Hammill Institute on Disabilities

The Effects of Instruction in an Inference Strategy on the Reading Comprehension Skills of

Adolescents with Disabilities
Author(s): Nanette Salim Fritschmann, Donald D. Deshler and Jean Bragg Schumaker
Source: Learning Disability Quarterly, Vol. 30, No. 4 (Fall, 2007), pp. 245-262
Published by: Sage Publications, Inc.
Stable URL: http://www.jstor.org/stable/25474637 .
Accessed: 14/06/2014 21:00

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Sage Publications, Inc. and Hammill Institute on Disabilities are collaborating with JSTOR to digitize,
preserve and extend access to Learning Disability Quarterly.

http://www.jstor.org

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
 THE
EFFECTS OF INSTRUCTION IN AN
INFERENCE STRATEGY ON THE READING
COMPREHENSION SKILLS OF ADOLESCENTS
WITH DISABILITIES

Nanette Salim Fritschmann, Donald D. Deshler, and Jean Bragg Schumaker

Abstract. The purpose of this study was to determine the effects

of teaching eight secondary students with disabilities, including
seven with learning disabilities, a strategy for answering a variety
of inferential questions. A multiple-baseline across-subjects design
was employed. Outcome measures included scores on researcher
devised comprehension quizzes, a standardized test of reading
comprehension, a strategy use test, a strategy test, and
knowledge
a reading satisfaction measure. Fidelity of implementation,
instructional time, and maintenance of skills were also measured.
Results suggest that students with disabilities can learn to use a
strategy to answer a variety of inferential and
questions, mastery
of its use can result in improved scores on criterion-based and
standardized measures of reading comprehension. In addition,
students' satisfaction with their reading improved.

NANETTE SALIM FRITSCHMANN, Ph.D., Lehigh University.

DONALD D. DESHLER, Ph.D., University of Kansas.
JEAN BRAGG SCHUMAKER, Ph.D., Edge Enterprises, Inc., Lawrence, Kansas.

The current educational climate and its calls for national standardized assessment exams. For example,
increased skill acquisition and rising performance the proposed 2009 National Assessment of Educational
demands are requiring students to learn higher-order framework reflects
Progress (NAEP) reading expansion
reading skills, like inference skills (e.g., American from its current 1992-2007 framework, to include the
Institute for Research, 2005; Partnership for 21st assessment of broader reading content and deeper cog
Century Skills, 2006). With few exceptions, all second nitive processes (American Institute for Research,
ary students, including students with disabilities 2005). This framework represents a shift from assessing
(SWD), are required to take rigorous state reading com skills at the literal/word level of reading comprehen
petency exams, most of which involve the use of infer sion to assessing higher-order skills within reading
ence skills. At present, 26 states administer exit exams, that emphasize and inte
comprehension "interpreting
and 19 of them withhold diplomas based on poor per matter, the very skills required to
grating" reading
formance on the exit exams (Center on Education make inferences.

Policy, 2005). This climate of increased reading demands in schools

Increased local demands appear to be rising in tan and on tests poses significant for
challenges struggling
dem with the level of reading skills evaluated by adolescent readers. For students who have a disa

Volume 30, Fall 2007 245

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
bility, increased demands are especially problematic "undergo integration, which results in a well-struc
(Bulgren, Marquis, Deshler, Schumaker, & Lenz, 2006; tured" (Kintsch, 1998, p. 95) understanding of the text.
Schumaker, Deshler, Bui, & Vernon, 2006). Some In the case of drawing inferences from text, Kintsch pos
research has shown that students with learning disabil tulated that various factors (e.g., text features, language
ities (LD) enter seventh grade reading, on average, at skills, and domain knowledge) contribute to compre
the fourth-grade level, and they do not make gains in hension and assist the reader in integrating information
reading achievement as they progress through the sec into a meaningful structure. When such integration
ondary grades (Deshler & Schumaker, 2006; Deshler et occurs, the reader is able to draw a successful inference
al., 2006; Warner, Schumaker, Alley & Deshler, 1980). related to the text. Also, according to Kintsch, this
Further, large proportions of these students are failing process may be either automatic (unconscious) or con
their state reading competency exams (Heubert, 2002), trolled(conscious and strategic).
as well as tests in their required high school courses Research conducted with elementary-level students
(Bulgren, Schumaker, & Deshler, 1988; Hughes, Deshler, lends support to Kintsch's theory. For example, some
Ruhl, & Schumaker, 1993; Wagner et al., 2003). studies have shown that the scores of poor comprehen
This is understandable, because, although some of ders on inferential comprehension questions improve
them have acquired some basic decoding skills (Catts, when they receive prompts to attend to integrative fac
Fey, Tomblin, & Zhang, 2002), they have not learned tors like those highlighted by Kintsch (1998), such as
many of the skills associated with reading comprehen text features and background knowledge (Cain &
sion, including inference skills (Gersten, Fuchs, Oakhill, 1999), and when they are given integrative
Williams, & Baker, 2001). The combination of more stimuli (e.g., a descriptive title) along with the passage
demanding academic requirements and their inade (Yuill & Joscelyne, 1988). Other research studies in

quate reading skills contributes to poor academic out which poor comprehenders have been explicitly taught
comes for students with LD (Leach, Scarborough, & how to make inferences have focused on teaching stu
Rescorla, 2003; Mastropieri, Scruggs, & Graetz, 2003; dents to attend to integrative factors similar to those
Wagner et al., 2003). specified by Kintsch. In each study, students were
Further complicating matters is the fact that higher taught one or some combination of the following skills:
order skills, such as those involved in reading compre activating their background knowledge, making predic
hension, in general, and making inferences, in tions, asking and answering questions, looking for clues
particular, can be much more difficult to teach students in the text, making connections between prior knowl
with LD to a point of proficiency than lower-order edge and information in the passage, and attending to
processes (Fisher, Schumaker, & Deshler, 2002; text structure.

Swanson, Hoskyns, & Lee, 1999). Making inferences has Although the results of the studies in this area are
been defined as the ability to "construct the text base somewhat mixed, three studies show positive treatment
and the mental models that go beyond the information effects for poor comprehenders (e.g., Dewitz, Carr, &
articulated in the text" (Snow, 2002, p. 108). Patberg, 1987; Hansen & Pearson, 1983; Yuill &
directly
The ability to generate inferences is typically tested by Jocelyne, 1988). However, none of these studies focused
on students with disabilities or on secondary students.
asking questions like, "Why did the boy take action like
he did?" or "What is the main message of this passage?" The poor comprehenders' average posttest scores on
The answers to such do not appear directly in some of the criterion-based tests hovered below the
question
none of these studies used
the text. Instead, the reader is expected to integrate clues passing range. Furthermore,
in the text with prior knowledge to create an answer. standardized measures of reading comprehension.
Thus, although the ability to generate inferences may be With regard to secondary students, researchers focus
critical to text comprehension, the generation of an ing on the use of conscious strategic reading compre
inference in is the result of the in hension processes have reported that students with LD
reading essentially
dividual reader's response to the ideas presented in the can learn to use comprehension strategies that conceiv
contribute to inference
text; this is somewhat dependent on his/her ability to ably could generation (Gersten
some et al., 2001; Swanson et al., 1999). Examples of such
connect or bridge those ideas with prior knowl
in text (Pressley, 2000). include summarization
edge and with clues provided comprehension strategies
To provide a theoretical framework for what happens & Salvia, 1992), activating background knowl
(Gajria
inference Kintsch (1998) suggested edge and prediction (Afflerbach, 1990), and clarifying
during generation,
that during the comprehension process, mental repre (Simmonds, 1992). Other research has shown that sec
sentations are constructed or formed about the infor ondary students with LD can learn complex reading
mation read in text. When the reader thinks, strategies and that their scores on criterion-based read
being
these measures can increase as a result of
talks, or writes about representations, they ing comprehension

Learning Disability Quarterly 246

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
 Comprehension
Composite
Scores
p'
GRADE

Standard
Grade
Age.

