embedded learning strategy instruction: story …grammar and no improvement on basal questions (no...

WINNER OF CLD'S 2006 AWARD FOR OUTSTANDING RESEARCHThe following article was selected by the Research Committee of the Council for Leaming

Disabilities as the winner of the 2006 Award for Outstanding Research. Presentedannually, this award is designed to promote and recognize doctoral or master's level

research conducted within the last five years. Winners receive a certificate and a cashaward during the Distinguished Lecture at the International Conference on Leaming

Disabilities sponsored by the Council for Learning Disabilities.

EMBEDDED LEARNING STRATEGYINSTRUCTION: STORY STRUCTURE PEDAGOGY

IN HETEROGENEOUS SECONDARYLITERATURE CLASSES

Michael Faggella-Luby, Jean S. Schumaker, and Donald D. Deshler

Abstract. The effects of using the Embedded Story-Structure (ESS)Routine in a literature course were investigated. A heterogeneous groupof 79 ninth graders, including 14 students with LD, were randomlyassigned to one of two conditions, with instruction occurring in groupsof 12 to 14 students in general education literature classes over a nine-day period. ESS instruction focused on three reading strategies: (a) stu-dent self-questioning, (b) story-structure analysis, and (c) summarizing.Instruction for the alternative condition, called comprehension skillsinstruction (CSI), was comprised of a package of research-based readinginterventions. Statistically significant differences were found betweengroups in favor of the ESS Routine on measures of strategy use, story-structure knowledge, and unit reading comprehension. Moreover,results indicated equivalent gains for ESS students regardless of disabil-ity versus nondisability category.

MICHAEL FAGGELLA-LUBY, Ph.D., University of Gonnecticut, Genter for Behavioral Education and Research.JEAN S. SGHUMAKER, Ph.D., University of Kansas, Genter for Research on Leaming.DONALD D. DESHLER, Ph.D., University of Kansas, Genter for Research on Leaming.

Meeting adequate-yearly-progress (AYP) goals in read- of secondary students, including a high percentage ofing is a challenge for secondary scbool practitioners, students witb learning disabilities (LD), lack tbe readingRecently released National Assessment of Educational skills necessary to succeed in school and the world ofProgress (NAEP) data indicate that more than two-thirds work (Perie, Grigg, & Donahue, 2005).

Volume 30, Spring 2007 131

This record of student failure leaves educators insearch of evidence-based practices that can be imple-mented across secondary settings in an attempt to closethe performance gap that emerges between students' lit-eracy skills that have plateaued at the fourth-grade leveland the increasing academic grade-level demands(Deshler & Schumaker, 2006). Moreover, because manyadolescents with LD are taught in general education set-tings (IDEA, 2004) so they can earn credits toward grad-uation, language arts teachers are increasingly beingasked to shoulder a major part of the burden for com-prehension instruction, especially as it relates to narra-tive texts. These teachers need evidence-based methodsof delivering direct and explicit comprehension instruc-tion that provides the necessary skill development forstudents with LD and other low-achieving studentswhile simultaneously challenging high-achieving stu-dents in the same classroom.

One approach to defining instructional design anddelivery to improve reading comprehension was sug-gested by Kintsch (2004), who argued that the goal ofliteracy instruction should be to get students engaged inprocesses equivalent to those that expert readersemploy. One such process involves categorizing infor-mation in light of certain text structures. According toKintsch, student knowledge and use of text structurefavorably impacts comprehension, just as knowledge ofsyntax or vocabulary can. Text structure is believed tobe most relevant to the reading process during encodingand during the reader's organization of the text intohigh-order units. While syntactic and semantic instruc-tion foster sentence-level knowledge construction, dis-course-level structure construction can be improved byteaching genre-specific text structures. Thus, accordingto this model, instruction should explicitly introducestudents to the use of narrative text structure (or storystructure) to aid in the conceptual understanding ofnarrative texts.

Indeed, awareness of underlying story structure hasbeen shown to improve basic academic performanceand lead to higher-order thinking, including causal rea-soning (Gersten, Fuchs, Williams, & Baker, 2001).Narrative text structure has been investigated withinthree successive phases of research (Olson & Gee,1988). Early research centered on developing empiricalevidence to support a taxonomy of narrative elementsthat can be used to develop a basic understanding ofstory construction during encoding (e.g., Mandler &Johnson, 1977; Rumelhart, 1975; Stein & Glenn, 1978;Thorndyke, 1977). Overall, this research supported theexistence of various theoretical models of story struc-ture, surfacing universal components of a canonicalstructure for encoding (e.g., character, setting, conflict,resolution, and theme). Moreover, studies in this phase

of research suggested that variations in reader knowl-edge of story structure are related to reader recall, whichled to the second phase of story-structure research.

The second phase of research centered on develop-mental differences between types of readers with regardto the flexible use and complexity of story structure(e.g., Mandler & Johnson, 1977; Ouellette, Dagostino, &Carifio, 1998; Stein & Glenn, 1978; Whaley, 1981). Alimited number of these studies explored developmen-tal differences for individuals with LD (e.g., Montague,Maddux, & Dereshiwsky, 1990; Weaver & Dickinson,1982). Results from this line of inquiry indicated thatthere were developmental differences related to story-structure knowledge, with story-structure knowledgeappearing initially at the first-grade level and becomingwell developed at the sixth-grade level for typical read-ers. In one study focusing on students with LD(Montague et al., 1990), no main effect was found forgrade between 4th- through llth-grade students withLD, indicating a story-structure plateau similar to thegeneral reading achievement plateau reported byWarner, Schumaker, Alley, and Deshler (1980).

Due to the findings of developmental differencesacross groups of school-age readers, a third phase ofresearch has focused on the instruction of the previ-ously validated components of story structure toimprove reading comprehension. While some studieshave included elementary school students (e.g.,Garnine & Kinder, 1985; Fitzgerald & Spiegel, 1983;Griffey, Zigmond, & Leinhardt, 1988; Idol & Groll,1987; Short & Ryan, 1984), four relevant studies wereconducted at the secondary level, including students ingrades 6 through 11 in instruction of story-structurecomponents.

In the flrst study. Singer and Donlan (1982) used acontrol-group design with 27 eleventh-grade students(no students with disabilities) randomly assigned toinstructional groups of 15 over a six-day period. Story-grammar instruction was provided to the experimentalgroup. Results on the lone measurement instrument, a10-question multiple-choice quiz following each of sixstories, showed that the only difference between groupswas in favor of the experimental group on the quiztaken after the fifth story.

Gurney, Gersten, Dimino, and Carnine (1990) used amultiple-baseline design with seven high school stu-dents with LD in instructional groups of two or three inpuUout classes over a nine-week period. The authorsconcluded that, on 7- to 10-question story-grammarand basal-type quizzes given every other day, studentsshowed improvement on questions related to storygrammar and no improvement on basal questions (nomean scores were provided on either question type).However, no individual student graphs were shown nor

Learning Disability Quarterly 132

were group means reported, making it difficult to inter-pret the results.

