Administrations of reading tests by NAEP (National Assessment of Educational Progress) have since 1971 confirmed what has long been part of the commonsense and intuitive knowledge of both teachers and laypeople: Children from more economically advantaged families score significantly higher than the less advantaged at all ages tested (nine, 13, and 17), and the gaps become greater with increasing age. The questions are: Why do these differences occur? Why are they so enduring and so universal?
A Developmental Model of Reading
In Chall's stages of reading development (1983, 1996), reading is conceptualized not as a process that is the same from beginning stages through mature, skilled reading, but as one that changes as the reader becomes more able and proficient.
Changes in reading development fall into six stages—from Stage 0 (prereading) to Stage 5 (the most mature, skilled level of reading in which readers construct and reconstruct knowledge from their own reading). Generally, Stages 1 and 2 (typically acquired in grades 1, 2, and 3) can be characterized as the time of "learning to read"—the time when simple, familiar texts can be read and the alphabetic principle is acquired (i.e., readers are able to decode words they do not immediately identify, and they become fluent, especially when reading texts that use language already within their experience and ability); Stages 3 to 5 can be characterized, roughly, as the "reading to learn" stages—when texts become more varied, complex, and challenging linguistically and cognitively. Beginning at Stage 3 (grades 4-8), students use reading as a tool for learning, as texts begin to contain new words and ideas beyond their own language and their knowledge of the world. Words and concepts in such material are beyond the everyday experience of children. In order to read, understand, and learn from these more demanding texts, the readers must be fluent in recognizing words, and their vocabulary and knowledge need to expand, as does their ability to think critically and broadly. If children are unable to make the transition from Stage 2 to 3, their academic success is usually severely challenged.
Using this developmental stage model of reading, we focused a research study on the critical transition from Stage 2 to Stage 3—from "learning to read" to "reading to learn." Teachers have often reported a "fourth-grade slump" in literacy development, particularly for low-income children—precisely at the point of the Stage 2-3 transition. We wanted to examine the skills and abilities of a low-income population to determine why some might meet the challenge of Stage 3 reading whereas others might not.
The subjects in the study (see Chall, Jacobs, and Baldwin, 1990) were 30 children—about 10 each from grades 2, 4, and 6, whom we followed for two years (through grades 3, 5, and 7 respectively). Low-income status was determined by the students' eligibility for a free-lunch program. Each child was given a series of individual tests of reading and language (as well as writing, but those findings are not presented here; see Chall and Jacobs, 1983). The reading measures were scores on the six subtests of the experimental version of the Diagnostic Assessments of Reading (DAR) (Roswell & Chall, 1992) that consist of word recognition, word analysis, oral reading, word meaning, reading comprehension, and spelling.
The most significant finding of the study for reading was that low-income children in grades 2 and 3 achieved as well as children in the normative population on all six subtests. However, as predicted by the theoretical model of reading used for the study, some of the students' scores started to decelerate around grade 4. This "fourth-grade slump," reported often by teachers of disadvantaged children, started in grade 4 on some tests and later on other tests.
The first and strongest to slip was word meaning. The low-income children in our study—in grades 4 through 7—had greatest difficulty defining more abstract, academic, literary, and less common words as compared with a normative population on the word meaning test. In grade 4, the children were about a year behind grade norms. By grade 7, they were more than two years behind norms. Next to decelerate were their scores on word recognition and spelling. Oral reading and silent reading comprehension began to decelerate later in grades 6 and 7.
Thus, if we view reading as composed of the three basic components proposed by Carroll (1977)—cognition, language, and reading skills—cognition seemed to be a lesser problem for them than language. In grades 4 to 7, they did best on the reading tests that relied on context and required understanding—reading comprehension and connected oral reading. The children had greatest difficulty in grades 4 through 7 in defining less common words and in recognizing and spelling them—skills and abilities needed for reading at higher levels of complexity (Chall, 1983, 1996).
