See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/270580077

Comprehension instruction in the 21st century

Article · January 2009

CITATIONS READS
3 767

2 authors:

Donna Caccamise Lynn E Snyder

University of Colorado Boulder University of Colorado Boulder
27 PUBLICATIONS   201 CITATIONS    45 PUBLICATIONS   1,521 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

eBRAVO View project

All content following this page was uploaded by Donna Caccamise on 01 October 2015.

The user has requested enhancement of the downloaded file.

Comprehension Instruction in the 21st Century
by Donna Caccamise and Lynn Snyder

F or several years reading researchers have kept a watchful

eye on the National Assessment of Educational Progress’
(NAEP) “national report card” to determine whether newer
to assist you in recognizing these processes in a proffered read-
ing program or to incorporate them into existing curricula. That
section will be followed by a discussion of different approaches
teaching methods, textbooks, and interventions were having to comprehension instruction, and a description of a couple of
any significant effect on students’ reading proficiency. Since promising applications already used successfully in some
this testing program began more than a dozen years ago, stu- schools.
dent performance has remained fairly flat. The profile of U.S
students has been characterized by a decline in proficiency Identifying Good Comprehension Instruction
during the middle school years, with a surprisingly low per- As teachers know, the market seems saturated with reading
centage of students reading with proficiency. This proficiency programs “guaranteed” to help students learn to read. In this
profile has remained essentially unchanged during the high section we will describe the latest and most promising ideas on
school years (Snyder & Caccamise, 2008). It is no wonder that what good comprehension processing involves. The foremost
U.S. students ranked 18th in reading among the nations in the scientific theory of reading and listening comprehension is
2007 international rankings, outranked by nations such as constructivist in nature, in which an individual constructs and
South Korea, Finland, and Hong Kong, (Organization for integrates meaning from text or oral discourse into a newly cre-
Economic Co-operation and Development, 2007). ated mental representation. A leading model of this process
In the United States, the early signs of this impending (Kintsch, 1998) describes this process as automatic, not effort-
problem can be found in an initial decline in students’ read- ful, until the reader fails to construct a coherent representation
ing comprehension in the 4th grade, historically called the of the text they are reading. When that happens, good readers
4th-grade slump. We and other researchers focused on engage in active comprehension strategies that activate more
analytic and effortful inferential processes.
According to the Kintsch model, there are levels of process-
ing that correspond to the mental representation of meaning for
. . . students find the comprehension of discourse, which result in a textbase and a situation model of
expository text more difficult than narratives. what has been read. The textbase is all the information
expressed in the text by the author. A complete textbase
requires the reader to engage in processes that build coherence
reading speculate that it is no coincidence that the 4th-grade or the logical connection of ideas, using processes such as
slump and subsequent decline in comprehension performance bridging inferences that link pronominal referents to the appro-
co-occurs with the introduction of expository text in class con- priate person or item to which the pronoun is referring within
tent. Prior to that, children’s experience with classroom text is and across sentences, and between paragraphs across text. In
almost exclusively in narrative genre, especially stories. addition, they use logical thinking processes to determine
Narratives follow a format that even young children have cause and effect relationships discussed in the text. Building
expertise in understanding. Expository text, on the other hand, the textbase gives the reader a coherent memory trace that they
provides new information and contains unfamiliar ideas, some can use to recall text units, but by itself, it does not represent a
of which are communicated with low frequency, often abstract, deep level of processing or comprehension. For that, reader
vocabulary words and concepts. Research studies in the U.S. needs to build a situation model. In building a situation model,
repeatedly report that students find the comprehension of the reader actively constructs a meaning of the text that com-
expository text more difficult than narratives (e.g., Duke, 2000; bines the textbase representation with their prior knowledge—
Best, Floyd, & McNamara, 2004; Snyder & Caccamise, in to create new knowledge.
press). A reader with poor background knowledge will have diffi-
Educators have come to realize that teaching students to culty building a situation model. Conversely, good readers with
decode, typically a focus of K-5 reading programs, is necessary background knowledge actively make inferences between the
if one is to learn to read. It is, nevertheless, not sufficient if textbase they have constructed and information stored in their
students are to understand the educational content they need long term memory, accessing these relevant pieces of knowl-
to advance through middle and high school coursework. edge, as they construct a situation model. These good readers
In the following sections, we will briefly describe the lead- use comprehension strategies such as
ing evidence-based theory of reading comprehension processes Continued on page 24

Acknowledgement: The work on this paper has been supported, in part, by The McDonnell Foundation, award #16084757L to the lead author. Any opinions, conclusions, and
recommendations expressed in this publication are those of the authors and do not necessarily reflect the views of the McDonnell Foundation.