In the third study, Dimino, Gersten, Carnine, andBlake (1990) conducted an experimental follow-up toGurney et al.'s study (1990) with 32 low-performingninth-grade students, including a subset of six studentswith LD and two students in Title I programs, over 19days. Students were randomly assigned to the groupsafter being matched for reading achievement. Resultson story-grammar and basal-type quizzes indicated sta-tistically significant differences in favor of the story-grammar instruction group related to scores onstory-grammar questions, hasal questions, theme ques-tions (a subset of story-grammar questions), and a sum-marizing task on the posttest and on the maintenancemeasure (mean raw score differences across all measuresranged from .3 [1,5% of total] to 2.5 [12.5% of total]).Results for the students with LD were not reported sep-arately from group means.

Drawing from the work of Dimino et al. (1990) andIdol and Groll (1987), Gardill and Jitendra (1999) con-ducted a multiple-baseline study with six middle schoolstudents with LD in instructional groups of two in pull-out classes. Results showed that (a) all students' scoreson story grammar and basal comprehension testsimproved concomitant with the onset of instruction,(b) all students' scores increased on generalization andmaintenance story-grammar tests from baseline and forfour students on hasal tests, and (c) five of the six stu-dents demonstrated an increase in the numher of storyelements when they orally retold stories over haseline.

In summary, when instruction in story structure wasimplemented, results across the studies cited above indi-cated (a) differences within or hetween groups on crite-rion-referenced story-grammar measures (i.e., answeringstory-grammar-related questions ahout a given passage)when instruction was multi-phased and at least fivehours in length; (h) inconsistent results on students'ahility to answer hasal (factual) questions; and (c) uncer-tain henefits for students with LD when taught in a het-erogeneous group.

While Gersten and colleagues (2001) are accurate intheir conclusion that instructing the awareness ofunderlying story structures has improved students' hasicacademic skills in some studies, the research is not com-plete. Several limitations are associated with the totalbody of research in this area, including hoth elementaryand secondary studies.

First, there is a paucity of research on heterogeneousgroups (or classes) of students that include studentswith LD, especially at the secondary level. Second, onlya few studies focus on even moderate teacher-to-studentratios, and these studies typically fail to disaggregatedata for students with LD, Hence, little is known ahout

the application of this type of instruction in larger class-room settings and outcomes for students with LD.Third, control-group design studies have involved onlylimited components of research-hased reading compre-hension instruction for the control group. Thus, thevalue added hy story-structure training cannot he deter-mined in comparison. Fourth, no secondary study hasconsidered the merits of packaging (or combining) ele-ments of story structure to intensify instruction. Fifth,because the measures in previous studies have heen cri-terion-hased related to isolated passages, limited infor-mation has emerged regarding such variahles as relatedknowledge (including knowledge of strategy compo-nents and literary terms), strategy use, written sum-maries of narrative passages, and student satisfactionwith instruction.

Given the status of the research on story structureoutlined above and the pressing need for secondary lit-eracy instruction to holster students' reading compre-hension, an instructional routine, the Emhedded StoryStructure (ESS) Routine, was designed for use in generaleducation classrooms serving heterogeneous popula-tions of students. The routine is hased on the findingsof Swanson's (1999) meta-analysis on effective compo-nents of explicit cognitive learning strategy instruction(i.e., advance organizer, skill review, demonstration,modeling, guided practice, independent practice, andcorrective feedhack), as well as the validated compo-nents of story structure (e.g., Mandler & Johnson,1977). This study investigated the effects of using theESS Routine in an inclusive ninth-grade literature classrelative to improving student use of reading compre-hension strategies, knowledge of strategy componentsand literary terms, comprehension of stories, and satis-faction with the routine.

METHODParticipants

Participants were 79 incoming ninth-grade studentsattending a summer school program for at-risk studentsat a private urhan high school in a southeastern state.At-risk status was determined hy school personnel inreading or mathematics hased on test scores fromEXPLORE (EXPLORE Test, 2005) taken during thespring of students' eighth-grade year. The student pop-ulation reflected heterogeneous characteristics, as somestudents exhihited limited reading comprehensionahility while others only exhihited limited abilities inmathematics (i.e., the latter group contained strongerreaders).

The researcher met with all parents and students priorto the experiment to solicit written permission for par-ticipation. Students with permission were subsequentlymatched according to four variahles.^ Memhers of each


Table 1Demographic Data on All Subjects

Category

Total number of subjects

GenderMaleFemale

Age (months)

Standard deviation

EthnicityCaucasianAfrican AmericanHispanicOther

Avg. EXPLORE Percentile (reading)

Frequency EXPLORE Scores80th - 89th Percentile70th - 79th Percentile60th - 69th Percentiie50th - 59th Percentile40th - 49th Percentile30th - 39th Percentiie20th - 29th Percentile10th - 19th Percentile

0 - 9th Percentile

ESSN %

39

22 56.417 43.6

171.44

4.63

30 77.06 15.41 2.62 5.1

47.97

167790423

CSIN

40

2119

172.08

6.25

29533

50.70

1688

120203

%

52.547.5

72.512.57.57.5

matched pair were randomly assigned to one of twogroups: (a) the embedded-story structure (ESS) groupor (b) the comprehension skills instruction (CSI)group. Students in each group were also assigned ran-domly to one of three classes, for a total of six classes(three in each condition). Matching variables were (a)disability diagnosis versus no diagnosis; (b) EXPLOREreading comprehension national percentile score(within six percentile points) (EXPLORE Test, 2005);(c) gender; and (d) age in months (within six months).(See Table 1 for demographic data for the ESS and CSIgroups.) Independent samples t-tests were employedto determine whether the groups were statisticallydifferent with regard to age and EXPLORE readingcomprehension national percentile. Results showedthat they were not, t{77) = .516, p = .608; t{77) = .609,p = .544, respectively.

An LD diagnosis was present for 14 subjects, based onthe school's existing psychological profiles. (Table 2provides demographic data related to these students inboth groups.) Full-scale IQ scores were available for 13of the 14 students on the Wechsler Intelligence Scalefor Children (Wechsler, 1991, 2003). The seven stu-dents in the ESS group earned a mean full-scale IQ scoreof 103.7 {SD = 7.2), whereas the six CSI students earneda mean IQ score of 104.0 (SD = 2.5). Independent sam-ples t-tests were employed to determine whether thegroups were statistically different with regard to age, IQscore, and EXPLORE reading comprehension nationalpercentile scores. Results showed that they were not,t{12) = -.050, p = .961; t(ll) = .093, p = .928; t{l2) =.315, p = .758, respectively.

SettingInstruction was provided in a typical classroom. The


total school population was 1,480, drawing studentsfrom across the city. Classroom desks were arranged infour rows. An overhead projector was positioned on atable at the front of the room, along with a large pres-entation screen and two side-by-side whiteboards.

