Think-Aloud Strategy

13 Pages·2007·0.11 MB·English

by  Lisa Kochel

© 2007 INTERNATIONAL READING ASSOCIATION (pp. 136–147) doi:10.1598/JAAL.51.2.5
TThhiinnkk--AAlloouudd SSttrraatteeggyy:: MMeettaaccooggnniittiivvee ddeevveellooppmmeenntt
aanndd mmoonniittoorriinngg ccoommpprreehheennssiioonn iinn tthhee mmiiddddllee
sscchhooooll sseeccoonndd--llaanngguuaaggee ccllaassssrroooomm
Regina G. McKeown, James L. Gentilucci
Teachers should not make the mistake of monitor their reading comprehension and employ
strategies to facilitate understanding oftext
considering all strategies as “good
(Baumann,Jones,& Seifert-Kessell,1993).It per-
teaching”and applying them equally to all mits these students to self-regulate the reading
process and improve comprehension
levels ofEnglish learners.
by employing “fix-up strategies”where
McKeown teaches at Cuesta
College in San Luis Obispo, needed (Cassanave,1988).Think-aloud
California. E-mail
also activates metacognitive monitor-
Given the largenumber ofU.S.federal [email protected].
Gentilucci teaches ing in those students for whom self-
and state accountability mandates as-
educational leadership at regulation has not yet become
sociated with English learners,K–12 the California Polytechnic
automatized (Ward & Traweek,1993).
State University in San Luis
teachers are struggling to find effective,
Obispo, California.
research-based strategies to help these The open-ended nature ofthe
students improve their learning,espe- Think-Aloud Strategy is a benefit for
cially their reading comprehension skills.Second- second-language learners because this nondirective
language learners (English is not their primary approach requires readers to stop and explore the
language) are exposed to unfamiliar idiom and text—a simple prescription for the reader to engage
cultural references while reading in a second lan- (Loxterman,Beck,& McKeown,1994).Think-
guage (L2) and,as a result,are required to repair aloud also makes sense conceptually because many
more gaps in the strategies they use to derive interpretive reactions are conscious (Pressley et al.,
meaning from text as they read (Block,1992). 1992) and can be made overt through the use of
Reading is a covert process actively controlled by this strategy (Baumann et al.,1993).
readers to create meaning from text,and the prac-
tice ofreaders “thinking about their thinking”
Background
while engaged in the reading process is known as
metacognition.One promising approach for acti-
Afflerbach and Johnston (1984) set the ground-
vating metacognition and thereby improving
work for the development ofthe Think-Aloud
reading comprehension among second-language
Strategy,first as a method ofmeasuring the cog-
learners is known as the Think-Aloud Strategy.
nitive reading process,then as an application in
The purpose ofthink-aloud is to help the metacognitive realm in which readers use the
second-language learners develop the ability to tool to monitor comprehension.Pressley et al.
136 JOURNAL OF ADOLESCENT & ADULT LITERACY 51:2 OCTOBER 2007
Think-Aloud Strategy: Metacognitive development and monitoring comprehension in the middle school second-language classroom
(1992) cited think-aloud as one ofthe “transac- judge whether comprehension is taking place and
tional strategies”because it is a joint process of that help them decide whether and how to take
teachers and students working together to con- compensatory action when necessary”(p.288).
struct understandings oftext as they interact with She suggested that reading researchers explore
it.Whereas readers independently derive meaning how second-language students can be helped to
from text by using background knowledge,inter- monitor their reading comprehension,focusing
ests,motivations,and purposes for reading, on the importance oftheir ability to access
group reading using think-aloud allows a small schema or background knowledge and their abili-
interpretive community to explore other view- ty to take necessary strategic action to correct
points and to negotiate understandings. comprehension deficiencies as needed (cf.August,
Flavell,& Clift,1984;Block,1992;Carrell,1989;
Kucan and Beck (1997) noted that think-
Carrell & Eisterhold,l983;Craine-Thoreson,
aloud achieves three goals:(1) It provides a method
Lippman,& McClendon-Magnuson,1997;
ofinquiry to understand cognitive processing relat-
Jiménez,1997;Wade,1990).Cassanave found that
ed to reading research;(2) it serves as a method of
students who used the Think-Aloud Strategy were
instruction;and (3) it is an aspect ofsocial interac-
able to improve the quality ofdialogues,generate
tion.Think-aloud has evolved,as discussed in
summaries ofthe main points that were no longer
Pressely et al.(1992),to incorporate an expanded
representation ofreading due to what Kucan and verbatim,and ask questions related to the main
Beck deemed “the potential influence ofsocial in- ideas ofthe text rather than less important details.
teraction in constructing meaning”(p.272).