The International Dyslexia Association Perspectives on Language and Literacy Spring 2009 23
Comprehension Processes continued from page 23

• summarizing the ideas in their own words Stand-Alone Versus Embedded

• identifying the main ideas Comprehension Strategy Instruction
• self questioning about the ideas in the text According to the findings of a national consortium of lead-
• using graphic and semantic organizers ing researchers in reading and education reported by Snow
• monitoring their comprehension (2002) for RAND, reading comprehension research in 2002
These are active processes the good reader uses to create his was not sufficiently elaborated to warrant major reform of read-
or her own unique meaning of the text. These processes will ing comprehension instruction at that time. The consortium
update the reader’s knowledge in memory, and thus deep com- reported that not much progress has been made regarding
prehension and learning occurs. When readers process text at content-area comprehension. They observed that while teach-
this level, they are able to use and build on the knowledge by ing reading comprehension strategies independent of content
using other thinking processes such as abstraction, generaliza- classes was possible, there did not seem to be transfer of these
tion, and application. These higher order processes allow read- new strategy skills to content-area classrooms. Thus, a question
ers to think about the bigger picture where this information that has not yet been thoroughly researched has to do with the
might apply and to use the information in new ways for new format of comprehension instruction. Is it better to teach read-
purposes. ing comprehension strategies as part of a reading/language arts
Good readers with poor background knowledge engage in class where the strategies themselves are the main focus of the
the same active comprehension processes to begin the knowl- learning, or is it better to teach reading comprehension strate-
edge building that will become their future bank of background gies as part of the curriculum in content area classes that
knowledge they can access for subsequent discourse process- depend upon text, such as science, history, social studies, etc.?
ing. Thus, background knowledge is separate from comprehen-
sion strategies, which can be viewed as a set of problem solving
skills that good readers use to understand difficult text. It takes
more than just knowledge to achieve the appropriate integration
. . . background knowledge is separate from
of information into a situation model. Difficult text often requires comprehension strategies . . .
that the reader use their knowledge strategically. Skilled readers
are always monitoring their comprehension, a process that falls
into the category of metacognition, where the reader is aware of Or, does it make a difference which way students learn about
his or her own cognitive processes and is able to monitor and reading comprehension strategies as long as they learn them?
regulate them. One of these, comprehension monitoring, occurs In this special issue, Margaret McKeown, and her co-authors
as good readers engage active processes to gain the most mean- report findings that argue for embedding reading comprehen-
ing from discourse, and to repair any comprehension failures. sion instruction into content curriculum. Danielle McNamara,
The comprehension effort that good readers employ requires also in this issue, has had success with her stand-alone strategy
active use of the comprehension strategies mentioned above. instruction using computer mediated tutors. Since building and
These strategies are employed through activities such as using background knowledge is an important piece in creating
• making inferences to connect background knowledge deep comprehension, we theorize that comprehension strategy
with the textbase instruction that is embedded in content area classes may
• rereading to clarify the textbase by engaging in backward emerge as having a decided advantage, but we think that the
and forward search strategies to cross-reference and necessary evidence to support that advantage is just beginning
identify information to verify or revise an interpretation to surface from scaled1 scientifically based studies.
• questioning to fill in gaps in the situation model At the present time, there is a serious practical problem with
• summarizing to galvanize a coherent gist or situation embedding reading instruction in content curriculum. We
model observed this during a 5-year scaling project testing the use of
• questioning the author’s frame of reference and goals a computer-based summarization tutor. Course goals and the
• deciding whether to interpret the information literally or rubric by which student and teacher performance are evaluated
figuratively tend to be based on subject matter specifically relevant to the
These strategies and activities are particularly helpful for content area of a course. Currently, in content courses such as
expository text because it does not have a story schema that science, history, and social studies, reading comprehension is
allows the reader to automatically fill in gaps. Rather, a good often just assumed to occur when a student is provided reading
reader must employ active reading comprehension skills if they material and instructional goals. Consequently, we often
hope to maximize the learning from informational text (e.g., received calls exclusively from language arts teachers, who
Pressley, 2000). wanted to participate in our study to improve comprehension

1 Scaled research is when the well-developed educational intervention is tested in many, diverse classrooms to determine whether, the intervention that worked so well in a
laboratory setting or just a couple of monitored classrooms, will work as well in any classroom without researcher support.