Instructional MaterialsEight short stories and a folktale were selected as pas-

sages to be used during instruction and student practiceand as the basis for comprehension assessment for allstudents. The materials included renowned classics (e.g.,Poe's The Tell-Tale Heart) as well as more contemporarystories that present accounts of dilemmas encounteredby adolescents (e.g., Myers' The Treasure of Lemon

Brown). Stories were 9-18 pages long, with lexile scoresranging from 600 (The Tell Tale Heart) to 1220 {TheMonkey's Paw; Metametrics, 2000).

The Instructional ProgramsEmbedded story-structure routine. The Embedded

Story-Structure (ESS) Routine targeted three strategies:(a) self-questioning (used during pre-reading), (b) story-structure analysis (used during reading), and (c) sum-mary writing (used after reading). A graphic device, theESS Organizer, was designed to facilitate the interactiveconstruction of knowledge between students andteacher and the integration of the three strategies (seeFigure 1).

Table 2Demographic Data on Students with Leaming Disabilities

ESS

SI

S2

S3

S4

S5

S6

S7

CSl

SI

S2

S3

S4

S5

S6

S7

Age(months)

175

174

166

165

166

177

168M= 170.1

SD = 5.0

162

167

175

177

168

166

175M= 170.1

SD = 5.0

Gender

M

M

M

F

F

M

M

M

M

M

F

M

M

M

Ethnicity

C

O

ccccc

co

AA

unmarked

C

C

unmarked

Note. Ethnicity: C=Caucasian; AA=African American; O=other•Disability: ADD=Attention Deficit Disorder;

IQ

114

101

93

101

112

102

103M= 103.7SD = 7.2

108

mid range

105

104

102

101

104M= 104.0

SD = 2.5

EXPLORENat Per

17

29

29

51

42

42

66M=39.4

SD = 16.2

51

42

7

51

42

51

51M=42.1

SD = 16.1

D=Dysiexia; R=Reading Deficit; ULD=UnspecifiedlQ=Fuli Scaie IQ score on the WISC-IV or WlSC-III.

Disability*

ADD D R ULD

• •• • •• •• ••

••

• • •••••

••

Learning Disabiiity.


Figure 1. Graphic organizer for the ESS routine.

Name Title .Date

Who are the main characters?

Characters©Protagonist©AntagonistOther

Clues/Description

ttWhat is the Central Conflict? Person vs. PersonPerson vs. NaturePerson vs. IdealPerson vs. Self

How does tile central conflict begin? (Initiating Event)

OWhen does the story take place? (Time)

e does tiie story taiie place? (Place and Background info)

-Which decision or event is tiie ciiniax (or turning point)? (Climax)

^How does tile central conflict end/resolve? (Resolution)

(5)W/iy does the autiwr tell us tire story in this way? (Theme)

Story Structure Picture:

Summary:

The self-questioning strategy involves students askingand answering a series of seven questions initially intro-duced by the teacher that relate to eight critical compo-nents of story structure (e.g., main character, initiatingevent, time, place, central conflict, climax/turningpoint, resolution, and theme) (e.g., Dimino et al., 1990;Gardill & Jitendra, 1999). When implementing thisstrategy, students use seven individual question words(e.g., "who," "what," "when," "where," "which,""how," and "why") as a mnemonic device to help themrememher the eight components and the related sevenquestions. (See Tahle 3 for the story components andquestions.) Students record their answers to each self-question on the ESS Organizer.

The second strategy, story-structure analysis, involvesfilling in a Story-Structure Diagram on the back of theESS Organizer by labeling specific events from a shortstory. To help them remember this strategy, students areprovided with a second mnemonic device that connectseach critical element of the Story Structure Diagram toa unique picture cue. Students draw these cues whenlabeling the Story-Structure Diagram.

Students use the third strategy, a summary writing

strategy, to generate a written summary of the shortstory on the back of the ESS Organizer based on theiranswers to the self-questions using a four-sentence for-mula that includes each critical element of story struc-ture.

Comprehension skill instruction. ComprehensionSkill Instruction (CSI) was designed to parallel the ESSinstruction by targeting three research-based strategiesfor instruction: (a) the LINCS Vocabulary Strategy (Ellis,2000; used during prereading); (b) Question-AnswerRelationships (QAR) (Raphael, 1982, 1986; used duringreading); and (c) semantic summary mapping (Englert,Mariage, Garmon, & Tarrant, 1998; used after reading).The CSI strategies were chosen because they have previ-ously been studied (separately) as mechanisms thatmight improve student reading comprehension. To par-allel the ESS instruction, comparison students received agraphic device, called the CSI Organizer, to facilitate theinteractive construction of knowledge between studentsand teacher.^

The LINCS Vocabulary Strategy involves the use of aset of mnemonic strategies, including a key word strat-egy, a visual imagery strategy, and a story strategy to


link known information to new vocabulary words andtheir definitions (Ellis, 2000). Students record LINCSinformation related to a teacher-chosen vocabularyword on the first page of the CSI Organizer.

For the QAR strategy, students ask themselves andanswer text-based and knowledge-based questions todevelop story understanding. Text-based questionsinclude (a) "Right There" questions, which require stu-dents to find answers explicitly stated right there in thetext (e.g., "What is Johnny's father's job?"); and (b)"Think and Search" questions, which require locatinganswers stated explicitly in the text but require exami-nation of more than a single location in the text (e.g.,"What are some of the challenges facing the kidnap-pers?"). Students record QAR questions and answers onthe first page of the CSI Organizer.

Students use the third strategy, semantic summarymapping, to visually summarize stories by mapping self-identified critical components of the story. The strategyinvolves identifying and listing critical components ofthe story before organizing them into a connectivesemantic/concept map of related ideas on the back ofthe CSI Organizer.

Measurement InstrumentsStrategy-use test. The Strategy-Use Test was adminis-

tered to all students in both groups at pretest (Day 1),progress test (Day 5), posttest (Day 9), and mainte-nance (8 weeks after the posttest). The purpose of thismeasure was to provide evidence that the ESS studentslearned the ESS strategies that they were taught (andthat the CSI students did not learn the ESS strategies)so that their learning could later be related to studentperformance on reading comprehension measures.Each of four forms of the test included three sections,each section corresponding to use of one of the threeESS strategies. The first provided seven lines on whichstudents could write prereading questions related tonovel passages. For the second section, students wereasked to read a 260- to 300-word narrative passage (7.6-8.2 GE) from the Analytic Reading Inventory (ARI)(Woods & Moe, 2003) and correctly label a Story-Structure Diagram. (Each form of the test containeda different passage.) The third section on the testrequired students to write a summary of the passagethat was approximately 60 words in length.

In order to control for passage difficulty, a counter-

Table 3Component, Question, and Picture

Story-Structure Component

Main Character(protagonist and antagonist)

Central Conflict/Initiating Event

Time

Place

Climax

Resolution

Theme

Cue for ESS Instruction

Story-Structure Question

Who is the main character?

What is the central conflictand how does it begin?

When does the story take place?

Where does the story take place?

Which decision or event is theturning point?

How does the central conflictend/resolve?

