> Reading Comprehension
Do Hi-Low Texts Support Comprehension?
More Closely at High-Interest, Low-Level Texts: Do They Support
by Stephanie A. Spadorcia, Ph. D
Reprinted with permission from the International
Dyslexia Association quarterly newsletter, Perspectives, Spring
2001, vol. 27, no. 2, pages 32-33, Stephanie Spadorcia,, Ph.D.
(It's worth joining IDA just to get Perspectives, and there
are many other benefits. Their website is http://www.interdys.org.)
Adolescents with reading disabilities can lack skill in any
of several areas (e.g., decoding and language usage, knowledge
of text and story patterns, making predictions, recognition
of textual elements, and monitoring of their own comprehension).
However, over the years, attention to the reading materials
available for adolescents with reading problems has been sporadic.
Up until recently, the only empirical research with a focus
on materials for older struggling students was conducted 20
years ago (see Mason, 1981). As a result, educators often
find themselves choosing from a selection of books that may
or may not be supportive of their instruction.
Texts can be viewed in terms of three different levels that
support and pose challenges to readers: word-level, sentence-level,
and passage-level (Kintsch, 1998). The word-level consists
of the words on the page as individual units that have to
be recognized. The sentence-level combines these words using
rules of syntax and semantic structure to form idea units.
The passage-level contains the network of connections among
the ideas, or the extent to which the ideas "hang together"
in a cohesive manner. All three levels proved context and
structure to the reader while reading for comprehension. And
all three levels provide a framework for looking more closely
at high-low books in terms of the support they offer to struggling
For instance, some high-low books are written to achieve
a targeted level on a readability formula. The assumption
is that if the word-level and/or sentence length is controlled,
then a reader will be able to comprehend the text. But areas
of difficulty for a student can lie beyond those characteristics.
At the sentence level, a less proficient reader may use language
skills in order to access the meaning of words (Hiebert, 1999).
High-low books written around readability formulas might give
less support to this process -- since the linguistic complexity
of these books at the sentence level is lower -- therefore
Coherence is also often ignored in the creation of these
books (Davison & Kantor, 1982). Many consistently exclude
linguistic markers of highly coherent text (Spadorcia, 2001).
Low coherence does not support less proficient readers' deep
comprehension of a book (McNamara, et al. 1996).
Analyzing High-Low Text
Recently I analyzed 60 high-low books written for adolescents
using a three-tiered framework of textual demands (Spadorcia,
2001). Word level demands were looked at with the
percentage of high frequency words and the percentage of decodable
words (from an onset-rime perspective). Sentence-level
demands were measured by sentence length and t-unit length.
Passage-level demands were evaluated with the use of
a holistic scale of coherence.
I found four categories of high-low books based on their
commonalities across the levels of textual demand: 1) decodable
texts; 2) literature-based texts; 3) readability-formula texts,
and 4) balanced texts.
Decodable Texts. Books within this group consistently
had a low level of high frequency words, a moderate number
of highly decodable words, low sentence-level sophistication,
and a moderate to low overall passage cohesion.
Books typical of this cluster included: France, M. (1998).
Deadly double. Novato, CA: High Noon Books.
This type of text would hinder comprehension for many less
proficient readers who rely on high sentence-level complexity
to identify unknown words and chain together idea units. But
only 8 % of the books I analyzed fell within this grouping,
indicating that few high-low texts control textual demands
in this way.
Literature-based text. This category of texts contained
high sentence-level complexity and overall coherence. But
this type of text did not seem to place parameters around
the word-level demands posed to readers. That is, equally
low numbers of high frequency and decodable words were found.
Therefore, a less proficient reader would have more word-level
demands to deal with, but the sentence- and passage-levels
would presumably provide more support for comprehension.
Books typical of this category included:
Brin, S. (1998). Trapped in the sixties. Costa Mesa,
CA: Saddleback Publishing, Inc.
Griffey, H. (1998). Secrets of the mummies. New York:
DK Publishing, Inc.
Gutelle, A. (1990). Baseball's best five true stories.
New York: Random House.
Books within this cluster also tended to contain high numbers
of "high-content" words. For example, Knucklehead
(Capstone PRess, 1993) tells the story of a young teenage
boy who just moved to town. He attempts to get involved with
the school community by playing on the high school baseball
team. To comprehend this book, the reader not only needs background
knowledge about baseball, but must be able to recognize less
frequent, content-related words like "clubhouse,"
"knuckleball," and "pitch."
Twenty-seven percent of the high-low books I analyzed fell
into this category.
Readability-formula text. Books in this category
contained shorter sentences and lacked syntactic complexity.
Overall coherence was also low, probably because words were
chosen and sentences were written without regard to their
impact on overall cohesion.
