reading Four ESL students (two bovs andhelped improve reading performance, two girls) in the first, second, and thirdin addition to improving reading moti- grades from an urban school district invation and parent/'child bonding. Leach West Texas participated in this studv.and Siddall (1990) compared paired They were all of Chinese origin andreading with direct instruction; a pause, had been residing in the United Statesprompt, and praise method; and hear- for onlv a short period, varying from aing methods of reading instruction. few weeks to a few months. All wereThese researchers found that paired enrolled in the regular education pro-reading and direct instruction tech- grams of their school, but each one ofniques resulted in significantly greater them was pulled out every day for aboutrates of progress in fluency rates than 40 minutes to study in an ESL class.

52/Reading Improvement

The four participants were selectedbased on the following criteria: (a) Eng-lish not as primary language, (b) limitedEnglish proficiency in the area of read-ing as determined by participants'performance on the Burns and RoeReading Inventory, and (c) reading levelone year or more below grade place-ment. To determine the students'reading fluency, the Burns-Roe Read-ing Inventory was administered to eachof them. They were asked to read a listof words and some short passages. Priorto the study, the parents of all partici-pants had expressed concern about theirchildren's reading problems and lan-guage learning difficulties. They werewilling to let their children participatein this study. The students' names usedhere are fictitious. Their characteristicsare described below.

Wen wenAt the time of the paired reading

intervention, Wenwen was a 9-year-oldthird grader. She had been living in theUnited States for less than a monthwhen the study began. She had diffi-culty with word recognition,comprehension, and fluency in read-ing English. Wenwen was a bright girland was a little shy. Her inadequateEnglish hindered her from interactingwith her schoolmates on a daily basis.At home, she and her parents commu-nicated mostly in Chinese. Wenwenloved animals and liked to keep littledogs and cats as pets.

KatieAn 8-year-old girl at the onset of the

study, Katie was in the second semes-ter of second grade, and she had beenliving in the United States for 3 months.Katie had an outgoing personality and

adapted to the new environment fast.Her performance on the Burns and Roeindicated that she was below grade levelin both word recognition and readingfluency. At school, she had difficultycompleting assignment and taking tests.Katie's parents were very concernedabout her learning and indicated theywould seek additional help for her. Likethe other three participants in the study,her primary language at home was Chi-nese. Katie enjoyed outdoor activitiesand was often playing games with herfriends.

RyanRyan was 9 years old and was in the

third grade when the reading interven-tion began. Having been in this countryfor less than a month, he demonstrat-ed weaknesses in listeningcomprehension, speaking, and readingin English. According to his parents,Ryan had broad interest. He lovedpainting, astronomy, and basketball inparticular. However, his English profi-ciency was very limited and his parentsexpressed great concern about his lan-guage difficulties. At school, thelanguage barrier not only preventedhim from effectively learning school-related subjects, but also restricted himfrom interacting with his peers.

JoeyA 7-year-old boy in the first grade,

Joey was happy and outgoing. WhenJoev first participated in the study, hehad been living in this country forjustthree months. Understandably, his read-ing level was well below grade level.He was experiencing difficulty in wordrecognition and passage comprehen-sion. His reading was characterized bylack of both fluency and accuracy. Joey

Using Paired Reading... / 53

was enthusiastic and eager to partici-pate in the study. His hobbies includedpainting, playing basketball, and play-ing computer games.

The IntervenerOne of the authors served as the

intervener (skilled reader) for the fourparticipants. The skilled reader was adoctoral student with past experienceof teaching reading to students whosemother tongue was not English. Hespoke both English and Chinese flu-entlv and was familiar with the culturalbackground of the participants.

Setting and MaterialsThe paired reading sessions were alI

conducted at the students' homes withpermission from their parents. Theenvironment was quiet and free fromoutside distractions. This instructionalsetting was chosen because the homeenvironment was considered morecomfortable and relaxing to the stu-dents. In this setting. it was anticipatedthat students would feel less pressureand restriction, and therefore, wouldmore likely read at ease. For the pur-pose of one-to-one instruction, theskilled reader and student in each read-ing session were seated side by side ata table so that they could read from thesame book. Children's storvbooksserved as the text materials for thestudv. The books were selected inaccoriance with the students' readinglevels. During the readitng sessions, theskilled reader was also provided withcopies of the stories to record students'reading errors.