59.75
76.0
Means
15.46
1.38
2.9

Student
Sex
Y-M
SES
Race
Disability
IQ
Score
Percentile
Equivalent

2.5
N
15-08
1M
SLD
67
Multi-racial
<1
<55
I N=neither
free
nor
reduced-cost
lunch;
F=free
lunch
orreduced-cost
lunch;learning
SLD= disability;
MR=Mentally
retarded;
LEP=Limite
Englis
pro
Specific
15-10 N 6F 15-05
N African-American SLD SLD 59<55<12.7 7
8 F 3F 15-02
15-01
F N/A MR
African-American 40<55<1
2.3 SLD
2M 16-05
F LEP/SLD 88744 3.9
Hispanic
3.3
115-01
66
105
SLD
White
M
F
7 SLD
5M F American-Indian
15-02 86661 3.3
White
African-American

Student
Information
. N/A=Not
.' '
available.

1
Table
U\

Oo
| bo ISO

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
the strategy instruction (see Schumaker & Deshler, Setting
2006, for a review). The school district is located in an urban midwestern
Each of the studies in this area has focused on one with a population of approximately
community
visual imagery)
reading strategy (e.g., self-questioning, 124,000. Instruction and testing took place in a class
that might be related to inference generation. None has room in the high school and was conducted with two
addressed inference generation as an outcome measure
groups of four students each during different class peri
or has employed a standardized reading test to measure ods.

changes in comprehension. Further, no study to date The Inference

has investigated a Strategy
teaching comprehensive package of The Inference Strategy is a reading comprehension
strategies that might be used for generating several
strategy designed to help students create meaning from
types of inferences. clues provided in text and respond to a variety of infer
Thus, the purpose of this study was to develop and
ence questions.
test the effects of an instructional to
Standardized reading tests were ana
program designed to determine the of inference questions that
teach an inference strategy to secondary students with lyzed types
students were expected to be able to answer. Four main
disabilities. Specifically, the study was designed to
examine the effects of explicit instruction in a multi types emerged: purpose, main idea/summarizing, pre
inference reading comprehension dicting, and clarifying (see Figure 1 for examples).
component strategy The Inference Strategy taught to the students in this
by assessing (a) student knowledge of the strategy, (b)
student use of the strategy while study consists of five steps. During Step 1, "Interact with
reading narrative pas
the passage and the questions," students first preview
sages, (c) student ability to answer four types of infer
the passage, paying particular attention to the title and
ential questions as well as literal comprehension
the length of the passage. Then they read the questions
questions, (d) student scores on a standardized measure
of reading comprehension,
and mentally identify two main categories of questions:
(e) student reading and
factual questions and think-and-seek (inferential) ques
strategy satisfaction, and (f) required instructional time
tions. Next, the students further classify the think
for students with disabilities in a secondary setting.
and-seek questions into four types: purpose, main
METHOD idea/summarization, prediction, and clarification ques
tions. During the second step, "Note what you know,"
Participants
students activate any background knowledge or experi
Participants were eight ninth-graders with disabilities
consent ences they may have related to the topic and questions,
whose parents had given for their participa
tion. Students were enrolled in learning-supported underline any key words in the questions that indicate
arts classes. A standardized what information to look for in the passage, and note
English/language reading
code letters next to each question to indicate the cate
test, the Group Reading Assessment and Diagnostic
Evaluation indicated that gory of the question and the question type. During the
(GRADE) (Williams, 2001),
scores fell at least five levels third step, "Find the clues," students carefully read the
the students' reading grade
below their current grade placement. Each student had passage and find and underline clues that are directly
been designated as a student with a disability and had related to key words in the questions. They then create
been placed in a resource program for a minimum of tentative answers to the questions mentally. The fourth
of amul more details," students to look
180 minutes per day on the recommendation step, "Explore prompts
for any additional clues in the passage that support the
tidisciplinary special services team with documented
tentative answers they have selected. The final step,
parent/guardian approval (see Table 1 for demographic
and test data on the students). Of the eight partici "Return to the question," calls for the students to go
seven had been classified by their school district back to each question and make sure that an answer has
pants,
as having a learning disability; one had been classified been selected and marked.
as having mental retardation. The participating district Thus, these steps were designed to cue students to
followed the IQ-achievement model for attend to their prior knowledge, to attend to the type of
discrepancy
identification of learning disabilities, a mini inference they were being asked to make, to attend to
requiring
mum discrepancy of 18 points between a student's IQ key words in the questions that would help them search
and achievement scores (J. Harrington, personal com for clues in the text, to search for those clues, and to
munication, July 21, 2007). Further, the state within problem solve once they had gathered the clues to make
which the participating district is located is one of five inferences about the information in the passage. The
states in this country that rely on the professional judg use of the generic strategy steps was constant regardless
ment of a team in the determination process (Reschly, of the type of question the student was addressing.
Hosp, & Schmied, 2005). However, students looked for different types of clues,

Learning Disability Quarterly 248

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
 Figure 1. Example inferential questions.

Purpose Questions:
1. What is the author's reason for writing this passage?

2. The author's for writing this is to ...

purpose passage

3. What is the most reason the author wrote this

likely passage?

Main Idea/Summarizing Questions:

1. Which of the following sentences best summarizes this passage?

2. This is mostly about ...

passage

3. The main topic of this passage is ...

Predicting Questions:
1. What is the next most likely event to happen after the end of this passage?

2. What would likely happen if the Olympics were held in Germany again?
3. Based on this in the future, Joe will ...
passage, probably

Clarifying Questions:
1. The man at the newsstand told Darren a dollar was a lot of money because ...

2. Why was Wilbur afraid?

3. What caused Charlie to lose his glasses?

depending on the type of think-and-seek question. (See student behavior, describe a step of the strategy or how
the Procedures section for a description.) to use the strategy, model the strategy, provide practice
The mnemonic device "INFER" was created from the opportunities with feedback, and provide a post organ
first letters of the steps to help students to remember the izer.