Why did the author tell us thestory in this way?

Picture Cue

U / ^

®


balanced design was used. That is, Form A was adminis-tered to half the class in the pretest and the other halfof the class in the posttest, and vice versa for Form B.During the progress and maintenance tests. Form C wasadministered to half the class in the progress test andthe other half of the class in the maintenance test, andvice versa for Form D. All sections of the test were scoredusing an answer key. The total number of points avail-able was 24, with eight points each available for pre-reading questions, for correct diagram labels, and for awritten summary.

Knowledge test. The Knowledge Test, administered toboth groups, measured student knowledge of ESS strat-egy components and literary terms associated with theESS strategies to provide evidence that students in theESS group learned knowledge about the ESS strategiesbut students in the CSl group did not, so that this learn-ing could later be related to students' performance onreading comprehension measures.

The knowledge test was divided into three sections.The first provided a blank Story-Structure Diagram andasked students to demonstrate knowledge of the loca-tion of story-structure elements by labeling each com-ponent (e.g., initiating event, climax, resolution) on thediagram. The second section asked students to demon-strate knowledge of the picture cues and their connec-tions to the story-structure components by listing theappropriate elements of story structure along with thecorresponding picture cue (e.g., initiating event - i i ,time - ®). Finally, the third section required studentsto answer four open-ended questions related to basiclanguage arts knowledge (e.g., "What are the types ofconflict?" and "What are the components of character-ization?"). The total score for the Knowledge Test was30 points, with 8 points each for the appropriate labelsand correct picture cue answers and 14 points for thelanguage arts knowledge questions. An answer key wasused to score the test which was administered at pretestand posttest.

Unit comprehension test. The Unit Reading Compre-hension Test measured student retention of informa-tion related to the eight stories read by both groupsduring instruction. The hypothesis was that students inthe ESS group would retain more information than stu-dents in the CSl group. The test included 40 short-answer and fill-in-the-blank questions. One point wasawarded for each correct answer for a total of 40 points.

To create two forms of the test, a test bank of 80 ques-tions was created with eight sets of 10 questions perstory. Each set contained six questions related to thestory-structure elements of the story and four higher-order questions (two induction and two deductionquestions). Question sets were paired for each storyaccording to the difficulty level of story-structure ele-

ment, and one question from each pair was randomlyassigned to each form of the test. To ensure measure-ment reliability, a counterbalanced design was used dur-ing the pretest and posttest, with Form A administeredto half the class in the pretest and the other half of theclass in the posttest, and vice versa for Form B. Two cer-tified secondary English teachers juried the measure andagreed that all questions were appropriate to the corre-sponding passage, all answers could be found in the pas-sage, the answers in the answer key were correct, andthat Forms A and B of the test were equivalent in diffi-culty (100% agreement in all areas).

Satisfaction surveys. To evaluate student satisfactionwith the interventions, two social validity measureswere also administered. The Reading Satisfaction Surveymeasured students' satisfaction with their own reading.Questions related to students' satisfaction with readingand remembering short stories or novels, writing sum-maries, and test performance on reading comprehen-sion tests. The Satisfaction with Instruction Surveymeasured student satisfaction with the ESS Routine (orthe CSl Routine) and its components. Questions hererelated to student satisfaction with how helpful instruc-tion was with regard to improving story understanding,how easy it was to identify the specific elements of storystructure, whether or not students could use what theyhad learned independently, and overall satisfactionwith the routine. Qn both surveys, items were format-ted on a 7-point Likert-type scale ranging from"Completely Dissatisfied" to "Completely Satisfied."The average rating was calculated for each item for eachgroup.

Fidelity of treatment checklists. A fidelity checklistwas used to assess the quality of teacher performance inimplementing the specified instruction in each lesson.Each checklist was comprised of a list of 29-46 items,cataloging all the instructional practices associated withboth routines and intervention-specific components foreach routine appropriate to the given lesson, including(a) review of the previous lesson, (b) use of advance andpost-organizers, (c) encouraging independent strategyuse, (d) administering a quiz, (e) giving notes and defin-ing terms related to ESS or CSl instruction, (f) readingthe same stories, (g) periodically engaging students indialogue, (h) demonstration and modeling of the ESS-or CSI-related strategies, and (i) teacher-led practicewith ESS- or CSI-related strategies. An independentscorer and the researcher listened to randomly selectedaudio recordings of 22% of the instructional lessons foreach group and completed the checklist. The fidelityscore was calculated by totaling the number of observedinstructional behaviors for the intervention, the num-ber of unobserved instructional behaviors for the alter-native intervention, and the number of observed


Table 4Summary of Instructional Procedures

Contentof Instruction

*Whole strategy

1st halfof each substrategycomponents

2nd halfof each substrategycomponents

*Whole strategy

Day(s)

1

23

4

5

6-89

Four-Phase Pedagogy of Instruction

Demonstration Teacher-Guided Cooperative Independent& Modeling Practice Peer Practice Practice Review

• •

• •• •

• •

• •

• •• •

*An components of specific instructional condition strategies {i,e,, self-questioning, story-structure analysis, and summarizing for ESS; andQAR, LINCS, and semantic summary mapping for CSI),

instructional behaviors common to both interventions.The total score was then divided by the total number ofpoints possible, ranging from 29 to 46.

Interscorer reliability. Interscorer reliability wasdetermined by having two scorers independently scorea random sample (20%) of each of the instruments ateach point in the data-gathering plan for each group.An item-by-item analysis was used to compare observerrecordings. For the Strategy-Use Test, scorers agreed on1,443 out of 1,536 opportunities to agree (range from83.3% to 100%) for 94% agreement. For the KnowledgeTest, scorers agreed on 914 out of 960 opportunities toagree (range from 83.3% to 100%) for 95.2% agreement.For the Unit Reading Comprehension Test, scorersagreed on 1,233 out of 1,280 opportunities to agree(range from 83% to 100%) for 96% agreement. Forfidelity of treatment, scorers agreed on 272 out of 282opportunities to agree (range from 90% to 100%) for96.5% agreement.

ProceduresPretest and progress measure procedures. Pretesting

occurred in two sessions. The first battery of measureswas administered during a 90-minute block, andincluded the Strategy-Use Test (20 minutes). KnowledgeTest (15 minutes), and the Unit Reading Comprehen-

sion Test (25 minutes). The Reading Satisfaction Surveywas administered during the first 20 minutes of the firstlesson. Additionally, the Strategy-Use Test was repeatedduring the last 20 minutes of the fifth lesson as aprogress measure.