Typical books included:
Pageler, E. (1997). The rich cake mystery. Novato,
CA: High Noon books.
Schultz, I. (1996). The great dinosaur hunt. Bothell,
WA: The Wright Group.
Wright, B. (1995). The haunted house mystery. Novato,
CA: High Noon books.
Since syntactic and semantic systems provide the connection
between oral and written communication (Hiebert, 1999), books
written in this manner can make it difficult for students
to make that connection and draw upon those systems when reading.
For example, consider the following excerpt from The Haunted
House Mystery (Wright, 1995):
"I'm not going," Tom said. "This is day time,
Tom. It will be OK," Eddie said.
"Anyway, we aren't going in it," Ricky said.
"We aren't? Why not? It looks okay to me," Tom said.
"You can't get in it. There's a big fence around it.
That's how they keep people out," Ricky said. (pg. 11-12)
Sections in these books were often not clearly introduced
or summarized with a closing statement. For instance, in the
same book, a chapter ended with the following sentence: "I
think those men are printing funny money there at night, he
said" (pg. 39). On the following page the chapter stared
with "Sergeant Collins didn't say anything. Everyone
just looked at Ricky" (pg. 40). There was no change in
the setting or action between these two chapters. The break
was blunt and not presented in a seamless manner, thereby
contributing to a text with low overall passage coherence.
Thirty-five percent of the books I analyzed fell into this
Balanced Text. Books in this category contained
equal amounts of high frequency words and regularly decodable
words, as well as engaging content and language, providing
students with a balance of opportunities for increasing reading
comprehension. As compared to the other categories, books
from this category were second in sentence-level sophistication
only to the literature-based category, and first overall in
Typical books included:
Bricker, S.D. (11993). Challenger. Costa Mesa, CA:
Saddleback Publishing, Inc.
Schraff, A. (1992). Swamp Furies. Costa Mesa, CA:
Saddleback Publishing, Inc.
Schraff, A. (1993). The phantom falcon. Costa Mesa,
CA: Saddleback Publishing, Inc.
Thirty percent of the books I analyzed fell into the Balanced
Each day teachers make on the spot decisions about reading
materials, whether for group reading instruction, or for students
who need independent reading materials. Often these decisions
are made while looking only at one factor, such as length
of book, type of words included, or interest connection for
the student (Hiebert, 1999). Customarily, issues such as linguistic
complexity and overall coherence are not taken into account
(Davison & Kantor, 1982).
To assist students in comprehension development, multiple
factors should be considered before a selection is made. Using
readability indices or word counts alone is not sufficient.
Instead, books should be examined through a multiple set of
lenses, recognizing the particular skills that less proficient
students need to learn to become better readers.
Hiebert (1999> speaks of "multiple-criteria texts"
for reading instruction at the primary grades. Within this
framework different book types are utilized in a cycle to
allow for practice in different areas of reading skills, including
increasing sight-word recognition and decoding skills, using
context to figure out unknown words, improving response to
text, and discussion. Older struggling readers would benefit
from the same approach.
Teachers of struggling readers face a unique situation and
have unique needs in locating materials for instruction and
reading. But there is a wealth of information that could be
shared between literacy and special educators in order for
headway to be made with this problem. Much information is
known about creating and utilizing effective reading for younger
beginning readers. The problems facing older struggling readers,
regardless of disability, are not different from this wealth
of knowledge, just foreign to the application of it. Under
the umbrella of collaboration between literacy and special
educators, as well as book publishers and authors, much progress
can be made in order to create materials that complement effetive
teaching for this group of students and their teachers.
Davison, A. & Kantor, R.N. (1982). On the failure of
readability to define readable texts: A case study from adaptations.
Reading Research Quarterly, 17, 187-209.
Hiebert, E.H. (1999). Text matters in learning to read. The
Reading Teacher, 52, 552-566.
Kintsch, W. (1998). Paradigms of comprehension. England:
Mason, G.E. (1981). High interest-low vocabulary books: Their
past and future. Journal of Reading, 24, 603-607.
McNamara, D.S., Kintsch, E., Songer, N.B., & Kintsch,
W. (1996). Are good texts always better? Interactions of text
coherence, background knowledge, and levels of understanding
in learning from text. Cognition and Instruction, 14,
Spadorcia, S. A. (2001). Analyzing the word-level, sentenc-level,
and passage-level of high-interest, low-level books. Manuscript
submitted for publication.
Stephanie A. Spadorcia, Ph.D. is an assistant professor
in the School of Education and a learning disabilities specialist
at the Center for Academic Achievement, at Lesley University,
Cambridge, MA. She is part of the faculty of the Center for
Literacy and Disability Studies. Her research interests include
the compatibility of reading materials and instruction, high-interest,
low-level texts for struggling readers, and assessment and
instruction of literacy skills for children with disabilities.