The number of these behaviors varied for each les

steps' names and their order in the strategy. Initially,
the strategy steps are to be used in the "I" to "R" order; son's checklist (from 7 to 11), depending on the content
of the lesson. For example, if one type of inferential
however, thereafter, the steps may be used in a recursive
and flexible manner to allow students to cycle back to question was to be introduced in a given lesson, there
were seven items on the checklist, corresponding to
any step if necessary.
those listed above. If two question types were to be
Instruments and Measures in a given
introduced lesson, then the strategy was
checklists. A checklist was used to measure
Fidelity described and modeled twice, once for each question
teacher adherence to the instructional sequence for the type, and nine items were listed on the checklist.
lessons. It listed several teacher behaviors: provide an The delivery of each of the instructional lessons was
advance organizer, discuss the purpose of the lesson and recorded using a tape-recorder and audiotapes. The
provide rationales for the lesson, state expectations for checklist was filled out by a scorer while listening to

Volume 30, Fall 2007 249

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
audiotapes of the lessons. If the scorer heard the teacher of 15 points per probe test (since there were three pas
emit one of the teacher behaviors listed on the check sages per probe test).
list, one point was awarded for that behavior. Zero test. The Strategy Knowledge
Strategy knowledge
points were awarded if the behavior was not emitted. A Test was used to measure the students' knowledge of
percentage score was calculated for the percentage of the steps of the Inference as a and
Strategy pretest
teacher behaviors in which the teacher engaged across posttest measure. It included five short-answer ques
the lessons. tions, which required students to list and explain the
Strategy use test. To obtainrepeated measures of strategy steps and their uses. An answer key specified
students' use of the Strategy, a pool of 30
Inference the parameters for correct answers. For example,
ninth-grade-level narrative passages from the Jamestown Question #1 was, "What is the first step a reader takes
Readers - Timed Readings in Literature Series (Spargo, when using the INFER Strategy?" (The correct answer
1989) was created; passages were randomly selected and was "Interact with the questions and the passage.") For
sequenced for each student from this pool. This level of each answer, a student could earn either zero points
passages (i.e., ninth grade) was chosen to provide infor (no information), 5 points (partial information [e.g.,
mation on performance at the students' current grade the student wrote "Interact with the questions"]), or 10
level because this is the level at which they are expected points (completely correct information [e.g., the stu
to perform in general education classes. Narrative pas dent wrote the whole name of the step]). A total of 50
sages were selected to control for the type of passage as points was available on the test.
well as to fulfill requirements for the English class in Standardized reading test. Two subtests of the
which the students were enrolled. GRADE (Williams, 2001) were administered to the
For each probe test, the student was asked to read students prior to and after instruction: Sentence
three of the 400-word passages silently and use the Completion and Passage Comprehension. The com
Inference Strategy in relation to the passage and the five bined scores from these subtests comprise the GRADE
factual and inferential questions that followed the pas comprehension composite score, which was the score
sage (see the next section for more about these ques used in this study. Forms A and B of the GRADE were
tions). To measure use of the strategy, students were used for pre- and posttesting, respectively. Reliability
awarded one point for each of the following strategic coefficients for the alternate forms and test-retest are in
behaviors: underlining key words in a question, under the .90 range. In addition, correlations between scores
lining clue words in the passage related to the question, on the GRADE and on the Gates-MacGinitie Reading
recording a code letter identifying the category of ques Test "range from .81 to .94, with half of the coefficients
tion (i.e., factual versus think-and-seek), and recording being .89 or higher" (Williams, 2001, p. 85).
a code letter identifying the type of think-and-seek ques Student satisfaction survey. The Student Satisfac
tion (i.e., purpose, main idea/summarization, predic tion Survey consisted of 10 questions, each formatted
tion, or clarification). using a 7-point Likert-type scale, ranging from "1"
Scorers used an answer key for awarding a total of four ("Totally wrong") to "7" ("Totally right"). This instru
points per think-and-seek question and three points per ment was administered to obtain a pretest and a posttest
factual question. (Students classified think-and-seek measure. On survey items, the students were asked to
questions by type. They did not have to classify the fac rate their attitudes toward reading in school and learn
tual questions by type, so they could not earn a fourth ing and toward using the Inference Strategy for the
point for these questions.) A total of 19 points were purpose of reading passages and responding to compre
available per passage for the strategy use score or 57 hension questions. Example survey items are "I am a
can take to
points per probe test (since three passages were read per good reader in school," "I know what steps I
probe test). make meaning from what I read," and "I feel that I can
Criterion-based comprehension test. As mentioned, use the Inference Strategy to help me understand what
each passage students
the read was followed by five I read in class."
written by a researcher specif Time required for instruction. Instructional time was
multiple-choice questions
ically for a given passage. The five questions consisted recorded in a journal kept by the researcher. Start and
of one factual question and one question for each of the stop times, including hours and minutes, were recorded,
four types of think-and-seek questions: purpose, main along with dates of instruction. Teacher time began
idea/summarization, predicting, and clarifying. Student when the teacher started (or restarted) the lesson with
answers to these were the criterion-based the students. It ended when an interruption occurred
questions
measure used in this study. Students (e.g., a phone call), or when the students began practic
comprehension
were awarded 1 point per correct answer (based on an ing the strategy. Student time began when the teacher
answer key) for a total of 5 points per passage and a total started (or restarted) the lesson with the students. It

Learning Disability Quarterly 250

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
ended when an occurred,
interruption or when the written by a researcher followed by two to four multi
students and handed in their work. ple-choice questions corresponding to the type of ques
stopped practicing
Thus, several start and stop times were potentially tion^) covered in each of the lessons. These short
recorded for each lesson. passages were used for practice activities immediately
For the fidelity checklists, two scorers after students had been introduced to how to use the
Reliability.
scored 40% of the lessons, and their strategy for a new question type.
independently
scores were compared item by item. An agreement was For the second type of practice material, narrative pas
tallied if both scorers had recorded the same score on an sages from the fourth-, sixth- and eighth-grade levels
item. The of agreement was calculated of the Jamestown Readers - Timed Readings in Literature
percentage by
series (Spargo, 1989) were so students could
dividing the number of agreements by the number of selected,
start practicing the strategy with relatively easy passages
agreements plus disagreements and multiplying by 100.
The scorers on 38 out of 39 (the fourth-grade passages). Gradually, across practice
agreed possible agreements
for a total percentage of agreement of 97% (range = 91% attempts, they progressed to more difficult passages
to 100%). (sixth-grade passages), and then to passages written one
Similar checks were for the grade level below their current grade (eighth-grade pas
reliability completed
Strategy Use Test, the criterion-based comprehension sages). The length of passages ranged from 200 to 400
tests, and the Strategy Knowledge Test for 40% of the words, increasing with the reading level of the passage.
tests taken during baseline and post-intervention condi For each passage, five multiple-choice questions (one
tions. Scorers were blind as to which tests were taken factual and four inferential questions, each correspon

