Department of Reading Education and Special Education
Appalachian State University
Submitted to Teaching Exceptional Children November ??, 2012.
DO NOT REPRODUCE, CITE, OR CIRCULATE WITH PERMISSION OF THE AUTHORS.
“You spelled the word, cat, Mark. We want the word, cap. What letter do you need to change in cat so that it spells cap? He wears a baseball cap.” Mark repeated both his word and the target word aloud and then carefully examined his word and the remaining letter cards. Finally he removed the –t and replaced it with a -p. He looked up and said, “Cap.”
“You’re right, Mark. You spelled the word, cap. Here’s the word card that spells cap, c-a-p. Just like yours. Let’s put cap in the pocket chart with the other three-letter words.”
Mark’s face changed from a look of concentration to that of celebration, and he cheered himself on with a self-congratulatory, “Woohoo! Gooooo Mark!” The process described above was repeated throughout a five-week pilot study of a phonics intervention, always with similar degrees of high student engagement and effort, intensive teacher feedback and guidance, and high expectations of success. Each day, Mark and three of his classmates met one-to-one with the first author for 20-minute sessions in which they engaged in systematic, guided invented spelling with targeted letters and words. The curriculum, Making Words (Cunningham & Cunningham, 1992; Cunningham & Hall, 2007), was based on scientific research evidence (National Reading Panel, 2000; Cunningham & Cunningham, 2002). The instruction, implemented with fidelity, was scientifically-based.
And student progress was negligible. None of the students showed substantial growth in their awareness of letter-sound correspondences across multiple measures. For example, on a kindergarten spelling list, the four students had 10 of 48 initial letter sounds logically represented at pretest and eight of 48 at posttest (e.g., watk for wish or B for bump). They had 8 medial or final letter sounds logically represented at pretest and 11 at posttest.
We found this outcome both surprising and troubling for two reasons. First, it played to an existing bias in the field regarding literacy instruction appropriate to students with a diagnosis of Down syndrome and moderate to severe intellectual disabilities. Widely-employed instructional materials are designed on the belief that such students cannot learn phonics and should be taught exclusively through sight word approaches (e.g., Edmark, 2011; Oelwein, 1995). Methods textbooks identify students like Mark and his classmates as those “who will not achieve the outcome of literacy,” although they “can benefit from sight word instruction” (Browder, 2001, pp. 179-180).
Second, Mark and his classmates attend school in a state adhering to the Common Core State Standards. In just a few years they are supposed to be reading at levels sufficient for college or career readiness, although at present they are best characterized as emergent readers. Mark and his classmates are soon to be engaged in higher-order thinking and rigorous content, although, despite the best efforts of schools for more than 10 years, they cannot currently spell words like cap without guidance and feedback.
As educators, we found that attributing the failure to Mark and his classmates provided an uncomfortable dead end. While it relieved us of responsibility for the students’ failure to respond, it also failed to suggest directions for future interventions. The RTI model provided no guidance, since the tiers of instruction had been exhausted, and Mark and his classmates were already receiving special education services. We feared the self-fulfilling prophecy of assuming students like Mark could not learn phonics; clearly they would not if we offered them no further phonics instruction and pursued instead a functional sight word approach.
Consequently, what remained was an assumption that the instruction we offered must be the source of the problem, but how could that be since we had selected scientifically-based instruction and implemented it diligently with fidelity? In order to answer these questions, we began a thorough review of our preparations for the intervention study. First, we reviewed the literature to make sure the intervention was indeed evidence-based.
The Scientific Basis of the Intervention
We had selected Making Words First Grade (Cunningham & Hall, 2007) as the intervention to be implemented, because we knew the students, while adolescents, were beginning readers. Making Words is described as a “guided invented spelling instructional strategy” (Cunningham & Cunningham, 1992, p. 107). It is organized on evidence-based practices in phonics that suggest explicit and systematic phonics instruction is superior to nonsystematic approaches, but that no specific approach fitting those criteria is superior to others (National Reading Panel, 2000; Stahl, Duffy-Hester, & Stahl, 1998). It employs evidence-based practices in phonemic awareness also, requiring students to blend and segment sounds and to do so by using letters, which increase the efficiency of acquisition of phoneme awareness and phonics (National Reading Panel, 2000).