Instructional procedures. Teacher-to-student in-structional ratios were 1:12, 1:13, or 1:14, dependingon the class enrollment. The same teacher (and firstauthor), a special education doctoral student with cer-tification in secondary-level English, provided allinstruction to both groups. Instruction for each classtook place over a nine-day summer school session with90 minutes of instruction on Days 1 and 9 and 120minutes of instruction on Days 2 through 8, for a totalof 17 hours of instruction. Instruction took place infour phases: (a) teacher demonstration and modeling ofthe targeted strategies, including think-aloud problemsolving; (b) student-teacher collaboration and co-con-struction of knowledge and strategy use; (c) studentpeer collaboration (cooperative learning) and teacher-guided practice; and (d) independent student practice.The teacher used the cooperative learning and inde-pendent practice sessions as opportunities to circulatethrough the classroom to provide personal correctivefeedback to individual students. The teacher also pro-

Voluine 30, Spring 2007 139

vided individual instruction at the teacher's desk whenneeded by a student. All lessons hegan with a review ofthe previous day's lesson and an advance organizer ofthe current day's instruction. All lessons ended with ahrief review of the day's lesson.

Instruction was delivered with consideration given totwo guiding principles: (a) the content of instruction(i.e., components of the three targeted strategies foreach instructional condition); and (h) the four-phasepedagogy of instruction described above. (See Table 4for a summary of instructional procedures.) The instruc-tional sequence for the ESS and CSI groups was thesame; only the content (the three strategies specified foreach group) differed.

Day 1 instruction involved teacher introduction, de-monstration, and modeling of all strategy componentswhile using the appropriate organizer to provide stu-dents with an overall "big picture" example of how thethree targeted strategies were to be used in combinationin relation to the first story. At the end of Day 1 instruc-tion, the teacher prompted the students to participatein using the strategies in relation to the end of the storyin a whole-group teacher-guided practice activity.

The first half of Day 2 instruction focused on teacherdemonstration and modeling of selected parts of two ofthe strategies in relation to Story #2. During the secondhalf of the period, the students were prompted to usethe strategies in a whole-group teacher-guided practiceactivity, also for Story #2.

On Day 3, students engaged in cooperative peer prac-tice of the strategy components covered on Day 2 forthe initial sections of Story #3 and independent prac-tice for the remaining sections of the story.

Day 4 instruction mirrored Day 2 instruction usingStory #4, but focused on the remaining strategy com-ponents that were not covered on Days 2 and 3.

On Day 5, students engaged in cooperative peer prac-tice of all three targeted strategies for some sections ofStory #5 and independent practice of all three strategieson other sections.

Days 6 through 8 again allowed for periods of peercollaboration and periods of independent practice ofthe combined strategies with sections of a new storyeach day and with students slowly transitioning tolonger periods of independent work.

The final session (Day 9) concluded with a class-widereview of the complete unit of eight short stories cov-ered. The teacher followed a written protocol for eachlesson and concluded each class with a class-wide dis-cussion of the day's short story. Students handed intheir completed graphic organizers periodically to theteacher for corrective feedback. (For more details aboutthe instructional procedures for each group, seeFaggella-Luby, 2006.)

Posttest and maintenance procedures. Posttestingoccurred in identical fashion to the pretesting with twogroup-administered sittings. During the first sitting, stu-dents completed the satisfaction surveys. In the secondsitting, 90-minutes was provided for the battery ofmeasures to be administered. For the maintenanceprobe, the Strategy-Use Test was administered eightweeks after the final lesson.

Design and Data AnalysesThe study employed a control-group design with ran-

dom assignment of members of matched pairs of stu-dents to two groups (ESS and CSI) to determine theeffects of the ESS Routine. The same teacher conductedall instruction to control for teacher effects. Instruc-tional time, stories, and the general instructionalmethodology were identical across the groups.

For the Strategy-Use measure, which was adminis-tered four times, a one-way analysis of variance(ANOVA) with one within-subjects factor and twobetween-subjects factors was conducted. Time with fourlevels (pretest, progress, posttest, and maintenance) wasthe within-subjects factor (independent variable).Instructional condition (ESS and CSI) and whether ornot students had a diagnosed disability (students withdisabilities and other peers) were the between-subjectsfactors.

For the Unit Reading Comprehension Tests andKnowledge Tests, which were each administered twice,a one-way ANOVA with one within-subjects factor andtwo between-subjects factors was conducted. Time withtwo levels (pretest and posttest) was the within-subjectsfactor (independent variable). Instructional condition(ESS and CSI) and whether or not students had a diag-nosed disability (students with disabilities and otherpeers) were the between-subjects factors.

To determine between-group differences on the UnitReading Comprehension Tests and Knowledge Tests,independent-samples t-tests were completed separatelyfor each administration of a measure. A BonferroniCorrection was used to control for Type 1 error (a =.05/2 = .025). For the Strategy-Use Test, to determinewithin-group differences, paired-samples t-tests werecompleted separately for the ESS and CSI groups to com-pare the pretest to posttest scores, the pretest to main-tenance scores, and the posttest to maintenance scores.Again, a Bonferroni Correction was used to control forType I error for all pairwise comparisons (a = .05/(2x3)= .0083).

RESULTSStrategy Use Test

Means and standard deviations for Strategy-Use Testscores are presented in Table 5 for students with disabil-


ities and the remaining students in the ESS and CSIgroups. The ANOVA results indicated no statistically sig-nificant time by condition by disability diagnosis inter-action, Wilks' A = .972, F(3,73) = .700, p = .555, multi-variate Ti = .028. However, a statistically significanttime-by-condition interaction was found, Wilks' A =.330, f(3,73) = 49.5, p <. 001, multivariate ^^ = .670,indicating a large effect size. Thus, regardless of whetheror not students had a disability diagnosis, ESS subjectsexhibited equivalent gains on this test.

Results of an independent samples t-test indicated nostatistically significant difference between the ESS andCSI students' performance on the strategy-use pretest:(77) = -.293, p = .770, d = .015. However, there were sta-

tistically significant differences between the groups, infavor of the ESS group, on all three remaining tests: theprogress test, t{S6.4) = -14.6, p < .001, d = .738; theposttest, t(60.5) = -15.9, p = <.OO1, d = .807, and themaintenance test, t(59.2) = -10.7, p = <.OO1, d = .542,indicating large and moderate effect sizes, respectively.

Results of the three paired-samples t-tests revealed sig-nificant differences on two comparisons for the ESSgroup: pretest to posttest, t(38) = -18.3, p < .001, A =4.27; and pretest to maintenance, t(38) = -10.84, p <.001, A = 3.70, indicating large effect sizes. There was nosignificant difference between the posttest and the

maintenance test, t{38) = 2.06, p = .046, A = .41. No sig-nificant differences were found for all three pairwisecomparisons for the CSI group: t(39) = -1.28, p = .210,A = .21; t(39) = .110, p = .913, A = .02; and t{39) = 1.38,p = .179, A = .30, respectively.

Knowledge TestMeans and standard deviations for the Knowledge

Test scores are presented in Table 6. The ANOVA resultsagain indicated no statistically significant time by con-dition by disability diagnosis interaction, Wilks' A = .99,f(l,75) = .138,p = .711, multivariate ri = .002. However,there was a statistically significant time-by-conditioninteraction, Wilks' A = .376, f(l,75) = 124.4, p < .001,multivariate r\^ = .624, indicating a large effect size.