Bereiter and Bird (1985) studied seventh-
and eighth-grade average readers as they learned
to monitor their comprehension using think-
Second-language studies
aloud.They found that students whose teachers
Fitzgerald (1995) addressed the impact ofthink- modeled think-aloud strategies for recognizing
aloud as a metacognitive strategy on English- comprehension problems and selecting repair
language learners in her extensive review of stratagems scored significantly higher on tests of
reading research in the United States.This analy- comprehension than those whose teachers did
sis was conducted to determine how readers ap- not.The authors stressed the importance offol-
proached texts and the ways in which they tried
lowing an instructional pattern that included (in
to repair miscomprehension.Ofthe metacogni-
this sequence) teacher modeling,direct instruc-
tive studies reviewed by Fitzgerald,10 used vari-
tion and explanation,and individual practice.
ous forms ofthink-aloud.
Cassanave’s (1988) findings supported this claim
The following were two main themes that by confirming the need for teacher modeling and,
emerged from Fitzgerald’s research:(1) Second- as with Fitzgerald’s (1995) study,demonstrated
language learners in the United States tend to the need to identify the monitoring and repair
monitor their comprehension,and (2) these stu- strategies successful second-language readers use.
dents employ a variety ofmetacognitive strategies
Carrell (1989) examined metacognitive
to accomplish this.Fitzgerald’s work is cited here
awareness in second-language readers and pro-
because it confirms that second-language learners
vided a useful overview ofsuccessful reading
actively employ comprehension-monitoring tools
strategies for these students.She explored what
such as think-aloud,and it challenges researchers
Hosenfeld (1977) and Block (1986) described as
and teachers to identify the most useful think-
successful and unsuccessful strategies used by
aloud strategies for these readers.
second-language readers.Successful strategies in-
Cassanave (1988) defined comprehension cluded keeping the meaning ofthe text in mind
monitoring as “any behaviors that allow readers to during reading,integrating ideas,reading in
JOURNAL OF ADOLESCENT & ADULT LITERACY 51:2 OCTOBER 2007 137
Think-Aloud Strategy: Metacognitive development and monitoring comprehension in the middle school second-language classroom
“broad phrases”(top-down versus bottom-up), whole”(p.332).Both Clarke’s (1980) study and
recognizing aspects oftext structure,skipping Cummins’s (1979) study hold that there is a
words that are unimportant to the total meaning threshold oflinguistic competence necessary for
ofthe phrase,and using personal and general successful second-language learning.Those who
knowledge and associations.Block (1986) noted lack this competence in both the native language
that unsuccessful strategies were often reflexive in and at least the oral level in the second language
nature because readers directed their attention will not be successful in reading and other learn-
away from the text and focused on their own ing.In a similar manner,students in Chamot and
thoughts and feelings rather than on the informa- El-Dinary’s (1999) study focused on the bottom-
tion presented in the reading. up type processes ofphonetic decoding,whereas
more proficient readers used background knowl-
edge and inferencing to understand text.