MeasuresOral reading fluency rate and read-

ing accuracy were the areas of interest

in this research. Oral reading fluencywas defined as reading connected textout loud with enough speed and accu-racy such that the words were groupedin thought units in a smooth and flow-ing manner. Fluencv was recorded asthe rate of words per minute and cal-culated as the number of words read,divided by the number of seconds, mul-tiplied by 60. Reading accuracy wasrecorded as a percentage and was cal-culated as the number of words readcorrectly without omission, substitu-tion. insertion, or miscue, divided by thenumber of words in the passage, mul-tiplied by 100. Self-corrections andrepetitions made by the student werenot counted as errors.

All sessions were tape-recorded.During each session, the intervener cal-culated and recorded the participant'sreading fluency rate and accuracy data.In addition, a trained graduate studentfunctioned as an independent observ-er. The independent observer calculatedreading fluency rates and recordedaccuracy data for 30% of the sessions.Using a stopwatch the independentobserver listened to the audio tapes,timed the readings, marked errors on acopy of the text, and recorded fluencyrates and accuracy data. The results ofthe intervener and the independentobserver were then compared to deter-mine interobserver reliability. It wascalculated by dividing the lower rateby the higher rate and multiplving by100, Results showed overall interob-server agreement was high for all 4students. The interobserver agreementon Wenwen's fluency rate was 97.80%,while the agreement on her accuracyrate was 99.63%. The interobserveragreement for Katie's reading fluencyand accuracy were 96.86% and

54/Reading Improvement

99.35%. respectively. For Ryan's read-ing, the percentages of agreement onfluencv and accuracv were 97.36%° and99.29%/e. In Joey's case, interobserveragreement was 93.08% for reading flu-ency and 98.93% for accuracypercentages.

Experimental Conditions andDesign

This study used a modified single-subject A-B design to examine theeffect of the paired reading method ofinstruction. This design provides aframework within which behavior canbe objectively measured under con-trolled conditions (Tawnev & Gast.1984).

BaselineTo determine the participants' read-

ing fluency and accuracy prior to theintervention, sessions were conductedto collect baseline data for each stu-dent. For Wenwen, data collectioncontinued until a relatively stable flu-ency level was observed. Duringbaseline, the intervener observed andrecorded her reading fluency rate andrecorded the accuracy percentage. Theother three students (Katie, Rvan, andJoey) were unable to read indepen-dentlv. As was determined byadministering the Burns-Roe ReadingInventory, these three students read thepreprimer or primer passages at thefrustration level. Their fluency rateswere from less than 15 words perminute to about 40 words per minute,while their error rates were over 40%.Because of their tremendous difficul-ty in reading, and to avoid furtherfrustration, intervention sessions start-ed immediatelv following theadministration of the reading invento-ry. The results of the inventory

administrations were used as the base-line fluency level of the three students.In order to monitor their independentreading fluency levels, probes wereconducted regularly, in which the stu-dents were asked to read passages bythemselves (without the paired readinginterventions).

IrnterventionThe intervention sessions began with

the skilled reader reading first. The stu-dent was instructed to follow alongduring reading. At the end of each pas-sage, the roles were reversed, and thestudent read the same passage. If thestudent miscued or omitted words, theskilled reader did not interrupt but madenotes of any errors. However, if the stu-dent hesitated, the skilled reader waitedfor 3 seconds and if the student did notrespond, the skilled reader pronouncedthe word for the student and recordedit as an error. During the first few ses-sions, the passages being read each timewere shorter so that the students didnot struggle along. Gradually, each pas-sage became longer (up to 30 or morewords). Each student read for approx-imately 20 minutes per session. As inthe baseline phase, probes were alsoconducted to examine students' inde-pendent performance without pairedreading.

MaintenanceA maintenance phase was used to

examine whether or not the effects ofpaired reading would be maintainedonce the intervention was not contin-ued on a daily basis. In this phase, thestudents still had paired reading; how-ever, the sessions were conducted onlyapproximately once every 10 days.