Fidelity of Implementation
Next we reviewed our teaching procedures to make sure that we had implemented the intervention with fidelity. Each Making Words lesson consisted of three steps. First, in the guided invented spelling step, students were given a pre-determined set of six to eight letters of the alphabet and led in spelling 10-15 words of increasing length. Students were directed to form certain words, beginning with two-letter words and then building up to increasingly longer words. For example, in the first step, students with the letters a, t, m, s, r, h, were told, “Take two letters and spell the word at. We are at school. At.” The students were then shown the word card and asked to compare their attempts to the model. If they were correct, they then put the word in a pocket chart. If they were incorrect, they self-corrected or were directed to self-correct (as in the vignette above) before putting the word in the pocket chart. Then they were directed, “Add one letter and spell hat. You have a baseball hat. Hat.” The process was continued, making two-letter words, three-letter words, and so on. For each new word, the students were guided to add or change one letter or to rearrange the existing letters (e.g., rearrange rat to spell art).
Next, in the sorting step, the students were supported in identifying patterns in the words they had spelled. For example, they might sort the words they spelled in the first step into those beginning with a vowel (e.g., as, am, at; is, it) or those that rhyme (e.g., at, sat, hat, rat, mat; am, ram, ham, Sam). Typically two patterns were identified and words sorted appropriately.
Finally, in the transfer step, students were directed to spell specific new words that shared rhyming patterns with the words they had previously sorted. That is, the intent was to teach them to generalize the use of patterns, to compare and contrast in order to read and spell unfamiliar words. The students might be given directions such as, “If this is how you spell am, ram, ham, and Sam, how do you think you could spell bam? Jam?” or “Which sorting pattern would help you spell bam? Jam?” Typically three to four new words were attempted in this way.
Our procedures were consistent with previous descriptions of the approach, directions in the front matter of the book, and sample scripted lessons (Cunningham & Cunningham, 1992; Cunningham & Hall, 2007).
A Theory of Reading Development
We returned to the literature to explore theories of how students become beginning readers and found an answer that helped us understand the failure of scientifically-based instruction in this instance. In studying Morris and colleagues’ theory of reading development (Morris, 1993; Morris, Bloodgood, Lomax, & Perney, 2003), we learned that Mark and his classmates required instruction targeted at a different set of skills before they would be able to benefit from the intervention we had provided. Morris and colleagues had found that beginning reading understandings develop in a predictable order (see Figure 1 below).
Figure 1. A Theoretical Model of Reading Development
Knowledge of letter names (ABC) is the earliest developing understanding. As parents and educators sing the alphabet song and point out letters in books, children’s names, and the environment, children acquire letter name knowledge in predictable ways (Justice, Pence, Bowles, & Wiggins, 2006). This knowledge of the alphabet leads children to attend to beginning consonant sounds (BC) in words, possibly because most letter names carry the sound they represent at the beginning (e.g., /b/, /k/, /t/) or ending (e.g., /f/, /m/, /r/) of the letter name (Foulin, 2005).
As children increase their knowledge of the alphabet and awareness of beginning consonant sounds, they are able to attend more specifically to print on the page when parents and teachers read aloud to them and point to words on the page. Children begin to attend to the spaces between words and develop concept of word (CW), an understanding that words are groups of letters on the page (Henderson, 1980). Additionally this growing attention to white space and understanding of beginning consonant sounds facilitates attention to the relationship of letters to sounds at the beginning and end of words (SPBE) (Morris, 1992, 1993).
These understandings, in turn, assist beginning readers in attending to the letters and sounds within words (PS), and, ultimately, as this understanding is firmed up through additional print experiences, to gain sight word knowledge (WORD) (Adams, 1990) and increase independence as beginning readers of text (READ).
From Theory to Application
When we looked at Mark’s skills and understandings, and those of his classmates, relative to the model, the problem became clear. Prior to the intervention, Mark was able to name 19 letters of the alphabet and 10 initial phonemes represented by pictures (e.g., seeing a picture of a ball and saying /b/). However, he was unable to apply that knowledge in spelling or segmenting words. He spelled words like trap as HoKicike, wish as otheweweii, and ship as Hatthew. He was able to identify just four words from the Qualitative Reading Inventory-3 (Leslie & Caldwell, 2000), and he had no strategies for reading or spelling words he had not memorized. He was able to read short texts composed of these memorized words if they were accompanied by supportive illustrations (e.g., reading aloud “The boy is in the red car” beneath a picture of this statement).