Additional studies
Many second-language studies employing
Think-aloud has been examined in several studies
the Think-Aloud Strategy also confirmed the
to obtain data on the schema-theoretic model
powerful impact ofsecond-language vocabulary
and,within this model,to determine how exactly on second-language processing.Jiménez,in his
second-language learners approach text.In refer- 1997 study employing think-aloud,recommend-
ence to top-down versus bottom-up processing, ed that “students appreciate and take advantage
Block’s 1992 study claimed,“Most people now ac- ofour Latino/a cognate vocabulary,as this also
cept that the two processes interact”(p.319). capitalizes on their Spanish-language proficiency”
Davis and Bistodeau (1993) confirmed these (p.241).Jiménez’s study also included a recom-
findings by further detailing how the two hy- mendation that for low-literacy students in par-
potheses interact.The “short circuit”hypothesis ticular,educators should use “a strategic approach
by Clarke (1980) asserted that students who are to interacting with text”that makes use ofcultur-
skilled top-down processors in their native lan- ally relevant children’s literature”(p.240).
guage are forced to “short circuit”into a bottom-
In contrast,Block’s 1992 study claimed that
up orientation when attempting to make
we should not “chew up the text”or attempt to
meaning from second-language text.
“predigest”the printed material for students be-
In contrast,the “bi-orientation”hypothesis cause this does not prepare them to “eat on their
by Lee (1991) claimed that “beginning language own”;comprehension does not depend on un-
learners who are sophisticated readers are orient- derstanding every feature ofthe text (p.337).The
ed neither from bottom-up nor from top-down; proficient readers in Block’s study did not have to
they are bi-oriented”(p.200).Davis and understand all the words to achieve comprehen-
Bistodeau confirmed both hypotheses in their sion,and she stressed that comprehension is not a
study stating that “low linguistic proficiency re- smooth process,furthering the notion that top-
sults in much greater attention to bottom-up down readers can discern which aspects ofthe
components ofcomprehension,but it also pro- text are important and which can be ignored.
vided evidence that top-down components exert
Developing strategic reading skills is critical
a powerful influence upon the strategies used by
to English-learner (and,arguably,all student) suc-
novice L2 readers,”thereby strengthening Block’s
cess in the English/language arts classroom and in
assertion that the two processes interact (p.468).
all content area learning.The California
Chamot and El-Dinary,in their 1999 think- Department ofEducation’s Strategic Teaching and
aloud study,also confirmed that “less effective Learningreport (Pritchard & Breneman,2000)
learners focus too much on the details,whereas identified a strategic reader as one who coordinates
more effective learners focus on the task as a successfully up to eight key comprehension strate-
138 JOURNAL OF ADOLESCENT & ADULT LITERACY 51:2 OCTOBER 2007
Think-Aloud Strategy: Metacognitive development and monitoring comprehension in the middle school second-language classroom
gies,knowing when and how to use them.These increasing pressure to hold teachers and students
students are actively involved in the reading;have a accountable for mastering English/language arts
running dialogue with the text (the internalized standards,it is important for educators to be
think-aloud);visualize,predict,and relate new top- taught empirically tested methods for diagnosing
ics to prior knowledge;read with a specific pur- individual needs and planning effective reading
pose,accept ambiguity,and monitor their own instruction for this population ofstudents.
comprehension;and apply the proper fix-up strate-
This study examined how the Think-Aloud
gies when needed (i.e.,metacognitive awareness).
Strategy affects content area reading comprehen-
Other strategies more specific to second- sion ofmiddle school English learners by at-
language learners include discerning important tempting to answer the question,Will middle
from nonimportant details,applying cognate vo- school English learners who employ the think-
cabulary,making extensive versus reflexive re- aloud reading strategy demonstrate greater con-
sponses to text,and focusing on the text as a tent area comprehension as measured by the
whole (top-down processing).These studies re- High Point Selection Comprehension Assessment
soundingly confirm the need to first identify read- than those who do not employ the strategy?
ing strategies that are most effective with English
learners with varying degrees ofliteracy in their
Subjects
L1 and L2,and then to heighten the metacognitive
awareness ofthese strategies within students so
Purposeful sampling provided information-rich
that they use them appropriately and strategically.
data regarding the usefulness ofthe strategy and
As informative as they are,however,these its value as a monitoring device for reading com-
studies leave two unanswered questions:(1) What prehension in second-language learner metacog-
is the extent ofmetacognitive awareness,specifi- nition.It also enabled us to determine ifthe
cally in the form ofreading strategies,that strategy is more useful with English-language
second-language learners possess? (2) Which learners who possess certain levels oflanguage
comprehension strategies are the most effective in proficiency.Twenty-seven English learners with a
helping these students repair “gaps”in their reading proficiency level ofEarly Intermediate
meaning-making strategies? It is to the second of (Level 2) or higher were included in the study.
these two questions we now turn our attention. Five Early Intermediate students (Level 2),11
Intermediate students (Level 3),and 11 Early
Advanced students (Level 4) were included in the
Method
sample (see Table 1).