during baselineand after instruction, respectively. With ding to one of the four types of think-and-seek ques
to the Use Test, the scorers agreed on tions) were written by a researcher.
regard Strategy
949 out of 969 possible agreements, for a total percent These long passages and accompanying questions
were used after all the question types had been intro
age of agreement of 98% (range = 93% to 100.0%). On
duced to the students so that they could practice using
the criterion-based comprehension tests, the scorers
on the strategy with a mixture of questions. Student per
agreed 100% of the 255 items scored. Finally, on the
formance on these activities was scored using an answer
Strategy Knowledge Test, the scorers agreed on 28 items
out of a possible 30, for a total percentage of agreement key and the guidelines described under the Measures
section for the Strategy Use Test and the criterion-based
of 93% (range = 80% to 100%).
comprehension test.
Reliability for teacher and student instructional time
Pretest procedures. Students completed a Reading
was determined for three of the instructional sessions
Satisfaction a
Survey and minimum of three probe tests
during which two scorers recorded the times separately.
containing a total of nine Strategy Use Tests and nine
Agreement was registered each time the two scorers'
corresponding five-item criterion-based comprehension
times agreed on a start or stop time to the minute. The
tests over a one-week period. For each probe test, a
total percentage of agreement on teacher time was
Strategy Use Score and a Comprehension Test Score
100%, and on student time it was 93%. The scorers'
were determined. When the third baseline probe was
times agreed on 10 out of 10 possible agreements for
completed, four students whose baselines were stable
teacher times and on 15 out of 16 possible agreements
for student times. (hereafter referred to as Cohort A) began instruction in
the Inference Strategy. Once the Cohort A students
Procedures showed an increase in their use of the strategy, the four
Instructor. The instruction was completed by the first other students at least one addi
(Cohort B) completed
author, who is a certified special education teacher with tional baseline probe until their baselines were stable.
five years' experience teaching students with disabili Then they began the instruction.
ties. She also is a certified strategic instruction model Intervention The students received
procedures.
professional developer with the University of Kansas instruction in the Inference Strategy in sessions ranging
Center for Research on Learning. in length from 60 to 75 minutes, on the
depending
Instructional materials. An instructional protocol school schedule. Instruction was based on a validated
(Fritschmann et al., in prep.) was written to ensure that instructional methodology for teaching learning strate
instruction was standardized across the two classes. It
gies to students with disabilities (Schumaker & Deshler,
was comprised of scripted step-by-step instructions for 2006).
each lesson plus visual devices to be used during the In the first instructional the students were
session,
instruction. asked to make a commitment to actively learn and use
Two types of practice materials were constructed. For the Inference Strategy. Also, in the first class session, the
the first type, nine short, one-paragraph passages were instructor explained and described in detail the steps of

Volume 30, Fall 2007 251

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
 the Inference Strategy. This was followed by instruction This process was modeled for the students prior to the
that how to identify and code the two main
included practice activities with the short passages. The instruc
categories of questions, (a) factual questions (i.e., ques tor provided group and individual feedback during and
tions with answers that are "right there" in the passage) after each practice activity, and mastery was required.
and (b) "think-and-seek" questions (i.e., questions that In subsequent sessions, students were provided prac
require the reader to really think about and seek out tice activities with the longer fourth-grade-level read
clues so that they can infer an answer). Additionally, ing passages, in which they were required to use all the
the students were instructed on how to identify and steps of the strategy in response to all question types. If
mark each type of think-and-seek question with code the student earned a score at or above the mastery lev
letters. els (80% on the Comprehension Test, 70% on the
The students were then taught through description Strategy Use Test), they were up to the next
moved
and demonstration how to look for key words embed level (sixth-grade level, then eighth-grade
reading
ded in the question to code and respond to factual ques level). They continued to practice and receive individ
tions. At the end of the first session, the students ual feedback until they reached mastery on a passage
completed practice activities for coding question cate written at the eighth-grade level. (See Fritschmann,
gories and answering factual questions for short pas 2006, for more details on the instruction.)
sages, and they received feedback on their work. If they Posttest of instruc
procedures. Upon completion
met mastery (i.e., earned 80% or more of the points), tion in the Inference Strategy and reaching the mastery
they proceeded to the next instructional session. If they criteria on one eighth-grade students
practice activity,
did not reach mastery on the practice activity, they took a posttest containing three 400-word, ninth-grade
completed additional practice activities until they met passages. After reading each passage and using the
mastery. strategy, students were asked to complete a five
During the second instructional session, the students criterion-based test. Each stu
question comprehension
were provided detailed information on purpose ques dent completed a minimum of one such posttest (con
tions and main and they
idea/summarizing questions, taining a total of 3 passages and 15 questions).
were taught how to identify and mark those questions these tests, the students were administered,
Following
with code letters. They were also taught how to look for as a group, the two selected subtests included in Form B
key words embedded in the questions and clues imbed of the GRADE (Williams, 2001). They were also admin
ded in the text and how to correctly respond to these istered the Student Satisfaction Survey and Strategy
of For for purpose questions,
types questions. example, Knowledge Test in a group setting. The students com
students were taught that authors may have three main and there was
pleted each instrument independently,
purposes as they write a passage: to entertain, to
no time limit for test completion. This posttesting
inform, and to persuade. Each type was defined, and
occurred during the last weeks of the school year.
students were taught to look for the key words Two maintenance tests
Maintenance procedures.
"author's or "author's reason" in the
during the next academic year. The
purpose" ques were administered
tion. Then as they read the passage, they were taught to after the posttests
first was administered eight months
ask themselves "Why do I think the author wrote this?" to six of the eight original subjects who were present
and attend to certain types of clues that would help The students did
on the day the test was administered.
them determine the author's purpose. For example, for
not receive a review of the Inference Strategy prior to
informative passages, they were taught to look for large test, nor had the strategy
like they might see in a taking the first maintenance
amounts of facts and details the
been reviewed with them since they had completed
textbook. For entertaining passages, they were taught months earlier. There was no time limit on
them happy or fearful. study eight
to look for sections that made
the test.
After instruction, discussion, and modeling, students
A second maintenance test was administered to four
were provided with short practice passages followed by 12
students who were still enrolled in the school
a mixture of factual, purpose, and main idea questions,
months after In this case, the students
the posttests.
and were given feedback on their efforts. Again, mas
took part in a 45-minute session where
review they
tery was required before students proceeded.
sessions focused on practiced using the Inference Strategy the day before
The third and fourth instructional
taking the maintenance test.
instruction of and practice with predicting and clarify
As in the previous sessions, Research Design
ing questions, respectively.
the researcher explained in detail how to identify the A multiple-probe-across-subjects design (Horner &
new and search for clues in the passage Baer, 1978) was to determine the effects of
type of question employed
to support a correct response to that type of question. instruction on students' strategy use and reading com

Learning Disability Quarterly 252

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
prehension performance. All students were given at RESULTS
least three reading probes before instruction began. The
Fidelity of Implementation Results
students who served as the second tier in the design
The instructor emitted 86 of 88 listed behaviors on
(Students 2, 4, 6, and 8) had at least four probes before the fidelity checklists for a total percentage of imple
instruction began. Their instruction commenced only mentation of 98%.
after the scores of students in the first tier of the
Use Test and Criterion-Based
design had improved. All students' baselines were sta Strategy
Test Results
ble before they began the instructional program. Addi Comprehension
tionally, a pretest-posttest analysis was employed to Figures 2-4 display the percentage of points earned

compare the standardized reading test scores earned by each of the eight participants on each Strategy Use
before and after instruction on the GRADE. Test (diamond symbols) as well as the percentage

Figure 2. Percentage of points earned for strategy use and percentage of comprehension questions
answered correctly by Students 1 and 2.