24 Perspectives on Language and Literacy Spring 2009 The International Dyslexia Association
by teaching better summarization skills. These teachers, after the textbase and situation model for the lesson’s text. The pro-
all, had state and district mandated instructional goals by gram uses discussion, graphic organizers, the application of
which they were specifically evaluated. Their teaching perfor- knowledge, and writing to do this. As the students work their
mance was judged by how well their students learned to sum- way through the chapters, they build content knowledge right
marize. Our research objectives, however, also led us to con- along with skills in the use of comprehension strategies.
stantly strive to “infiltrate” content area classes with our sum- • strategies consistent with the CI model of comprehen-
marization tool, but this was generally viewed as a “luxury” in sion that are emphasized in this program include
an already over-filled classroom hour designed to meet perfor- • summarizing, that is, restating the text in the readers
mance goals in the content area. But neither we, nor other own words. This strategy requires readers to grasp the
reading comprehension researchers, are put off by these prac- gist of the text by creating a situation model.
ticalities. Below, we report on a couple of promising compre- • explaining how ideas and sections of text relate to one
hension programs that are designed to be embedded in content another, that is, building text coherence
area curriculum, including our work on summarization. • developing domain specific vocabulary
• answering open ended questions that require domain
IDEAS specific vocabulary and making connections within the
Vitale and Romance (2007) have developed a knowledge- text
based instructional framework, called IDEAS that unifies con- • developing an awareness of text structures so that stu-
tent-area reading comprehension with reading comprehension dents can develop metacognition that will help them
strategies. They successfully implemented a combined read- build a coherent textbase
ing—science content area 2-hour class in a grade school set- • creating graphic organizers that display the ideas in the
ting—grade 5, which involved over 50 teachers and 1200 eth- text and the students’ understanding of this information
nically diverse students. This class replaced the reading curricu- and how the ideas relate
lum at these schools. Control classrooms were matched to • discussing related ideas across texts (i.e., ideas that are
experimental classrooms. The students in control classrooms independent but are related by theme)
studied reading comprehension in the context of the usual dis- • writing and discussion to demonstrate understanding
trict basal reading program which focused on narrative texts. Currently, this program has been successfully implemented
The students in the experimental classrooms studied reading in a few classrooms as a supplemental program—after school
comprehension strategies embedded in their science instruc- or pull-out—and does not replace a school’s reading curricu-
tion. Their tasks included three main types of strategies: text lum or English language arts program. However, the format
elaboration via discussions, which included the identification could be adapted to content classes. Our research lab is look-
of key ideas, question answering to link the text and engage it ing into this theory-based program to gain further understand-
with students’ prior knowledge; concept mapping, in which ing of the factors that provide the most efficient and effective
students identify, arrange and link together the core ideas, sub- application of the constructs of the CI model. More specifically,
ordinate ideas, and examples; and summarization/writing, in we are evaluating the program with regard to concrete improve-
which students organize and communicate the ideas generated ment in and therefore transfer of comprehension skills learned
from the text. in executing these lessons as well as an increase in learning the
In IDEAS, teachers model the appropriate comprehension knowledge that makes up the reading content and its transfer to
strategies and monitor the students’ use and progress using subsequent related learning.
the strategies themselves. This program required a reported The summarization tool we reported earlier in this article is
30 hours of training during the summer and 30 hours of plan- called Summary Street, and is currently offered commercially
ning time on the part of the teachers, along with continued by Pearson Knowledge Technologies, (http://www.pearsonkt.
mentoring from the researchers. A similar level of professional com/). Our research with this application found that if students
development was required by, Literacy Navigator, the next used this computer tutor, their ability to summarize was sig-
application we discuss, which also focuses on content-based nificantly improved as compared to matched control students
comprehension instruction. who only used a word processing application to summarize the
same material (Caccamise, Franzke, Eckhoff, & Kintsch, 2007).
Literacy Navigator Summarization is only one comprehension strategy that good
Literacy Navigator (Hampton, 2007), offered by America’s readers use to more deeply comprehend material, but an
Choice, is a teacher-led, instruction program designed to important one.
improve reading comprehension in grades 4-10. It is designed
for students who can adequately decode text, but nevertheless, Future Directions
struggle to comprehend informational text and have problems The National Science Foundation has identified personal-
on state reading tests. This program is based on Kintsch’s ized learning as a 21st century research initiative. As future
Construction Integration (CI) theory of reading comprehension. research continues to affirm content-embedded reading com-
The lessons are organized around a single topic (in this case, prehension programs, we envision a time when not only content
adaptation) and generally progress from easy to more difficult and comprehension training will be integrated, but assessment
content. Each lesson is specifically targeted to demonstrate and and performance accountability will also be a seamless part of
engage students in the active use of reading strategies to build Continued on page 26