Due to the wide range ofreading abilities in the All five Level 2 students had been enrolled
English-language development classroom and the in K–12 schools in the United States for an
Table 1
Characteristics of the study sample (N = 27)
English-learner level Female (N= 10) Male (N= 17)
Level 2 (N= 5) 2 3
Level 3 (N= 11) 4 7
Level 4 (N= 11) 4 7
JOURNAL OF ADOLESCENT & ADULT LITERACY 51:2 OCTOBER 2007 139
Think-Aloud Strategy: Metacognitive development and monitoring comprehension in the middle school second-language classroom
Table 2
Language backgrounds of students in the study
sample (N = 27)
Native language Female (N= 10) Male (N= 17)
Spanish (N= 24) 9 15
Urdu (N= 1) 0 1
Hebrew (N= 1) 1 0
Tagolog (N= 1) 0 1
average oftwo years and had an average instruc- Design and procedure
tional reading level of2.5 as measured on the
A pretest and posttest ofrelated samples were
STAR Reading Assessment.In addition,all had an
used to test the hypothesis that there would be a
overall proficiency-level scaled score ofEarly
statistically significant positive difference between
Intermediate (Level 2) on the California English
mean scores ofpre- and posttests ofthe sample
Language Development Test (CELDT).This
groups,signifying that the use ofthe Think-
state-mandated assessment includes listening and
Aloud Strategy is an effective intervention for im-
speaking (50% ofthe scaled score),reading,and
proving reading comprehension among the
writing (25% each ofthe scaled score).
English-learner population.Although use ofa
The Level 3 students were a mixed group; control and experimental group design was ini-
four had arrived in the United States just two tially considered as part ofthe pre- and posttest
approach,we decided against it for two reasons.
years prior,three had been U.S.residents for four
First,we refused to deny halfthe students the
years,and the remaining four were born in the
possible benefits oflearning and applying the
United States but had not been able to make sat-
Think-Aloud Strategy as quickly as possible.
isfactory progress in school to exit the English-
Although Regina (first author) could have taught
learner program.All had an average instructional
some students the strategy ifthe study had
reading level of3.5,and all were in the
proven it effective,she decided to include all stu-
Intermediate range on the CELDT (Level 3).
dents in the treatment because time is critical
Nine ofthe 11 Early Advanced group (Level when attempting to recoup academic deficits.
4) were at or above grade level in their native lan- Second,the sample groups were small,ranging
guage,had supportive families who made school from 5 to 11 students,and to divide them in half
a priority,did their homework,and had a positive for treatment and control purposes would have
attitude about school.Two ofthese 11 had further limited the significance ofany findings
climbed four levels ofEnglish proficiency in one due to the smaller number ofeach group.