STUDENT 1
Baseline
100n Instruction Generalization
90
#^^^
t> sen -^
xB..^irrrrt
- ''
<? 60n P
& 50 n /
jS 40- /

'*
fe 20n ^^^
?" *^
10-
-
0-- -!? ?'-;-1-1-1-j-1-1

9 9 9 4446899
Reading Passage Grade Level

STUDENT2 I I
100-
j
^^

" t3 80^ -"'"' ^hI ^-m

t 70^
5 60
fl) 50

3 40-i
son "
? .B_
" m
"
201
*fe
101
o-h-*?? ?!? ?!? ?I-!-1-1-1-1-1-1
999946899
Reading Passage Grade Level

? - Use - - -- -
Strategy Comprehension

Volume 30, Fall 2007 253

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
 Figure 3. Percentage of points earned for strategy use and percentage of comprehension questions
answered correctly by Students 3, 4, and 5.

STUDENT3
Baseline 100 Instruction Generalization
90

?- "
? 70-J
60- m^~~m^^ ^^
$
50- Jtr ^^*
&
o> x^
S 40- /
JU x *
o 0
fe 20- x' /~
?"
10
0^-
y -1- -1- -*?\-1-!-1-1-1-1
9994446899
Reading Passage Grade Level

STUDENT4 I

90- *''
m^*^ S^^a
?? 7o^/
o
eo^
50-
^^"^^""
"' /
O)
JS 40

8 30"
fc 20
?"
10
oH? ?i? ?i? -1-1-1-1-1-1-1-1

999 46899
Reading Passage Grade Level

STUDENT5 I
100
^m%%

5 som/ x* w/
S 60- m
o 50

S 40
30 ''
3 ** ~B^.
o 20-
?-
10
- '
oH? ?i? ?i? ?i? ?I-1-1-1-1-1-1-1
*'*
999446899

Reading Passage Grade Level

?#?
.^; Strategy Use --*-- Comprehension

y.jr#-^te-^toJg>.-.:. .- .
._..^^_ _

Learning Disability Quarterly 254

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
 ' - mini i . i.u j y ? i
I1.wmipi'W'^'1-.,
-!
^"' { J'W Y-Y
"' W""'-"
.
.';""i luimj 11 j. ; u,m. m j^i IM^^VTC^
'' - = '
??-- -'^ .. . -.: , ,:..' -_, ....:..- ^y-\ : ;- >' ; ,^
,V 'v.--- i- V^:>:-'- -_; v-y ;>?
?; W^-^^wMfiafe;_,

Figure 4. Percentage of points earned for strategy use and percentage of comprehension questions
answered correctly by Students 1 and 2.

STUDENT6
Baseline
100 I I
Instruction Generalization

& 50^ /
2 40i B^ /
8 30" ^.
" ^^*
fe 20 n
a
io-|
Oi ?i? ?i? i i i i i i i
9994446899
Reading Passage Grade Level

STUDENT7 I
100

'
o 601 - /
& 5o / ^/
S 40- i y / \
8 301 ^S*
a> 20H ?r

oH? ?i? ?i? ?I-1-1-1-1-1-1-1-1

9 9 9 4 4 6 8 9 9
Reading Passage Grade Level

? - Use - - -- -
Strategy Comprehension
, ;,. ,.: iM,,?,..,.,?,..M ;i ' 7"" '-'
^-1J............MU-...
Y-^'v-'
<?^?^>?y<<^.? ,;,^-',r;
T?- -;-; -""t111"1" '",-;""'"?.&? '-'"-'""
r1"1"-"." :"'**>>rr,rwfm^^*"?XW~.

of comprehension questions answered correctly (square The figures show that the percentage of comprehen
symbols) on each criterion-based comprehension test. sion questions answered correctly increased with the
(Each symbol represents performance on three passages onset of instruction and maintained a positive trend
and three sets of questions.) The grade level at which through the remainder of instruction and posttesting.
each reading passage was written is shown along the During baseline, the students answered an average of
x-axis of each graph. Student 3's data were graphed 31.74% of the
comprehension questions correctly.
with those of Students 4 and 5 because his graph part During instruction, they answered an average of
ner moved from the school and could not continue 77.39% of the questions correctly; during the posttest
with the study. Likewise, Student 6's data were graphed condition, they answered an average of 82% of the
with those of Students 7 and 8 because his graph part questions correctly. Similar results were evidenced with
ner was excluded midway from the study due to a high regard to the Strategy Use Test. During baseline, the
absentee rate. students earned an average of 0% of the points on the

Volume 30, Fall 2007 255

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
Strategy Use Test. During instruction, they earned an at levels that were higher than during baseline and that
average score of 66.39%; during the posttest condition, approximated their posttest performance more closely
they earned an average score of 81.94% on the Strategy than when they took the test without a review.
Use Test. A Friedman Test was conducted to evaluate differ
Table 2 displays the results from the maintenance ences between the median for the Strategy Use Test
tests. On the left-hand side are listed the comprehen scores during baseline =
(median 0%), posttest (median
sion (M= 41%, SD = 11.73) and strategy use percentage =
82%), and maintenance without review (median = 19
scores (M = 19%, SD = 14.19) earned by the six partici %) conditions. (Data from the maintenance with
pants who took a maintenance test eight months after review test were not included because of the low num
instruction was completed without participating in a ber of subjects.) Significant differences were found, x2>
review Test 1). On the right-hand side = = < and the Kendall coefficient
(Maintenance (2,N 6) 11.565, p .01,
are the scores earned by the four participants who took of concordance size index) of .964 indicated
(effect
the test 12 months after instruction was completed and strong differences among the three median scores.
after a brief review (Maintenance Test 2). The scores Follow-up pairwise comparisons were conducted
earned by the latter group of students suggest that with using aWilcoxon Signed-Ranks Test (Siegel & Castellan,
a brief review and practice, students were able to apply 1998); the LSD procedure was used to control Type I
the steps of the strategy and answer questions correctly errors across these comparisons at the .05 level. The

Table 2
Percentage Scores on the Maintenance Tests

Maintenance Test 1 Maintenance Test 2

Comprehension Score Strategy-Use Score Comprehension Score Strategy-Use Score

Student Without Review3 Without Review3 With Reviewb With Reviewb

1 40 0 60 79

2 n/a n/a 67 84

3 26 15 53 77

4 60 22 n/a n/a

5 47 40 n/a n/a

6 33 29 n/a n/a

7 40 10 n/a n/a

8 n/a n/a 80 88

Mean (SD) 41(11.3) 19(14.19) 65(11.52) 82(4.97)

n/a=not available for testing; an=6, bn=4.