Using Paired Reading..." 55

Results

Readitig FluencyFigures I to 4 present the reading

fluencv of Wenwen, Katie, Rvan. andJoeyv respectivelv. During the baselinephase, Wtenwen was reading at an aver-age rate of 34 words per minute. Afterthe intervention of paired reading start-ed, her reading showed immediateimprovement in fluency. During thefirst two sessions of intervention. shewas readincg at 78 and 79 words permin.ute. During the following sessions.,her rate showed variations. However,

the general trend was on the rise. Themean of her fluency rate during base-line was 86 words per minute. Wenwenwas able to make more progress dur-ing the maintenance phase by readingstories at an average rate of 1 1 2 wordsper minute. This was considered a sig-nificant improvement over her baselineperformance. To evaluate her indepen-dent reading, probes were conductedduring both intervention and mainte-nance. The result showed her readingperformance with paired reading wasbetter than her independent reading per-formance.

Figure 1Wenwen's Reading Fluencv Performance

1 3 5 7 9 1t 13 I5 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71

I_ __ i .. - - _ .,

C

II11�Ei

0

56/Reading Improvement

Katie, Ryan, and Joey were studentswho experienced tremendous amountof difficulty in reading independently.They had very limited vocabulary andknew few phonological rules for read-ing words out loud. Because of this, nobaseline sessions were conducted forthese three students. Instead. their per-formances on the Burns-Roe ReadingInventory were used as baseline data.

As shown in Figure 2, Katie's flu-ency rate in reading two passages ofthe Burns-Roe Reading Inventory waslow, averaging 35.5 words per minute.This rate was taken as her baseline per-formance. When she started pairedreading with the intervener, there wasan apparent increase in her fluency. Shewas able to read at 73 words per minute

at the very first session of intervention.During this session, she was also askedto read a story by herself. Her rate ofindependent reading was approxi-mately 47 words per minute, muchlower than that of paired reading. Dur-ing the following intervention sessions,her reading showed variations in termsof fluency rate. However, a therapeu-tic trend was observed from the graph.Overall, her average paired reading flu-ency rate during the intervention phasewas 69 words per minute. On the otherhand, her probed independent readingfluencv average during the same phasewas 48 words per minute. After theintervention sessions stopped, Katiewas still reading at an average rate of66 words per minute.

Figure 2Katie's Reading Fluency Performance

Intenention j Mintenance

l

.

L z_*_uwok R-V aPhwpam 1

Baselne

1X, j0 I IIs80

r60

.40

20

l

/4

, . , , , , ~~~..,. .. . . . . . . . . .

1 2 3 4 5 6 7 8 9 1011121314151171819 202122232425262726 293031323334 353637839s404

Sulkn

Using Paired Reading...t 57

Ryan's reading of a Burns-RoeReading Inventorvy passage indicatedthat he had great difficulty with read-ing fluenqc because he had very limitedvocabulary. He was reading at a rate ofless than 15 words per minute. Duringthe intervention sessions, Ryan demon-strated a gradual but steady increase inhis reading fluency rate. On average, heread the stories at the rate of nearlv 65words per minute. His average rate ofindependent reading during this phasewas about 42 words per minute. Dur-ing maintenance, Ryan was able to readstories even more fluently, with a meanof 96 words per minute.

Joev's fluency rate is shown in Fig-ure 4. Prior to intervention, his readingfluency was determined to be 35 wordsper minute by using the Burns-RoeReading Inventorv. When he was intro-duced to paired reading, his fluencyrate showed increase. During the first7 intervention sessions, he was alsoasked to read story passages indepen-dently. Compared with this independentreading, his performance in paired read-ing during these sessions was better interms of fluency rate. During later ses-sions, his independent reading fluencywas probed only occasionally. His aver-age paired reading fluency rate duringintervention was 54.59 words per

Figure 3Ryan's Reading Fluency Performance

.~~~~~~~~~~~~~~~~~~~~

asebim

rhkrerflo,

II

a~~~FAS;~ .R i *< jv' : : a :

: : : : : :: ::: ::a

II

140i4

II 10tS

:f

I :

I .tt_

t$ S r B i ¶1 "I 1S t 2t 22 2s 27S 2?23t M U7 rw 4f41 0 t 4g 54)t S5557 s 6: 63

___ a Pa-fti4Raadtt

I

I

58/Reading Improvement

minute, higher than that of baseline.During maintenance, his fluency ratewas even higher, averaging 67.33 wordsper minute.