The International Dyslexia Association Perspectives on Language and Literacy Spring 2009 25
Comprehension Processes continued from page 25

this process. The integration of these components will bring both Vitale, M. R., & Romance, N. R. (2007). A Knowledge-Based framework for unifying
content-area reading and comprehension and reading comprehension strategies.
efficiency and a higher level of effectiveness in the education In D.S. McNamara. Reading comprehension stragegies: theories, interventions
practices of tomorrow (Caccamise & Snyder, 2005, Caccamise, and technologies. (pp. 73–104). NY: Erlbaum-Taylor and Francis Group.
Snyder and Kintsch, 2008). Teachers will be an indispensible
catalyst and facilitator in making it all happen.
Donna Caccamise A.B. (SDSU), M.A., Ph.D. (Univ. of
References Colorado), Psychology. Dr. Caccamise is the Associate
Best, R. M., Floyd, R. G., & McNamara, D.S. (2004). Understanding the fourth-grade
slump: Comprehension difficulties as a function of reader aptitudes and text Director of the Institute of Cognitive Science, and the
genre. Paper presented at the 85th Annual Meeting of the American Educational Director of Academic Programs for ICS at the University of
Research Association.
Colorado, Boulder. She is cognitive psychologist/cognitive
Caccamise, D., Snyder, L., & Kintsch, E. (2008). Constructivist theory and the situa-
scientist whose research is at the nexus of cognition, and
tion model: Relevance to the future assessment of reading comprehension. In C.
Block & S. Parris, Comprehension instruction. NY, NY: Guilford Press. computer mediated educational interventions. Her recent
Caccamise, D., Franzke, M. Eckhoff, A., & Kintsch, W. (2007) Automated tutor for work focuses on reading comprehension issues and instruc-
teaching summarization and deep comprehension. In Mc Namara, D., ed. tional applications in grades K-14.
Reading Compre. Strategies: Theory, Interventions, and Technologies. Erlbaum.
Caccamise, D., & Snyder, L. (2005) Theory and pedagogical practices of text com-
prehension, Topics in Language Disorders, 25, 5–20.
Lynn Snyder, B.A., M.A., CCC-SLP, Ph.D., is Professor
Duke, N. K. (2000). 3.6 minutes per day: The scarcity of informational texts in first
grade. Reading Research Quarterly, 35, 202–224.
Emerita in the Department of Speech, Language and
Hampton, S. (2007). Literacy navigator. Washington, DC: America’s Choice
Hearing Sciences and Director Emerita of the Center for
(National Center on Education and the Economy). Language and Learning at the University of Colorado at
Kintsch, W. (1998). Comprehension: A paradigm for cognition. New York: Boulder. Her research focuses on reading comprehension in
Cambridge University Press. children and adolescents. She is a Fellow of the American
Organisation for Economic Cooperation and Development, (2007). Report of the Speech-Language-Hearing Association and the International
Progress in International Reading Literacy Study, Programme for International
Academy for Research in Learning Disabilities.
Assessment, November 28, 2007.
Pressley, M. (2000). What should comprehension instruction be the instruction of?
In M.L. Kamil, P.B. Mosenthal, P.D. Pearson & R.Barr (Eds.), Handbook of reading Correspondence may be sent to Donna Caccamise, Ph.D.,
research, (Vol. 3, pp. 545-561). Mahwah, NJ: Erlbaum. Institute of Cognitive Science, University of Colorado at
Snow, C. E. (2002) Reading for understanding: Toward a research and development Boulder, Campus Box UCB 344, Boulder, CO 80309-0344
program in reading comprehension. Santa Monica, CA: RAND. e-mail: [email protected]
Snyder, L., & Caccamise, D. (In preparation) Comprehension Processes for
Expository Text: Building Meaning and Making Sense. In M. Nippold and C. Scott
(Eds.)., NY: Taylor and Francis.

26 Perspectives on Language and Literacy Spring 2009 The International Dyslexia Association
View publication stats