Results of an independent samples t-test indicated nostatistically significant difference between the ESS andCSI groups on the pretest: t(77) = -.890, p = .376, d =.045. However, there was a statistically significant dif-ference between the groups on the posttest: t(77) =-4.11, p < .001, d = .208, indicating a small effect size infavor of the ESS group.

Unit Reading Comprehension TestMeans and standard deviations for the Unit Reading

Comprehension Test scores are presented in Table 7.ANOVA results indicated no statistically significant timeby condition by disability diagnosis interaction, Wilks'

Table 5Strategy-Use

ESS

SWD*

Other

CSI

SWD*

Other

Test Results

N

7

32

7

33

*SWD=Students with disabilities.

PretestMean

6.86

6.09

6.43

6.00

SD

0.69

2.49

2.15

2.51

ProgressMean SD

18.14

16.25

7.00

6.06

3.13

4.08

2.08

2.01

PosttestMean SD

15.93

16.02

5.71

6.76

2.83

3.32

1.60

1.87

MaintenanceMean SD

17.00

14.19

5.43

6.15

3.37

4.51

2.44

2.49


Table 6Knowledge Test Results

ESS

SWD*

Other

CSl

SWD*

Other


JV

7

32

7

33

PretestMean

6.57

4.19

3.86

4.70

SD

2.57

2.79

2.41

2.77

PosttestMean

23.71

22.47

5.43

6.33

SD

5.96

6.20

2.37

3.04

A = .992, f (1,75) = .621, p = .433, multivariate rf = .008.However, again there was a statistically significant time-by-condition interaction, Wilks' A = .908, F{1,75) =7.61, p = .007, multivariate yf = .092, representing amedium effect size. Thus, regardless of whether or notstudents had a disability diagnosis, the ESS subjects

exhibited equivalent gains on the Unit ReadingComprehension Test.

Results of an independent-sample t-test comparingthe pretest scores of the two groups indicated no differ-ence between the groups: t{77) = -.104, p = .917, d =.005. However, there was a statistically significant dif-

Table 7

Unit Reading Comprehension

ESS

SWD*

Other

CSl

SWD*

Other


N

7

32

7

33

Test Results

PretestMean

2.00

2.09

1.14

1.72

SD

2.83

2.18

2.61

2.20

PosttestMean

21.86

24.44

17.14

19.67

SD

3.67

5.17

6.15

5.06


Table 8Reading Satisfaction Survey

Ql

Q2

Q3

Q4

Q5

Q6

Q7

Q8

GrandMean

Pretest

M

4.64

4.59

4.05

4.72

5.58

5.79

5.17

4.68

4.89

Note. Item score can range from 1

SD

1.56

1.50

1.64

1.40

1.27

1.26

1.21

1.34

0.92

to 7.

Results

ESSPosttest

M

5.90

6.15

6.13

6.15

6.18

5.77

5.85

6.05

6.02

SD

0.99

0.90

1.11

1.11

0.97

1.14

0.90

0.94

0.73

Pretest

M

4.65

4.28

4.18

5.15

5.68

5.25

4.75

4.3

4.78

SD

1.27

1.40

1.80

1.23

1.19

1.41

1.52

1.22

0.99

CSIPosttest

M

6.15

6.12

5.79

6.08

6.30

5.77

6.03

6.07

6.04

SD

0.83

0.94

1.20

0.96

0.79

0.99

0.92

0.86

0.65

ference between the groups on the posttest: t(54.4) =-15.3, p < .001, d = .776, in favor of the ESS group. Thisrepresents a large effect size.

SatisfactionReading satisfaction survey. Means and standard

deviations for the Reading Satisfaction survey scores arepresented in Table 8. Results from an independent sam-ples t-test indicate no statistically significant differencesbetween the groups on either the pretest or the posttestgrand mean ratings: t{77) = -.516, p = .607, d = .03; andt(77) = .114, p = .909, d = .006, respectively. Two follow-up paired samples t-tests were conducted to examinewithin group differences from pretest to posttest. ABonferroni Correction was used to control for Type Ierror for both pairwise comparisons (a = .025). Resultsindicated statistically significant differences and largeeffect sizes for both the ESS, f(38) = -8.62, p < .001, A =1.23, and the CSI, t{39) = -9.62, p < .001, A = 1.23,groups between pretest and posttest ratings, with highermean scores on the posttest indicating higher levels ofreading satisfaction than at pretest.

Satisfaction with instruction survey. Means and stan-dard deviations for ratings on the Satisfaction withInstruction Survey are presented in Table 9. A grandmean rating of 6.27 {SD = .54) was calculated for the ESSstudents, and a grand mean rating of 6.26 (SD = .54) wascalculated for the CSI students. An independent sam-ples t-test revealed no statistically significant differencebetween groups' satisfaction ratings, t{77) = -.053, p =.958, d = .003.

Fidelity of TreatmentThe teacher completed a mean of 97% of the instruc-

tional steps across the analyzed lessons for the ESSRoutine with a range from 90% to 100% for individualESS lessons. A mean of 98% of the instructional stepsacross the analyzed lessons was completed for the CSIIntervention with a range from 94% to 100% for indi-vidual CSI lessons. No content components of the CSIintervention were taught during the ESS intervention.Likewise, no components of the ESS intervention weretaught during the CSI intervention.


DISCUSSIONThe methodological and theoretical presuppositions

inherent in this study's design are directly related to theearliest work surrounding story-structure instruction.The study was designed to answer the question: Canstory-structure components be taught to heterogeneousgroups of learners, in general education settings, toimprove reading comprehension without sacrificing thelearning of the higher achieving peers?

The results are similar to those of previous studiesregarding (a) positive effects of multi-phase instruction

with moderate-sized groups including a variety oflearner types and using authentic literature with sec-ondary students (e.g., Dimino et al., 1990; Gardill &Jitendra, 1999; Idol, 1987); and (b) improved perform-ance on criterion-referenced story grammar measures(e.g., Dimino et al., 1990; Gardill & Jitendra, 1999; Idol,1987).

In addition, this study extends previous research andcontributes to the field in several important ways. First,it supports Kintsch's (2004) suggestion to instruct com-ponents of narrative text structure to improve reading

Table 9Posttest Satisfaction with

Ql

Q2

Q3

Q4

Q5

Q6

Q7

Q8

Q9

QIO

Q l l

Q12

Q13

Q14

Q15

Q16

GrandMean

Note. Item scores can range from 1 to 7.