If you think of the theoretical model above as a bookshelf, Mark and his classmates had bookend knowledge. They knew letters and words, but had no knowledge they could access between those bookends. Our classroom observations led us to believe that the students had acquired this bookend knowledge primarily through repetition and memorization in isolation rather than through by application in reading or writing meaningful texts. They had little knowledge of letter-sound relations and seemed to understand concept of word only in short, memorized texts. They were unable to look within words at letter-sound relationships, because they had little meaning to attach to abstract phoneme awareness and insufficient practice reading, writing, or listening to texts.
Learning from Our Mistakes
What we learned from our mistake and subsequent research is that scientifically-based interventions in the absence of a theory of learning are insufficient. They are neither better nor worse than random selection of interventions, and they may be worse than eclectic instruction, which sometimes offsets the impact of poor instructional choices by providing a wide variety of learning experiences and opportunities. With instructional choices in classrooms increasingly restricted to the well-worn paths of evidence-based instruction, if the instruction does not match the needs of the students, then there are fewer opportunities to balance out the error.
Using the theoretical model as a planning tool, we learned that the kind of intervention most likely to yield positive literacy outcomes for Mark and his classmates would look quite different than what we had provided. Our intervention emphasized guided invented spelling of two- to six-letter words along with sorting and transferring of rhyming patterns. It was too difficult given the students’ existing understandings of print. A successful intervention would also look quite different from the kinds of instruction Mark and his classmates were receiving in their classroom, emphasizing memorization of alphabet letters and sight words and the reading of short texts composed solely of known sight words. It would even look quite different than functional literacy approaches emphasizing sight word learning in the absence of the many skills and understandings (e.g., beginning consonant knowledge or phoneme segmentation ability) and experiences (e.g., wide reading) that would lead to more efficient growth in a sight word base through application rather than drill and practice.
A successful intervention would need to emphasize the letter-sound knowledge available in letter names already known to the students. A successful intervention would need to emphasize invented spelling in writing along with reading of a greater variety of easy texts to promote the connections between concept of word and attending to letter-sound relationships in medial and final positions in words. Finally, a successful intervention would need to provide opportunities to explore and apply letter-sound knowledge in reading and spelling new words.
The mistake we have admitted to above is common in many classrooms which select evidence-based practices in hopes of assisting students in learning to read at supplemental tiers or in special education placements. We believe that a three-step process would reduce the potential for error in identifying students as non-responders and in providing them with instruction ill-matched to their instructional needs. The three steps include use of a theoretical model of beginning reading, assessment referenced to that model, and selection of evidence-based practices that address the students’ identified needs. We describe this process below.
Step 1. Selection of a Theoretical Model
Evidence-based practices answer the question of what to teach, but provide little guidance in the case of struggling readers or students in special education about who, when, or why. RTI provides guidance about who to modify instruction for and when (i.e., whoever fails to thrive at one or more tiers of instruction) but no explanation as to why. Theoretical models answer the question of why. We have described already the Morris, Bloodgood, Lomax, and Perney’s (2003) beginning reading model that helped us understand our students’ failure to respond to evidence-based instruction. Next we describe a diagnostic assessment aligned to this theory.
Step 2. Assessment Referenced to the Theoretical Model
One diagnostic assessment that aligns directly with the model we selected is the Early Reading Screening Instrument (ERSI; Morris, 1998). Originally designed as a tool for screening at-risk beginning readers for additional instructional assistance, the ERSI requires approximately 20 minutes to administer individually. Students’ relative performance on the sub-tests can be compared in order to answer the question, “What is preventing this student from progressing as a beginning reader?” That is, the student’s greatest needs can be identified and subsequently addressed more intensively and specifically than previous instruction or tiered interventions that have not assisted the student in progressing satisfactorily. The ERSI has been demonstrated to be both valid and reliable for the use proposed here (Lombardino, Morris, Mercado, DeFillipo, Sarisky, & Montgomery, 1999; Morris, Bloodgood, Lomax, & Perney, 2003; Perney, Morris, & Carter, 1997).