ReadingAccuracvAnother aspect of reading under

investigation in this study was students'reading accuracy. It was assessed bycalculating the percentage of wordsread correctlv by the students. Theresult is presented in Table 1.

Figure 4Joey's Reading Fluency Performance

!. s . .......T ,,,. ...,,-

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57

Sessions

Li - Idependent Readtng - -Pared Reading

Over the course of the baseline phaseof data collection, Wenwen's readingaccuracy ranged from 79% to 91% witha mean of 85%. When the paired read-.ing intervention was initiated,Wenwen's accuracy increased while thenumber of her reading errors decreased.On average, her accuracy rate increasedto 96% during the intervention phase.In the maintenance phase, her accura-

cy rate averaged 99%. Katie's accura-cy percentages were recorded asfollows: baseline phase, mean accura-cy percentage was 75%; interventionphase, accuracy in paired readingranged from 72% to 96% with a meanof 88%; maintenance phase: accuracyranged from 84% to 90% with a meanof 87%. Both Rvan and Joey had greatdifficulty in reading accurately during

120

100

s80

IL

0.

, 40

20

* a

.

"emrtion

Using Paired Reading . . 59

the baseline phase. Error rate in theirreadings was recorded over 40%, Afterthe introduction of paired reading, how-ever. thev both inmproved in readingaccuracy. In the intervention phase.Rvan's accuracy ranged from 56% to100% with a mean of 89%. His accu-racy in the maintenance phase was evenhigher. with a mean of 96%. Joey'saccuracv in the intervention phaseranged from 74% to 99% with a meanof 91%. In the maintenance phase, hisaccuracy ranged from 94% to 98% witha mean of 96%.

DiscussionIn this study, the intervention of

paired reading helped the 4 studentsimprove both reading fluencv and read-ing accuracy. Compared with their ownperformances at the beginning of thestudy, the students were able to readmore fluently and more accurately dur-ing the maintenance phase of theexperiment. This result provides fur-thier evidence that one-to-oneinstruction is effective in helping stu-dents develop efficient reading skills(Wasik & Stavin, 1993; Juel, 1996).The effectiveness of one-to-one instruc-

Table IMean accuracy percentages by student and phase of studv

Name Baseline Phase Intervention Phase Maintenance PhaseProbe Paired Reading

Wenwen 85% 92% 96% 99%

Katie 75 % 79% 88% 87%

Ryan 52%* 80% 89% 96%

Joee 57.%* 76% 91% 97%

IuLEc. myau anu .oey nau onjy one reaaing during the baseline phase.

tion is not limited to students who speakEnglish as a native language. Thosewho speak English as a second lan-guage can also benefit from suchinstructional strategy.

For ESL students who have limitedEnglish proficiency, paired reading hasa number of advantages over tradition-al classroom reading instruction. First,paired reading reduces students' pres-sure and anxietv, which they often

experience in reading classes at school.Because many ESL students have poorcommand of English, they tend to beshy or reluctant to speak out. Thisreduces their chance to practice the newlanguage. Paired reading is a form ofindividualized tutoring. In this learm-ing context, students feel less pressurethan in class instruction. Thus thev maybe more likelv to focus on the readingactivity. Consequently, they are likely

----------

__

60/Reading Improvement

to make progress and experience suc-cess.

Second, paired reading gives ESLstudents more opportunity to practicereading in a new language. ESL stu-dents with limited English proficiencvneed constant modeling and feedback,without which they will end up withinaccurate pronunciations and intona-tions. The students in this study allcame from families in which Englishwas rarely spoken among family mem-bers. The students had little opportunitvto read books with experienced readersat home. At school, their opportunity toread orally to a teacher or other skilledreaders was limited. This lack of oppor-tunity may hinder their progress indeveloping efficient reading skills.When the researchers of this study con-tacted the parents about the pairedreading project, all of them were excit-ed and willing to let their childrenparticipate in the study. The frequentpractice undoubtedly helped thembecome more fluent in reading. Theconstant modeling and feedback werebelieved to have helped them improvereading accuracy.