Learning Disability Quarterly 256

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
 Table 3

Percentage Scores on the Strategy Knowledge Pretest and Posttest

Pretest
Student Posttest
j

0 1 100 j
I 0 2 96 | j
3 0 80

0 4 100

5 0 96

6 0 92

7 0 80

8 0 90

Mean (SD) 0 91.75* (8.031)

*Significant difference found.

median score for the Strategy Use posttests was signifi cantly higher than the median score for the baseline
than the median score for the baseline = and it was also signifi
cantly higher comprehension tests, p .012,
= was also significantly than the median score for the mainte
tests, p .012, and it higher than cantly higher
score for the maintenance without-review nance = score was
the median test, p .028. The median maintenance
= .028. The median score for the maintenance than the median baseline score,
tests, p significantly higher
=
without-review test was significantly higher than the p .027.

median baseline score, p = .043. Test Results

Strategy Knowledge
A Friedman Test was also conducted to evaluate the Table 3 lists the percentage scores earned by the eight
differences in the medians for the percentage of com students on the Strategy and
Knowledge pretest
prehension questions answered correctly during the As illustrated, each student earned a score of
posttest.
baseline = 33.51 = = =
(median %), posttest (median 0% on the pretest (M 0.00%, SD 0.00). Percentage
and maintenance-without-review conditions scores on the posttest =
78.42%), ranged from 80% to 100% (M
= were found, =
(median 41.00%). Significant differences 91.75%, SD 8.031). The Wilcoxon Signed-Ranks Test
= =
X2(2,N 7) 12.00, p < .01, and the Kendall coefficient (Siegel & Castellan, 1988) was used to determine
of concordance (effect size index) of 1.00 indicated whether there were significant gains from pretest to
strong differences among the three median scores. posttest on this measure; the LSD was used to
procedure
Follow-up comparisons
pairwise were conducted using control Type I errors across these comparisons at the .05
aWilcoxon Signed-Ranks test (Siegel & Castellan, 1998); level. A significant difference was revealed between the
the LSD procedure was used to control for Type I errors = = .00. The effect
pretest and posttest scores, z -2.530, p
across these comparisons at the .05 level. The median size for this gain was r = 0.99, representing a large gain
score on the posttest comprehension tests was signifi according to Cohen (1988).

Volume 30, Fall 2007 257

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
Standardized Reading Test Results Table 6 displays the mean pretest and posttest ratings
Table 4 lists the mean standard scores and grade-level and the standard deviations for each item on the
scores earned by students on the pretest and posttest on questionnaire. The that received
items the highest
the GRADE. The Wilcoxon Signed-Ranks Test revealed a posttest ratings relatedto enjoying the exercises and
while the strategy =
significant difference between the pretest and posttest passages learning (M 6.62) and
standard z = = .012. The effect size for the to aid understanding in class
scores, -2.521, p using strategy
this gain was r = 0.91, a large gain according to Cohen (M=6.37).
(1988). This gain represents an average increase of 2.82 Time
= 1.4 to 3.6 Required for Instruction
grade levels (range grade levels) in reading Instructor time. The total amount of instructor time
comprehension.
required to deliver the initial Inference Strategy instruc
tion ranged from 280 min. to 350 min. =
Satisfaction Results (M 300 min.,
The mean pretest and posttest reading satisfaction or approximately 5 hours).
ratings are reported in Table 5. As illustrated, signifi Student time. The total amount of student time
cant differences were found between the pretest and included time
spent working with
(see the instructor
posttest mean scores for the 7-point Reading the time reported above) and working independently
=
Satisfaction Questionnaire using aWilcoxon Test, z on practice activities with reading passages. It ranged
= 0.12. The effect size for this difference was from 770 min. to 1040 min. = 905 min.,
-2.524, p (M approxi
r = 0.95, a large gain according to Cohen (1988). mately 15 hours).

Table 4
The GRADE Comprehension Standard Scores and Grade-Level Equivalents

GRADE GRADE Grade-Level Grade-Level

Student Pretest SS Posttest SS Equivalent Pre Equivalent Post

551 88 2.5 5.6

742 88 3.9 5.6

553 87 2.3 5.3

724 94 3.8 7.2

665 83 3.3 4.7

556 91 2.7 6.3

58
7 89 2.9 5.9

558 89 2.5 5.9

Mean (SD) 61.25 (8.17) 88.63* (3.16) 2.99 (0.52) 5.81 (0.73)

*Significant difference found.

~ - ''" "?'- ~Y Y- ' ~

-<.. j...?-.., _ . ----v. 7 < , Yt.. "^-'' - ' ' X:'-' '- "7
^I^^S&^it^i'V^ YY7-YY ;Y-"^-Y- ?>>?.\7^- - -Y*Y '-: ..: "*-7/-,7;;'
' Y -Y
I .feMjftife ",- ;/ ,^n7--,Y7 -VY7 Y^YYYYmV V ^T 'V. , . Y' 7 7^ ^ , I

Learning Disability Quarterly 258

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
 - "" '-""". - .'.'"" .il" '
[[ m" " U l ."" ' -. I I II 1 I I'll I ?W?^^M
lJMINIUIIII|ipil
'[.1^'"' '^"-V., iK-' . 1 ^1
nnrfrr mhi if,nin iri >Im
i n,,.', i , .'.n.Iim" ) , in .? ?? in ..in..'. in wr mm ,,,,.?,, ," rf., , u ,-,, ,! ? nr ,il m ,,'.,?.
..',? hv , i i 'n'..%?
n ,,,.,i 1. \
l^iii^T.iilnWitm^ftiyin'i.;,

Table 5
Overall Mean Ratings for All Students on the Reading Satisfaction Questionnaire

Student
Pretest Posttest

1 1.1 5.8

2 1.6 4.9

3 1.5 5.1

42.3 5.5

5 1.5 4.4

6 1.8 6.4

7 1.3 6.5

8 1.3 6.6

Mean (SD) 1.55 (3.75) 5.65* (.815)

*Significant difference found.