The ERSI assesses all areas of the beginning reading model except contextual reading, which is easily assessed interactively and observationally or with other instruments such as an informal reading inventory (e.g., Johns, 2012). Letter name knowledge (ABC) is assessed by asking children to point to and write upper and lower case letters. The Concept of Word (CW) task requires children to engage in finger-point reading of two short texts once they have observed the examiner’s modeling of this behavior. Word recognition (WORD) is assessed through reading aloud both high frequency words and decodable CVC words. Beginning consonant (BC), spelling the beginning and ending sounds of words (SPBE), and phoneme segmentation (PS) are assessed through a single developmental spelling task: writing a short list of words with different spelling patterns. Children who spell, for example, back as B are demonstrating beginning consonant knowledge, as BK or BC have knowledge of beginning and ending sound spelling, or as BAC or BAK understand phoneme segmentation.
All ERSI materials, with additional directions and suggestions, as well as interpretation guides for the developmental spelling assessment, can be found in Tyner (2004) or Lombardino, Morris, Mercado, DeFillipo, Sarisky, & Montgomery (1999). Copies of the two short texts used for finger-point reading, that include color photo illustrations instead of the original black-and-white line drawings, can be found and used online, in print, or downloaded as PDF files for use on handheld devices such as iPads: Katie (http://tarheelreader.org/2009/02/11/katie/) and My Home (http://tarheelreader.org/2009/02/11/my-home/).
The Phonological Awareness Screening-Kindergarten (PALS-K; Invernizzi, Meir, Swank, & Juel, 2003), which was developed initially as part of Virginia’s Early Intervention Reading Initiative, is another assessment with demonstrated reliability and validity that might be used to identify a student’s area(s) of greatest need relative to the beginning reading model. Materials, video demonstrations of the assessment procedures, underlying research, support materials, and links to purchase information are found at the PALS-K website (https://pals.virginia.edu/tools-k.html). The ERSI can be obtained and used without materials costs; the PALS-K cannot. A comparison of the tasks employed to assess key beginning reading skills relative to the beginning reading model is found in Table 1 below.
Table 1. Comparison of ERSI and PALS-K Assessment Tasks
Upper and lower case letter name recognition and upper case production
Lower case letter name production
Spelling variety of words
Identify picture with same beginning sound as target word
Spelling variety of words
Spelling CVC words
Concept of word
Finger-point reading two short texts following examiner model, pointing to specific words in text
Finger-point reading aloud familiar nursery rhyme, pointing to specific words in text, reading specific words from text out of context
Spelling beginning and ending sounds
Spelling variety of words
Spelling CVC words
Reading aloud high frequency sight words and decodable CVC words
Optional reading aloud preprimer, primer, or first grade words as individually appropriate
Step 3. Selection of Evidence-Based Instruction
Armed with knowledge of a theory of beginning reading and student performance, teachers and schools are prepared to consider what kind of scientifically-based instruction will be most likely to address their students’ needs at appropriate levels. One source of information about evidence-based practices in beginning reading is the What Works Clearinghouse (http://ies.ed.gov/ncee/wwc/).
Given the assessment results, Mark and his classmates were adolescents who read at beginning levels. They needed a program that would help them acquire knowledge of beginning consonant sounds (BC), letter-sound relationships (SPBE, PS), and concept of word, while continuing to strengthen the understandings they had demonstrated in other aspects of beginning reading.
Ladders to Literacy: A Kindergarten Activity Book (O’Connor, notary-Syverson, & Vadasy, 1998) represents one evidence-based alternative for beginning readers. The What Works Clearinghouse (U.S. Department of Education, 2007) reported that students in experimental studies of these activities made a 25 percentile gain in alphabetics (i.e., the phonological awareness, phonics, spelling, and concept of word knowledge our students needed to acquire).
Examining Ladders to Literacy closely, we find activities to address beginning consonant knowledge (e.g., letter sound of the week, pp. 119-123), concept of word (e.g., morning/afternoon message and news, pp. 43-46), spelling beginnings and endings of words (e.g., integrating spelling and reading, p. 86), phoneme segmentation (e.g., segmenting into three phonemes, pp. 149-152), and word reading (e.g., transition to reading words, pp. 84-85). As the authors note, this is a book of activities, not a curriculum. Consequently, it is particularly useful in the hands of educators armed with an understanding of beginning reading theory and their students’ relative needs and abilities, who can then purposefully select those activities particularly needed by their students. Such educators can thoughtfully observe learning outcomes as activities are implemented and then select additional activities and experiences based on the relative responses of their students.