Third, paired reading was flexibleand easy to make adaptations. The stu-dents in this study began with very lowreading skills. To accommodate to theirspecial needs, the storybooks wereselected according to their diffexentreading levels. As they progressed.more difficult books were used. Suchadaptations were to ensure that thebooks were neither too difficult nor tooeasy for the students. In addition, theintervener varied the reading speedaccording to the difficulty level of thereading material and the reading levelof the students. Thus, the studentsalways felt the challenge, which was

needed to stimulate their greater effort.At the same time, the challenge wasnot so overwhelming that it frustratedthe students. The frequent adaptationsassured students' steady progressthroughout the studv.

Although the results were positivefor the four participating students, thereare a number of limitations to this study.First. the small sample size limits thegeneralizability of the results. Furtherreplication of this research across moresubjects is warranted in order to gain aclearer picture of the effectiveness ofthe intervention for students of differ-ing ability levels, cognitivecharacteristics, temperaments, and agelevels.

Second, the study lasted for morethan eight months (except for Katie,whose participation was shortened tofive months because her parents wererelocated to another city). During thisperiod of time, the students were attend-ing a public school where they hadopportunities to interact with teachersand peers. There was reason to believetheir overall language skills hadimproved as a result of school learn-ing and daily exposure to English atthis natural environment. This maturi-ty factor thus could also havecontributed to the students' increase inreading fluency and accuracy. For thisreason, this study used probes toexplore the students' natural improve-ment. Results indicated theirindependent reading did increase overthe course of the studv. Therefore, theresult of the study should only be inter-preted in the light of this understanding.

ConclusionIn conclusion, the intervention of

paired reading can be effectively used

Using Paired Reading...! 61

to improve ESL students' reading per-formance in English. With its uniquecharacteristics, this one-to-one inter-vention helps ESL students byproviding constant modeling and feed-back as well as plenty opportunity topractice. The students who participat-ed in this studv all demonstrated steadvincreases in reading fluency and read-ing accuracy. This finding isparticularly significant to those educa-tors who are seeking ways to help ESLstudents with reading in English.

ReferencesJuel, C. (1996). What makes literacv tutoring effective?

Reading Research Quarterls, 31. 268-289.

Leach. D.J .& Siddall, S,W. (1990). Parental involve-ment in the ceachbng of reading; A comparison ofhearing reading. paied reading. pause. pronmpt andpraise, and direct instruction methods. British Jour-nal of Educatie aI Psvchology, 60, 349-35S.

Mathes, P. G.. Fuchs. D.. Fuchs. L, S., Henlev A, M., &Sanders,A. ( 1994). Increasing strategic reading prac-tice with Pcabody Classwide Peer Tutoring.Learning Disahilries Research and Practice. 9. 44-4X.

Nathan. R, G., & Stanovich. K. E. (1991). The causesand consequences of differences in reading fluency.Theory Into Practice, 30. 176-184.

Rasinski. T: V., & Fredericks, A. D. (f991). The AkronPaired Reading Project The Reading Teacher. 44.514-515.

Sahan, A. (1994). A bridge to hooks: Reading aloud tochildren in primary education. Reading Improve-mnent, 31(1). 23-27.

Shanahan, T.. & Barr. R. (1995). Reading Recoverv: Anindependent evaluation of the effects of an earlyinstructional intervention for at-risk learners. Read-ing Research Quarterby, 30. 958-996.

Tawney, J. W.. & Gast. D. L. (1984). Single subjectresearch in special education. Columbus, OH: Mer-rill.

Wasik, B. A., & Slavin, R. E. (1993). Preventing earlyreading failure with one to one tutoring: A reviewof five programs. Reading Research Quarter/v, 28.i ?9-200.

Author's NoteThe study reported in this paper was

supported by a grant from the TexasHigher Education Coordinating Board.

COPYRIGHT INFORMATION

TITLE: Using paired reading to help ESL students become fluentand accurate readers

SOURCE: Reading Improvement 38 no2 Summ 2001WN: 0119600995001

The magazine publisher is the copyright holder of this article and itis reproduced with permission. Further reproduction of this article inviolation of the copyright is prohibited.

Copyright 1982-2001 The H.W. Wilson Company. All rights reserved.