Instruction Survey

M

6.46

6.28

5.97

6.24

6.46

5.79

5.78

5.90

6.13

3.01

6.36

6.54

6.61

6.72

6.59

6.53

6.27

ESSSD

0.76

0.79

1.04

1.05

0.76

1.07

0.95

0.94

0.92

0.95

0.90

0.79

0.64

0.76

0.68

0.60

0.54

Results

N

39

39

39

38

39

38

39

39

39

39

39

39

38

39

39

38

39

M

6.58

6.50

5.67

6.02

6.33

5.80

6.13

6.07

6.18

6.05

6.13

6.37

6.50

6.83

6.53

6.55

6.26

CSISD

0.64

0.93

1.02

1.07

0.89

1.07

1.01

1.07

0.98

1.18

1.02

0.74

0.72

0.50

0.64

0.68

0.54

N

40

40

40

40

40

40

39

40

40

40

40

40

40

40

40

40

40


comprehension. Second, it furthers investigations ofoutcomes for heterogeneous groups of learners (includ-ing individuals with disabilities). Moreover, while rela-tively modest in scope, the study is the largest of its kindto date, involving 79 students, including 14 studentswith disabilities, in a control-group design. Third, thestudy extends the research into secondary or adolescentliteracy, a growing but minimally explored area ofstudy. Fourth, the study demonstrates that, unlike pre-vious studies of story-structure instruction in which theinclusion of individuals with disabilities requiredinstructional ratios of 3:1 (e.g., Gardiil & Jitendra, 1999;Gurney et al., 1990) to produce gains, students can betaught successfully in instructional groupings of 14:1.Fifth, this is the first study related to teaching storystructure that used research-based reading comprehen-sion instruction in the comparison condition, providingsubstantial support to claims that differences wererelated to the benefits of the ESS Routine over anotherstrong instructional model. Sixth, this study providespreliminary evidence about the effects of packaging ele-ments of story structure to intensify instruction by com-bining student self-questioning, story mapping, andsummarizing. Finally, the study adds three new assess-ment tools to the research on story structure, whichenable measurement of intervention efficacy: (a) a strat-egy-use test, (b) a knowledge test, and (c) a unit test overall of the short stories covered during the unit. The strat-egy measures demonstrate whether or not students havelearned and can use the strategies that have beentaught. Such a demonstration helps to eliminate thealternative explanation that reading practice alone pro-duced the differences in reading comprehension.

Implications for PracticeStudy results support the use of the ESS Routine as an

instructional intervention for improving reading com-prehension of at-risk learners in heterogeneous inclu-sive secondary classrooms. Thus, the Strategy-Use Testand Knowledge Test results indicated that studentsreceiving the ESS Routine instruction not only outper-formed CSl students, but gains were equivalent regard-less of the disability/no disability category. Moreover,participation in the routine led to significant improve-ments in strategy use from pretest to posttest, and frompretest to maintenance. No such trend existed for CSlstudents, whose Strategy-Use Test scores remained lin-ear and low.

Results from the Unit Reading Comprehension Testindicated, once again, that students, including thosewith disabilities, benefited equally from use of the ESSRoutine. Significant differences between students in theESS group and the CSl group on the reading compre-hension measure and the Knowledge Test are particu-

larly significant, given their similar nature to criterion-referenced assessments common in language artsclassrooms. These findings support the belief thatinstruction in the ESS Routine may bolster educationaloutcomes for individuals with disabilities in such class-rooms.

The socially significant findings across all surveymeasures are positive indicators of the palatability ofthe ESS Routine. More specifically, differences withinboth groups from pretest to posttest scores for theReading Satisfaction Survey indicate that, regardless ofinstruction type, students were more satisfied with theirreading abilities after instruction. Further, student sur-vey scores on the Satisfaction with Instruction Survey inboth groups indicated not only overall satisfaction withthe instruction, but also a feeling that this instruction:(a) helped them understand short stories, (b) was funand interesting, and (c) would lead to likely independ-ent use of the strategies. Again, these results are positiveindicators to support the likelihood of instructionalpalatability.

Finally, an important aspect of this study design isthat ESS condition scores on strategy-use and knowl-edge test measures when evaluated in light of scores onthe unit reading comprehension test provide strong evi-dence to suggest that a relationship exists between stu-dent use and knowledge of ESS strategies and growth inreading comprehension.Limitations

Several limitations apply to this study. First, theteacher for the study was also the researcher. While thisarrangement controlled for teacher effects, the effectsthat may be produced by other teachers are unknown,as are other teachers' satisfaction with the routine andtheir ability to implement it.

Second, ESS Routine efficacy was measured in relationto heterogeneous groups of students, in part becausefederal law continues to advocate the need to educateindividuals with disabilities in the least restrictive envi-ronment. The effects of the ESS Routine when imple-mented in supportive learning environments and underthe optimal conditions of effective learning strategyinstruction described by Swanson (2001) are unknown.

Third, the sample population included only 14 stu-dents with disabilities. While this percentage of thestudy population (18%) is larger than the reportedprevalence rate of 6% of students with LD in readingfound nationally in schools (National Research Centeron Learning Disabilities, 2002), this small number ofstudents with LD limits the generalizability of theresults with regard to this type of student. Moreover,the subjects were all students referred to a summerschool program.


Fourth, as a group, although the ESS student per-formance on the Unit Reading Comprehension Testwas significantly higher than the performance of theCSl group, the level of their performance on theposttest may not be considered educationally signifi-cant (M = 24.0 or 60%, SD = 4.99). This is partially theresult of rigorous application of a scoring rubric and theinitial design of a difficult test to ensure against a ceil-ing effect. A second consideration is that the questiontype was short-answer, proving more challenging thanthe multiple-choice or fill-in-the-blank format com-monly used in schools. Though the total number ofhours of instruction (17) was typical for demonstratingeffective learning strategy instruction (e.g., Deshler etal., 2001), the relatively short number of instructionaldays (9) may have produced an additional cognitivechallenge for the specific student population. That is,complex packages of strategies (e.g., self-questioning,story mapping, summary writing) require sufficienttime to master. Perhaps additional practice sessionswould have yielded improved results.

Finally, a significant challenge to all the research todate on story-structure instruction is the lack of stan-dardized reading comprehension measures. This studyis no exception. In part, no such measure was usedbecause of the relatively short amount of instructionaltime and the prediction that scores on a standardizedreading test would not have a chance of changing afteronly nine sessions of instruction.

Future ResearchFuture research should include studies that involve

implementing the ESS Routine by a teaching cadreother than the researcher across multiple classes andover a longer period of time. A second future directionis to implement the ESS Routine in a supportive learn-ing environment for students with LD using the eightstages of instruction from the Strategic InstructionModel's Learning Strategies Curriculum (see Deshler &Schumaker, 2006) to allow maximum interventioneffectiveness to be measured. Another study mightfocus on a component analysis of the routine.

All of these studies might involve standardized read-ing comprehension measures. Such research will beimportant to the field of special education and studentswith disabilities because it will lead to the developmentof successful pedagogy that is appropriate for all levelsof learners in inclusive classrooms.

REFERENCESCarnine, D., & Kinder, D. (1985). Teaching low-performing stu-

dents to apply generative and schema strategies to narrative andexpository material. RASE, 6(1), 20-30.

Deshler, D. D., Schumaker, J., Bulgren, J., Lenz, K., Jantzen, J. E.,Adams, G., Carnine, D., Grossen, B., Davis, B., & Marquis, J.(2001). Making learning easier: Connecting new knowledge to

things students already know. Teaching Exceptional Children,33(4), 82-85.