year and were scheduled to exit the English-learner
program into regular ninth-grade English the fol-
Instrumentation and classroom
lowing year.The average reading level for this
procedures
group was 4.5.Ten ofthe 11 received an Early
Advanced score on the CELDT (Level 4),with We chose to employ the High Point
one student receiving an Intermediate score Comprehension Assessment (Schifini,Short,&
(Level 3).In addition,four language backgrounds Tinajero,2002) for this study because it was al-
were represented in the study (see Table 2). ready embedded in the school’s instructional pro-
140 JOURNAL OF ADOLESCENT & ADULT LITERACY 51:2 OCTOBER 2007
Think-Aloud Strategy: Metacognitive development and monitoring comprehension in the middle school second-language classroom
Table 3
Comparison of High Point expository selections
Text selection Student Subject High Point Word count
level level
“Grandfather’s Nose” (pretest) 2 Science Level A 372
“When Disaster Strikes” (posttest) 2 News Level A 1,114
“Talking Walls” (pretest) 3 Photo essay Level B 824
“Teammates” (posttest) 3 Biography Level B 1,180
“Home, Sweet Home Page” 4 How-to Article Level C 1,136
(pretest)
“Space Exploration” (posttest) 4 Science Level C 1,740
gram.This obviated any disruption ofinstruc- Assessment.Regina then applied the treatment by
tional routines and consequent loss ofinstruc- teaching students to use the Think-Aloud
tional time.In addition to its usefulness for Strategy following Bereiter and Bird’s (1985) ap-
providing formative assessments to guide instruc- proach for teacher modeling ofdesired learning
tion throughout the year,the High Point strategies.This explicit teacher modeling oc-
Language,Literacy,and Content series was the curred over a period oftwo weeks during the 50-
only curriculum adopted by the State of minute reading class and lasted from 20 to 30
California as both a systematic English-language minutes,three days each week.Regina used social
development program and reading intervention science texts as well as the novel The Outsidersby
program.Furthermore,High Point provides S.E.Hinton (2007,Puffin),which she read aloud
English proficiency-level placement in one of to the whole group.After every two or three lines
four levels ofdifficulty.Finally,we chose High oftext,she stopped and restated what she
Point due to a wide selection ofexpository arti- thought was happening,asked herselfa question,
cles within the series. clarified,or made a prediction,thus modeling her
own meaning-making strategies for the students.
Selections from the same form ofHigh Point
were used to pre- and posttest the three groups of During the third and fourth week ofthe
English learners.The assessments included six study,students began applying the Think-Aloud
multiple-choice questions,and the publisher pro- Strategy to their daily reading assignments in so-
vided a scoring guide,thereby eliminating scorer cial studies,or they used an Accelerated Reader
subjectivity.Each group was assessed using an ex- novel to practice strategic application ofthe strat-
pository article appropriate for the students’level egy.Regina monitored students as they read
oflanguage proficiency.All articles were examined aloud,prompting them to respond aloud about
using Fry’s (1977) readability index,and all fell “what was going on in their heads”and encour-
within the middle school range for readability aged them to think aloud in whatever language
(sixth to eighth grades) as well as reading age was more comfortable.
(11–13).The only significant difference among
Regina also used an instructional model
the articles was their length (see Table 3).
based on the work ofBaumann et al.(1993).The
During the pretests,group members silently model stressed that good readers ask questions,
read the expository selection and then completed determine ifthe information they receive consti-
the High Point Reading Comprehension tutes new or prior knowledge,predict,clarify,and
JOURNAL OF ADOLESCENT & ADULT LITERACY 51:2 OCTOBER 2007 141
Think-Aloud Strategy: Metacognitive development and monitoring comprehension in the middle school second-language classroom
reread or retell stories.If,after using this model, Results and discussion
the text still did not make sense,students were
Use ofthe Think-Aloud Strategy did not yield
taught to employ additional metacognitive fix-up
strategies taken from Bereiter and Bird’s (1985) homogeneous results across the English-learner
approach for using the think-aloud.This model subgroups (i.e.,Level 2,Level 3,and Level 4).A
stressed substituting another word for an un- briefdiscussion ofthe effects ofthink-aloud on
known word,using pictures and graphs to clarify the difference between mean pre- and posttest
the meaning ofsurrounding text,and ignoring scores ofeach group follows.
confusing portions ofthe text that could be clari-
fied later with Regina’s help or that ofa peer.
Pre- and posttest findings, Early
These fix-up strategies were laminated and post-
Intermediate students (Level 2)
ed in the classroom,and Regina reviewed the fix-
up strategies before each reading session over a
When the pre- and posttest scores for the Early
period oftwo weeks to reinforce their effective-
Intermediate English learners (n= 5) were tested
ness as comprehension tools.Students practiced
using a two-tailed related samples t-test,the t-criti-
the strategies both independently and in pairs.
cal value for the paired means was 2.776 and the t-
To practice particular fix-up strategies or observed value was zero.For this group,the results
think-alouds during this timeframe,Regina show that use ofthe Think-Aloud Strategy did not
would flip the light switch during reading ses- help improve the English learners’comprehension
sions and say,“OK,stop reading.Tell your partner ofexpository text;individual scores were nearly
(or “say out loud”for independent work) what’s identical on pre- and posttests (see Figure 1).