The language tests assessed vocabulary meaning, metalinguistic awareness, and grammatical judgment. All but the vocabulary measure (the WISC-R vocabulary subtest) were developed by Carol Chomsky.
Our population seemed to do well on measures of basic language abilities through the third grade. After the third grade, they began to decelerate first in knowledge of the meanings of words (on the WISC-R as on the DAR)—especially the less common, more academic words found in books used in the intermediate and upper elementary grades and higher. The children's basic linguistic competencies, as shown in their grammar and language awareness scores, were stronger than their word knowledge, especially after the third grade.
In sum, the low-income population in our sample achieved as well in literacy and language as a normative population through the third grade. Beginning with the fourth grade, however, they exhibited signs of a slump. First and strongest to decelerate was word meaning. The slowest to decelerate were reading comprehension and oral reading.
Causes of the Fourth-Grade Slump
One possible reason for the fourth-grade slump may stem from lack of fluency and automaticity (that is, quick and accurate recognition of words and phrases). We found this particularly among the poorest readers who read slowly and hesitatingly in grade 2 and beyond. Lack of fluency tends to result, ultimately, in children's reading less and avoiding more difficult materials (see Chall, 1983, 1996; Stanovich, 1986).
The deceleration of the scores on word meaning, beginning with grade 4 and continuing through grade 7, suggests a difficulty that, if not improved, ultimately affects children's reading comprehension as well. The high correlation of word knowledge with reading comprehension has been found consistently in the research literature from the turn of the century to the present time (see Anderson & Freebody, 1981; E.L. Thorndike, 1917; R.L. Thorndike, 1973-1974).
It is important to realize that the vocabulary scores of our population, when they were in grades 2 and 3, were on a par with the general population. At these levels, the words tested were of high frequency and were familiar. It is when the words became less common (at grade 4 and beyond) that they presented a problem.
Importantly, too, the vocabulary lag did not at first seem to affect negatively the children's silent reading comprehension scores. Although their vocabularies decelerated in grade 4, their comprehension still held up well against norms. It was at grade 6 that their comprehension began to decline. This suggests that the students used context well to compensate for their weakness in word meanings. But when there were too many difficult words, their comprehension declined as well.
Why should low-income children have greater difficulty with word meanings at about the fourth grade? One reason is that the words at fourth grade and above are less familiar. Although the children's language seemed to have been sufficient for the first three grades, they were not prepared to meet the challenge of the greater number of abstract, technical, and literary words characteristic of the reading materials of grades 4 and beyond. Such language—often termed literary and abstract—is more complex than that used by children in everyday, oral interaction.
* * *
A follow-up study of our low-income children, five years later when they were in grades 7, 9, and 11, found decelerative patterns of scores similar to those the children exhibited when they were in the elementary grades (see Snow et al., 1991). On most tests, their scores were below norms, and the discrepancies between their scores and norms were greater in each succeeding grade. By grade 11, their reading scores were in the 25th percentile—considerably below their achievement in grades 4 through 7.
Similar to the deceleration in the earlier grades, the seventh- and ninth-grade students decelerated slowest on tests that permitted the use of context. However, by the 11th grade, even their reading comprehension scores had fallen. Thus, similar to their achievement in the earlier grades, the children scored relatively well, as long as their meaning vocabularies and word recognition were not too weak and when they could compensate through the use of context on reading comprehension. But when there was too great a mismatch between the students' word-meaning knowledge and the difficulty of the text, their use of context was not sufficient to help them to understand the text.
The trends for grades 2 through 7 and grades 7 through 11 suggest that we cannot be sanguine when students do well in silent reading comprehension despite apparent difficulty in word meanings and word recognition. As predicted by the model of reading development, if children are lacking in certain aspects of reading, later reading development will usually suffer. Because of the developmental nature of reading, the later one waits to strengthen weaknesses, the more difficult it is for the children to cope with the increasing literacy demands in the later grades. Moreover, those who have reading difficulties in the intermediate grades will, most likely, have serious trouble with the study of science, social studies, literature, mathematics, and other content study that depend, in great part, on printed text.