Deshler, D. D., & Schumaker, J. (Eds.). (2006). Teaching adolescentswith disabilities: Accessing the general education curriculum.Thousand Oaks, CA: Corwin Press.

Dimino, J., Gersten, R., Carnine, D., & Blake, G. (1990). Storygrammar: An approach for promoting at-risk secondary stu-dents' comprehension of literature. The Elementary SchoolJoumai, 91(1), 19-32.

Ellis, E. S. (2000). The LINCS vocabulary strategy (2nd ed.).Lawrence, KS: Edge Enterprises.

Englert, C. S., Mariage, T. V., Garmon, M. A., & Tarrant, K. L.(1998). Accelerating reading progress in early literacy projectclassrooms: Three exploratory studies. Remedial and SpecialEducation, 19(i), 142-159.

EXPLORE Test. (2005). Iowa City, IA: ACT, Inc.Faggella-Luby, M. (2006). Embedded learning strategy instruction:

Story-structure pedagogy in secondary classes for diverse learners.Unpublished doctoral dissertation. University of Kansas.

Fitzgerald, J., & Spiegel, D. L. (1983). Enhancing children's readingcomprehension through instruction in narrative structure.Journal of Reading Behavior, 15, 1-17.

Gardiil, M. C, & Jitendra, A. K. (1999). Advanced story map in-struction: Effects on the reading comprehension of students withlearning disabilities. Journal of Special Education, 33(1), 2-17.

Gersten, R., Fuchs, L. S., Williams, J. P., & Baker, S. (2001).Teaching reading comprehension strategies to students withlearning disabilities: A review of research. Review of EducationalResearch, 71(2), 279-320.

Griffey, Q. L., Jr., Zigmond, N., & Leinhardt, G. (1988). The effectsof self-questioning and story structure training on the readingcomprehension of poor readers. Learning Disabilities Research,4(1), 45-51.

Gurney, D., Gersten, R., Dimino, J., & Carnine, D. (1990). Storygrammar: Effective literature instruction for high school stu-dents with learning disabilities. Journal of Learning Disabilities,23(6), 33-343.

Individuals with Disabilities Education Improvement Act. (2004).P. L. 108-446.

Idol, L. (1987). Group story mapping: A comprehension strategyfor both skilled and unskilled readers. Journal of LearningDisabilities, 20(4), 196-205.

Idol, L., & CroU, V. J. (1987). Story mapping training as a means ofimproving reading comprehension. Learning Disability Quarterly,10, 214-229.

Kintsch, W. (2004). The construction-integration model of textcomprehension and its implications for instruction. In R. B.Rudell & N. J. Unrau (Eds.), Theoretical models and processes ofreading (5th ed., pp. 1270-1324). Newark, DE: InternationalReading Association.

Mandler, J. M., & Johnson, N. S. (1977). Remembrance of thingsparsed: Story structure and recall. Cognitive Psychology, 9(1), 111-151.

MetaMetrics. (2000). The Lexile framework for reading. Durham,NG: Author. Retrieved September 15, 2004, from http://lexile.com/faq.htm.

Montague, M., Maddux, G. D., & Dereshiwsky, M. I. (1990). Storygrammar and comprehension and production of narrative proseby students with learning disabilities. Journal of LearningDisabilities, 23(3), 190-197.

National Research Center on Learning Disabilities (NRGLD).(2002). Researcher consensus statement. Information DigestNumber 3. Retrieved November 19, 2005, from http://www.nrcld.org/html/information/articles/digests/digest3.html


Olson, M. W., & Gee, T. C. (1988). Understanding narrative: Areview of story grammar research. Childhood Education, 64(5),302-306.

Ouellette, G., Dagostino, L., & Carifio, J. (1998). The relationshipof children's expectations for structure in story to their knowl-edge of literature and their reading ability. Reading Improvement,35(3), 98-113.

Perie, M., Grigg, W., & Donahue, P. (2005). The nation's report card:Reading 2005 (NCES 2006-451). U.S. Department of Education,National Center for Education Statistics. Washington, DC: U.S.Government Printing Office.

Raphael, T. (1982). Question-answering strategies for children. TheReading Teacher, 36(2), 186-190.

Raphael, T. E. (1986). Teaching question-answer relationships,revisited. The Reading Teacher, 39(6), 516-522.

Rumelhart, D. E. (1975). Notes on a schema for stories. In D. G.Bobrow & A. M. Collins (Eds.), Representation and understanding(pp. 211-236). New York: Academic Press.

Short, E. J., & Ryan, E. B. (1984). Metacognitive differencesbetween skilled and less skilled readers: Remediating deficitsthrough story grammar and attribution training. Joumai ofEducational Psychology, 76(2), 225-235.

Singer, H., & Donlan, D. (1982). Active comprehension: Problem-solving schema with question generation for comprehension ofcomplex short stories. Reading Research Quarterly, 166-185.

Stein, N. L., & Glenn, C. G. (1978). An analysis of story composi-tion on elementary school children. In ?? Friedle (Ed.), Newdirections in discourse processing (Volume 2, pp. 53-120).Norwood, NJ: Ablex.

Swanson, H. L. (1999). Instructional components that predicttreatment outcomes for students with learning disabilities:

Support for a combined strategy and direct instructionalmethod. Leaming Disabilities Research & Practice, 14(3), 129-140.

Thorndyke, P. W. (1977). Cognitive structures in comprehensionand memory of narrative discourse. Cognitive Psychology, 9, 11-110.

Warner, M. M., Schumaker, J. B., Alley, G. R., & Deshler, D. D.(1980). Learning disabled adolescents in public schools: Arethey different from other low achievers? Exceptional EducationQuarterly, 1(2), 27-35.

Weaver, P. A., & Dickinson, D. K. (1982). Scratching below the sur-face structure: Exploring the usefulness of story grammars.Discourse Processes, 5(3), 225-243.

Wechsler, D. (1991). Wechsler Intelligence Scale for Children (3rded.). New York: Psychological Corporation.

Wechsler, D. (2003). WISC-IV technical and interpretive manual. SanAntonio, TX: Psychological Corporation.

Whaley, J. F. (1981). Readers' expectations for story structures.Reading Research Quarterly, 1(9), 107-111

Woods, M. L., & Moe, A. J. (2003). Reader's passages to accompany:Analytic Reading Inventory. Upper Saddle River, NJ: MerrillPrentice Hall.

NOTES'Please note that there were 79 students who had permission toparticipate; thus, one student did not have a matched partner.After all the other matches were made, the 79th student wasassigned to the CSI condition by a coin flip.^ Comprehension Skills Organizer available upon request.

Correspondence should be addressed to: Michael Faggella-Luby,University of Connecticut, 249 Glenbrook Rd., Unit 2064, Storrs,CT 06269; e-mail: [email protected]

Volume 26, Winter 2003 147

embedded learning strategy instruction: story …grammar and no improvement on basal questions (no...

Documents