going on in the story right now,”or “What did
this passage remind you of?”or “Is the story mak-
Pre- and posttest findings,
ing sense?”or “Say what you think is going to
happen next.”Partner work and guided inde- Intermediate students (Level 3)
pendent practice kept students on task and
The t-observed value for paired means for this
helped them practice using the strategies inde-
group ofEnglish learners (n= 11) was -2.185
pendently.After two weeks ofpaired reading and
with a t-critical value of2.228.Although the dif-
independent practice,Regina administered the
ference in pre- and posttest means was not statis-
posttests to students in each group to measure the
tically significant,the data suggest measurable
effect ofthe think-aloud treatment.
growth in students’reading comprehension be-
tween pre- and posttests.Seven of11 students in-
Data analysis creased their posttest score,2 remained the same,
and 2 scored lower on the posttest than on the
Mean differences between pre- and posttest
pretest (see Figure 2).
scores for each group were calculated to deter-
mine the effect ofthe treatment.Two-tailed t-
tests (p< .05) ofrelated samples (i.e.,paired Pre- and posttest findings, Early
means) were used to determine ifthere was a sta-
Advanced students (Level 4)
tistically significant difference between the mean
scores ofpre- and posttest assessments.The dif- The t-observed value of2.951 for this group of
ference between means was tested under the null readers (n= 11) exceeded the t-critical value of
hypothesis X1= X2(i.e.,there would be no sig- 2.228.For most ofthis subgroup ofEnglish learn-
nificant difference between the means at the .05 ers,the think-aloud metacognitive strategy actu-
level ofsignificance). ally hindered reading comprehension (i.e.,8 of11
142 JOURNAL OF ADOLESCENT & ADULT LITERACY 51:2 OCTOBER 2007
Think-Aloud Strategy: Metacognitive development and monitoring comprehension in the middle school second-language classroom
Figure 1
Pre- and posttest scores of Level 2 students (N = 5)
35
30
25
20
s
e
r
o
c 15
S
10
5
0
1 2 3 4 5
Individual students
Pretest Posttest
Figure 2
Pre- and posttest scores of Level 3 students (N = 11)
35
30
25
s 20
e
r
o
Sc 15
10
5
0
1 2 3 4 5 6 7 8 9 10 11
Individual students
Pretest Posttest
JOURNAL OF ADOLESCENT & ADULT LITERACY 51:2 OCTOBER 2007 143
Think-Aloud Strategy: Metacognitive development and monitoring comprehension in the middle school second-language classroom
had lower posttest scores,2 had higher posttest understand text.Clarke (1980) posited that stu-
scores,and 1 remained the same (see Figure 3). dents reading in a second language are often
“short circuited”into becoming bottom-up read-
The findings suggest that while English
ers,and students in the Early Intermediate sam-
learners successfully use metacognitive strategies
ple group were clearly “stuck”at the word level
such as think-aloud,the efficacy ofthe strategies
due to decoding and vocabulary difficulties.
depends on the unique needs ofeach particular
Furthermore,Davis and Bistodeau (1993) stated
level ofproficiency as they approach the text.For
that English learners with a low level oflinguistic
example,each subgroup in this study presented
proficiency in English pay much greater attention
distinct and differing data:the Early Intermediate
to bottom-up components.Students in this sam-
students showed no measurable difference in pre-
ple group clearly followed this pattern.
and posttest mean scores,the difference in mean
scores for Intermediate students presented strong
The increase in the length ofthe second ar-
(although not statistically significant) evidence of
ticle may also be a reason for the lack ofprogress
think-aloud’s usefulness as a comprehension
on the posttest.Given the findings from earlier
strategy,and the statistically significant difference
studies,however,it makes sense that the sub-
in means scores ofthe Early Advanced students
group with the lowest level ofEnglish proficiency
contradicted the study’s hypothesis.