The late Jeanne S. Chall is remembered as a great researcher who devoted over fifty years to helping children learn to read well. She authored such classics as Learning to Read: The Great Debate and Stages of Reading Development. Vicki A. Jacobs is associated director of the Teacher Education Program and lecturer at the Harvard Graduate School of Education. This article is based on research originally presented in The Reading Crisis: Why Poor Children Fall Behind by Jeanne Chall, Vicki Jacobs, and Luke Baldwin; it is excerpted with permission from "Reading, Writing, and Language Connection" in J Shimron (ed.) Literacy and Education: Essays in Memory of Dina Feitelson. Cresskill, N.J.; Hampton Press, Inc. 1996, pp. 33–48.
Anderson, R.C., & Freebody, P. (1981). Vocabulary knowledge. In J. T. Guthrie (ed.), Comprehension and teaching: Research review (pp. 71-117), Newark, Del.: International Reading Association.
Carroll, J.B. (1977). Developmental parameters of reading comprehension. In J.T. Guthrie (ed.), Cognition, curriculum, and comprehension (pp.1-15). Newark, Del.: International Reading Association.
Chall, J.S. (1983). Stages of reading development, New York: McGraw-Hill.
Chall, J.S., & Jacobs, V.A. (1983). Writing and reading in the elementary grades: Developmental trends among low-SES children. Language Arts, vol. 60, no. 5, pp. 617-626.
Chall, J.S., Jacobs, V.A., & Baldwin, L.E. (1990). The Reading Crisis: Why Poor Children Fall Behind. Cambridge, Mass.: Harvard University Press.
Chall, J. S. (1996). Stages of reading development (2nd ed.). Fort Worth, Tex.: Harcourt Brace.
National Assessment of Educational Progress (NAEP) (1971). 1969-1970 writing: Group results for sex, region, and size of community. Washington, D.C.: U.S. Government Printing Office.
Roswell, F.G., & Chall, J.S. (1992). Diagnostic assessments of reading and trial teaching strategies (DARTTS). Chicago: Riverside Publishing.
Snow, C.E. Barnes, W.S., Chandler, J., Goodman, I.F., & Hemphill, L. (1991). Unfulfilled expectations: Home and school influences on literacy, Cambridge, Mass.: Harvard University Press.
Snow, C.E., & Dickinson, D.K. (1991). Skills that aren't basic in a new conception of literacy. In A. Purvis & T. Jennings (eds.), Literate systems and individual lives: Perspectives on literacy and schooling (pp. 175–213). Albany: State University of New York Press.
Stanovich, K.E. (1986). Matthew effects in reading: Some consequences of individual differences in the acquisition of literacy. Reading Research Quarterly, 21, (pp. 360–407).
Thorndike, E.L. (1917). Reading as reasoning: A study of mistakes in paragraph reading. Journal of Educational Psychology, 8, (pp. 323-332).
Thorndike, R.L. (1973–74). Reading comprehension education in fifteen countries (International Studies in Evaluation III). Stockholm, Sweden: Almquist and Wiksell.
Reading Comprehension Requires Knowledge—of Words and the World (PDF)
Scientific Insights into the Fourth-Grade Slump and the Nation's Stagnant Comprehension Scores
By E. D. Hirsch, Jr.
The Classic Study on Poor Children's Fourth-Grade Slump
By Jeanne S. Chall and Vicki A. Jacobs
How Words Are Learned Incrementally Over Multiple Exposures (PDF)
By Steven A. Stahl
Oral Comprehension Sets the Ceiling on Reading Comprehension
By Andrew Biemiller
Basal Readers (PDF)
The Lost Opportunity to Build the Knowledge That Propels Comprehension
By Kate Walsh