DISCUSSION were large. During the instruction, the students gradu

Several conclusions may be drawn from the results of ally worked up to reading passages written at their
this study. First, instruction in the Inference grade level and responded to associated inference com
Strategy
increased students' use of strategic skills to a mastery prehension questions at mastery levels. They also
level. Second, the instruction produced a positive earned scores above 90% on a test of their strategy

change in the students' ability to respond to inference knowledge at the end of the study.
type questions on criterion-based tests. Increases in In addition, the results of the Student Satisfaction
strategy use and ability to answer inferential questions Questionnaire suggest that students who participated
were found only after students participated in the in the Inference Strategy instruction were more satis
Inference Strategy instruction as demonstrated through fied at the end of the study than at the beginning of the
the multiple-baseline design. Third, students' posttest study with how they felt about reading and different
scores on the GRADE were significantly higher than comprehension processes. On the one question related
their pretest scores following instruction in the to their reaction to the instruction, the students indi
Inference the GRADE scores indi cated that they enjoyed the instruction = 6.62 on a
Strategy. Moreover, (M
cated that, on average, the students made a 2.8 grade 7-point scale). Finally, instruction of the Inference
level gain in reading comprehension within 15 hours Strategy required five hours of initial instructional
of instruction. The effect sizes related to the gains asso delivery and then another 10 hours of supervising prac
ciated with all three of these major outcome measures tice activities and providing feedback.

Volume 30, Fall 2007 259

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
 Table 6 I
Mean Pretest and Posttest Ratings, Standard Deviations, and Ranges for All Students on
j
Each Item on the Reading and Strategy Satisfaction Questionnaire

Fill in the box that best describes how you feel ... Pretest Posttest

; Reading in school is boring. Mean 6.12 4.375

SD 0.83 2.39
Range 5-7 1-7

I am a good reader in school. Mean 2.00 3.63*

SD 1.07 2.64

Range 1-4 1-6

Reading assignments in school confuse me. Mean 6.37 6.00

SD 1.06 1.19

Range 4-7 1-7

I know what an inference in reading is. Mean 1.37 4.75*

SD 0.52 2.25
Range 1-2
1-7

Coming to a conclusion from what I read in class Mean 6.50 5.37

can be difficult for me. SD
0.76 1.60
Range 5-7 3-7

I know what steps I can take to make meaning Mean 2.06 4.56*
fromwhat I read. 0.64
SD 1.92
Range 1-3 2-7

I enjoy reading in class. Mean 2.25 3.75*

SD 1.28 2.12

Range 1-4 1-6

I felt comfortable using the Inference Strategy Mean 1.00 5.87*

while I read in class. 0 SD 1.55
Range 1 3-7

I feel that I can use the Inference Strategy to help Mean 1.00 6.37*
me understand what I read in class. 0 SD 0.52
Range 1 6-7

I enjoyed the exercises and passages that were Mean 1.00 6.62*
used while learning the Inference Strategy. 0 SD 0.52
Range 1 6-7

*
Represents amean gain of at least one point.

Learning Disability Quarterly 260

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
 Several limitations apply to this study. First, although
score the
importance of continued practice and peri
the sample size is adequate for the experimental odic maintenance probes if students are expected to
design
used, it was small. Expanding the research design to continue to use a given strategy.
include more classrooms would increase the sample size On the one hand, these data indicate that compre
and, therefore, eliminate the problem of committing a hension strategies learned by students with disabilities
that are not will not continue to be an active part of their repertoire
Type II error and making generalizations
based on a sufficiently large sample of students. of reading behaviors unless students are given opportu
Another possible limitation relates to the characteris nities to practice the strategies over a sustained period
tics of the participants. They represent students with of time. On the other hand, these data are encouraging,
severe reading comprehension deficits, beyond those because they clearly demonstrate that proficiency with
evidenced in the of students with the targeted behavior can be quickly regained when a
typically population
at large. How much instructional relatively modest review procedure is used. Clearly, the
learning disabilities
time would be required and the kinds of gains that instructional dynamics surrounding maintenance and

might be made by students exhibiting lesser deficits sustained use of complex strategies such as the
still need to be determined. Another limitation is that Inference Strategy require considerably more study.
test scores on the students' decoding skills were not Research is needed to determine the effects of
available. Thus, it is unknown whether they had higher Inference Strategy instruction with larger, more di
levels of decoding skills than comprehension skills, verse, and more carefully described groups of students,
which enabled them to tackle the higher reading-level and with teachers, as opposed to researchers, providing
passages.
the instruction. Maintenance procedures need to be
An additional concern is that the instruction was explored in more depth and detail. In addition, further
provided by a researcher. Whether other teachers can analysis of the data focusing on student performance
the same types of reading gains is unknown. on each of the question types may provide helpful
produce
Further, expository passages were not included in the information. Last, research is needed that focuses on
instructional materials or the criterion-based tests. use of the strategy with expository as well as narrative

Although they were included on the GRADE, how the passages.

students performed on those particular passages is not
Implications for Practice
known. Whether the strategy can be used successfully The instruction described here for the Inference
with passages remains to be determined.
expository Strategy has the potential of impacting education more
This is an important consideration because expository in can
broadly the following ways. First, the instruction
passages represent a considerable portion of the mate potentially serve as a vehicle through which small
rials used in content-area classes and are
secondary groups of students with disabilities can be taught the
included within state
nationaland assessments. inference skills required on today's standardized read
Another concern
is that the results were not disaggre tests. As a schools will be to
ing result, better equipped
gated according to the type of question on the tests. meet and students' needs.
increasing requirements
Thus, at this time, whether the students responded teachers may be able to combine this type of
Second,
more successfully to some types of questions than oth instruction with other forms of explicit reading com
ers is not known.
prehension strategy instruction (e.g., see Schumaker &
A final limitation relates to the maintenance data col for a because of similar instruc
Deshler, 2006, review)
lected eight months after the initial intervention. tional design features. Third, the Inference Strategy
While these data were collected on a relatively small instruction be used
might with student populations
subset of the entire sample, some trends require an low socio-economic and ethnically diverse
representing
explanation. Specifically, performances on both strat as well as those who have severe reading
groups,
egy use and reading comprehension probes were rela deficits.
tively low. Educational research has shown that the
maintenance of a targeted behavior as a result of an
intervention has largely been an unrealized
REFERENCES
goal of Afflerbach, P. P. (1990). The influence of on
prior knowledge
behavioral interventions
(McConnachi & Carr, 1997). readers' main idea construction
expert strategies. Reading
While features of the intervention itself need to be con Research Quarterly, 25(1), 31-46.
sidered as possible reasons for the low performance, American Institute for Research. (2005, Spring). Reading framework
the 2009 National Assessment of Educational
other factors may be responsible. In this study, the stu for Progress, pre
publication edition. Washington, DC: Author. Retrieved March 3,
dents received no additional instruction in the use of
2006, from http://www.nagb.org/pubs/pubs.html
the strategy nor prompts to use it after the study was D. D., Schumaker,
Bulgren, J. A., Marquis, J. G., Deshler, J. B., &
completed. These maintenance data seem to under Lenz, B. K. (2006). Instructional context of inclusive secondary