might not benefit from think-aloud because they
These heterogeneous outcomes appear to are “short circuited”into bottom-up reading
indicate that Early Intermediate English learners strategies,attempting to decode and make sense
may focus on bottom-up processes ofphonetic ofvocabulary and syntax while not able to imple-
decoding,whereas more proficient readers may ment a top-down metacognitive strategy such as
use background knowledge and inferencing to think-aloud in their second language.As a conse-
Figure 3
Pre- and posttest scores of Level 4 students (N = 11)
35
30
25
s 20
e
r
o
Sc 15
10
5
0
1 2 3 4 5 6 7 8 9 10 11
Individual students Pretest Posttest
144 JOURNAL OF ADOLESCENT & ADULT LITERACY 51:2 OCTOBER 2007
Think-Aloud Strategy: Metacognitive development and monitoring comprehension in the middle school second-language classroom
quence,the Think-Aloud Strategy may not be ap- traces generally tend to be weaker in the nonna-
propriate for beginning or Early Intermediate tive language.Think-aloud may have provided
English learners.The instructional implications the opportunity to strengthen memory traces in
ofthese findings suggest that reading instruction these second-language students’memories.
for English learners should focus primarily on
The Early Advanced English learner find-
vocabulary development,reading fluency,and
ings suggest that the Think-Aloud Strategy actu-
sentence structure.
ally disrupted the comprehension ofthis group of
The Intermediate English learners’success students.While somewhat surprising to us,extant
with the Think-Aloud Strategy indicates that they literature does suggest this possibility.Afflerbach
possess enough linguistic development in their and Johnston (1984) warned,“All probes (e.g.,
native language to effectively develop metacogni- think-aloud) can be considered disruptive to the
tive comprehension strategies.The findings of reading task at hand”(p.311).The findings for
this study suggest that,like their Early this subgroup confirm Afflerbach’s and
Intermediate peers,Intermediate students initial- Johnston’s assertion because use ofthink-aloud
ly become stuck at the word level;however,their in this study appeared to force Early Advanced
knowledge ofvocabulary and decoding ability English learners to regress to bottom-up reading
(fluency) has progressed enough in their second strategies.The increase in the length ofthe sec-
language and their linguistic threshold is high ond article may be related to the decline in scores
enough to allow them to become top-down read- ofthe Early Advanced students;however,the
ers with proper modeling and practice ofthe findings suggest that these readers already possess
strategy. metacognitive skills developed in their second
language—they did not need to make metacogni-
Cassanave (1988) stated that readers need to
tion covert through the use ofthink-aloud to in-
judge whether comprehension is taking place,
crease comprehension.
and that they need to stop and “decide whether
and how to take compensatory action when nec- Another possible explanation ofthink-
essary”(p.288).The Intermediate students,after aloud’s negative effect on this subgroup may lie in
applying this self-questioning technique to repair the fact that comprehending expository writing
their comprehension using think-aloud,were able requires readers to focus on the author’s message
to negotiate meaning in the expository text more and does not provide many opportunities for
successfully and apply Hosenfeld’s (1977) and personal connections (reflexive associations) with
Block’s (1986) successful comprehension strate- the text.In fact,making such personal connec-
gies such as keeping the meaning ofthe text in tions actually directs the reader’s attention away
mind during the reading,integrating ideas,read- from the information.This is an important point:
ing in “broad phrases,”and recognizing aspects of Pausing to make personal connections with the
text structure.In addition,the Intermediate stu- material,as the Early Advanced students were
dents may have been more successful in tapping asked to do with think-aloud,may distract stu-
prior knowledge (schema) without difficulties in dents from keeping the meaning ofthe text in
decoding or vocabulary as experienced by Early mind during reading.Many ofthe Early
Intermediate students.Although the length ofthe Advanced students in this study complained dur-
posttest article increased (as it did with the other ing think-aloud practice and asked the teacher,
two groups) students still achieved measurable “Why do I have to stop and say what’s going on if
gains in comprehension on the assessment. I already get it?”In fact,this group ofEnglish
Finally,Davis and Bistodeau (1993) related that learners was already able to “read in broad phras-
specific to L2,restatement ofcontent is useful in es,”to judge whether comprehension was taking
refreshing a reader’s memory because memory place,and to know how and when to stop and
JOURNAL OF ADOLESCENT & ADULT LITERACY 51:2 OCTOBER 2007 145