Volume 30, Fall 2007 261

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions
 general education classes: Teachers' instructional roles and prac for students and teachers.
struggling Learning Disability
tices, curricular demands, and research-based practices and stan Quarterly, 26(2), 103-116.
dards. Learning A Contemporary
Disabilities: 4(1), 39-66.
Journal McConnachi, G., & Carr, E. G. (1997). The effects of child behav
Bulgren, J. A., Schumaker, J. B., & Deshler, D. D. (1988). Effective ior problems on maintenance of intervention Behavior
fidelity.
ness of a concept routine in enhancing the perform 123-158.
teaching Modification, 97(2),
ance of LD students in secondary-level mainstream classes. for 21st Century Skills.
Partnership (2006). Learning for the 21st
Learning Disability Quarterly, 11(1), 3-17. century. A report and mile guide for 21st century skills. Retrieved
Cain, K., & Oakhill, J. (1999). Inference making ability and its rela March 29, 2006, from http://www.21stcentury skills.org.
tion to comprehension failure in young children. and M. should
Reading Pressley, (2000).What comprehension instruction be
Writing: An Interdisciplinary Journal, 11(5), 489-503. the instruction of? In M. L. Kamil, P. B. Mosenthal, P. D.
Catts, H. W., Fey, M. E., Tomblin, J. B., & Zhang, X. (2002). A lon & R. Barr
Pearson, (Eds.), Handbook of reading research, Volume II
gitudinal investigation of reading outcomes in children with
(pp. 545-562). Mahwah, NJ: LEA.
language impairments. Journal of Speech, Language, and Hearing D. J., Hosp,
Reschly, J. L., & Schmied, C. M. (2005). And miles to
Research, 45, 1142-1157. go ...: State SLD requirements and authoritative recommendations.
Center on Educational (2005). High school exit exams. Policy
Policy. Retrieved from www.nrcld.org/research/states/
July 31, 2007,
brief 1, December. Retrieved April 1, 2006, from http://www.cep index, shtml
dc.org/highschoolexit Schumaker, J. B., & Deshler, D. D. (2006). Teaching adolescents to
Deshler, D. D., & Schumaker, J. B. (2006). Teaching adolescents with be strategic learners. In D. D. Deshler & J. B. Schumaker (Eds.),
disabilities: Accessing the general education curriculum. New York:
Teaching adolescents with disabilities: Accessing the general educa
Corwin Press.
tion curriculum (pp. 121-156). New York: Corwin Press.
Dewitz, P., Carr, E., & Patberg, J. (1987). Effects of inference train
D. D., Bui, Y., & Vernon, S. (2006). High
Schumaker, J. B., Deshler,
ing on comprehension and comprehension monitoring. Reading schools and adolescents with disabilities: at every
Research Quarterly, Challenges
22(1), 99-121.
turn. In D. D. Deshler & J. B. Schumaker (Eds.), Teaching adoles
Fisher, J. B., Schumaker, J. B., & Deshler, D. D. (2002). Improving cents with the general
disabilities: Accessing education curriculum
the reading comprehension of at-risk adolescents. In C. C. Block
(pp. 1-34). New York: Corwin Press.
& M. Pressley (Eds.), Comprehension instruction: Research-based
Siegel, S. & Castellan, J. N. (1988). Nonparametric statistics for the
best practices (pp. 351-364), New York: Guilford.
behavioral sciences (2nd ed.). Boston: McGraw-Hill.
Fritschmann, N.
S. (2006). The effects of instruction in an inference
Simmonds, E.P.M. (1992). The effects of teacher training and
strategy on the reading comprehension skills of adolescents with
implementation of two methods of improving the comprehen
learning disabilities. Unpublished doctoral dissertation, Univer
sion skills of students with learning disabilities. Learning
sity of Kansas, Lawrence.
Disabilities Research and Practice, 7(4), 194-198.
Fritschmann, N. S., Schumaker, J. B., & Deshler, D. D. (in prepara
Snow, C. E. (2002). Reading for understanding: Toward a research and
tion). INFER: The inference strategy instructors manual.
development program in reading comprehension. Santa Monica,
Gajria, M., & Salvia, J. (1992). The effects of summarization
CA: Rand.
instructions on text comprehension of students. Exceptional
Spargo, E.
(1998). Timed readings in literature - Jamestown Series.
Children, 58(6), 508-520.
Providence, RI: Glencoe/McGraw-Hill of McGraw-Hill Educa
Gersten, R., Fuchs, L. S., Williams, J. P., & Baker, S. (2001).
tion.
Teaching reading comprehension strategies to students with
Swanson, H. L., Hoskyn, M., & Lee, C. (1999). Interventions for stu
learning disabilities: A review of research.
of Educational Review
dents with learning disabilities: A meta-analysis of treatment out
Research, 71(2), 279-320.
comes. New York: Guilford.
Hansen, J., & Pearson, P. D. (1983). An instructional study:
the inferential of good and poor Wagner, M., Marder, C, Blackorby, J.; Camento, R., Newman, L.,
Improving comprehension
Levine, P., & Davis-Mercier, E. (with Chorost, M., Garza, N.,
fourth-grade readers. Journal of Educational Psychology, 75(6),
821-829. Guzman, A., & Sumi, C). (2003). The achievements of youth with
race disabilities during secondary school. A report from the National
Heubert, J. P. (2002). Disability, and high-stakes testing.
MA: National Center on the General Longitudinal Transition Study-2 (NLTS2). Menlo Park, CA: SRI
Wakefield, Accessing
Curriculum. Retrieved March from International. Retrieved April 3, 2006, from http://www.
30, 2006, http://www.
cast.org/publications/ncac/ncac_disability.html nlts2.org
R. D., & Baer, D. M. A Warner, M. M., Schumaker, J. B., Alley, G. R., & Deshler, D. D.
Horner, (1978). Multiple-probe technique:
variation of the multiple baseline. Behavior (1980). Learning disabled adolescents in the public schools: Are
Journal of Applied
189-196. they different from other low achievers? Exceptional Education
Analysis, 11(1),
Hughes, C. A., Deshler, D. D., Ruhl, K. L., & Schumaker, J. B. Quarterly, 1(2), 27-35.
instruction for adolescents with Williams, K. T. (2001). GRADE: Group Reading Assessment and
(1993). Test-taking strategy
behavior disorders. Emotional and Behavioral Diagnostic Evaluation. Shoreview, MN: Pearson AGS Globe.
Journal of
188-199. Yuill, N., & Joscelyne, T. (1988). Effect of organizational cues and
Disorders, 1(3),
A paradigm on good and poor comprehenders' story understand
Kintsch, W. (1998).
Comprehension: for cognition. New strategies
York: Press. ing. Journal of Educational Psychology, 80(2), 152-158.
CambridgeUniversity
Leach, J.M., Scarborough, H. S., & Rescorla, L. (2003). Late-emerg

ing reading disabilities. Journal of Educational Psychology, 95(2),

211-224. Please address correspondence to: Nanette S. Fritschmann,
M. A., T. E., & Graetz, J. E. (2003). Reading School of Education, Lehigh University, Bethlehem, PA 18015;
Mastropieri, Scruggs,
comprehension instruction for secondary students: Challenges [email protected]

Learning Disability Quarterly 262

This content downloaded from 195.78.108.37 on Sat, 14 Jun 2014 21:00:42 PM

All use subject to JSTOR Terms and Conditions