We are hopeful that others can learn from our mistake and new appreciation of the role of theory in informing instructional decision-making. With the growth of RTI, we are concerned that others may be selecting scientifically-based interventions without considering the match or mismatch to student needs. As we discovered, while many types of phonics instruction are scientifically-based, they are not equally appropriate. We are concerned that students may fail to respond, not because they cannot learn, but rather because the chosen instructional approach is not the best fit to their needs or existing skills and understandings. The resulting, and unnecessary, waste of increasingly scarcer resources in response to a failure attributed to the student would be unfortunate at best.
We are not suggesting that theory solves all instructional difficulties, and we recognize that there are a good many theories to be considered in trying to develop more thoughtful approaches to student learning difficulties. It is important, however, that school administrators, program coordinators, and teachers recognize that instruction cannot be considered high quality unless it address the needs of intended students at appropriate levels of support and challenge. Likewise, an intervention cannot be considered an intervention until and unless it addresses a student’s needs. Assumptions of student literacy learning incapability will remain failures of instruction or program administration until and unless questions of the (mis)match of intervention and student needs are carefully considered.
Adams, M. J. (1990). Beginning to read: Thinking and learning about print. Cambridge, MA: MIT Press.
Browder, D. M. (2001). Curriculum and assessment for students with moderate and severe disabilities. New York: Guilford.
Cunningham, P. M., & Cunningham, J. W. (1992). Making words: Enhancing the spelling-decoding connection. Reading Teacher, 46, 106-115.
Cunningham, P. M., & Cunningham, J. W. (2002). What we know about how to teach phonics. In A. E. Farstrup & S. J. Samuels (Eds.), What researchhas to say about reading instruction(3rd ed., pp. 87–109). Newark, DE: International Reading Association.
Cunningham, P. M., & Hall, D. P. (2007). Making words first grade. Upper Saddle River, NJ: Pearson.
Foulin, J. N. (2005). Why is letter name knowledge such a good predictor of learning to read? Reading and Writing, 18, 129-155.
Henderson, E. H. (1980). Developmental concepts of word. In E. H. Henderson & J. Beers (Eds.), Developmental and cognitive aspects oflearning to spell: A reflection of word knowledge(pp. 1-14). Newark, DE: International Reading Association.
Invernizzi, M., Meier, J. D., Swank, L., & Juel, C. (2003). Phonological awareness literacy screening for kindergarten (PALS-K): Administration and scoring guide. Charlottesville: University of Virginia.
Johns, J. (2012). Basic reading inventory: Pre-primer through grade twelve and early literacy assessments (11th ed.). Dubuque, IA: Kendall Hunt.
Justice, L. M., Pence, K., Bowles, R. B., & Wiggins, A. (2006). An investigation of four hypotheses concerning the order by which 4-year-old children learn the alphabet letters. Early Childhood Research Quarterly, 21, 374-389.
Leslie, L., & Caldwell, J. (2000). Qualitative reading inventory-3 (3rd ed.). Boston: Allyn and Bacon.
Lombardino, L. J., Morris, D., Mercado, L., DeFillipo, F., Sarisky, C., & Montgomery, A. (1999). The early reading screening instrument: A method for identifying kindergarteners at risk for learning to read. International Journal of Language and Communication Disorders, 34, 135-150.
Morris, D. (1992). Concept of word: A pivotal understanding in the learning-to-read process. In S. Templeton & D. Bear (Eds.), Development of orthographic knowledge and the foundations of literacy(pp. 55-77). Hillsdale, NJ: Erlbaum.
Morris, D. (1993). The relationship between beginning readers' concept of word in text and phoneme awareness in learning to read: A longitudinal study. Research in the Teaching of English, 27. 133-154.
Morris, D., (1998). Assessing printed word knowledge in beginning readers: The early reading screening instrument (ERSI). Illinois Reading Council Journal, 26 (2), 30-40.
Morris, D., Bloodgood, J. W., Lomax, R. G., & Perney, J. (2003). Developmental steps in learning to read: A longitudinal study in kindergarten and first grade. Reading Research Quarterly, 38, 302-328.
National Reading Panel (2000). Teaching children to read: An evidence-based assessmentof the scientific research literature on reading and its implications for reading instruction (National Institute of Health Pub. No. 00-4769). Washington, DC: NationalInstitute of Child Health and Human Development.
O’Connor, R. E., Notari-Syverson, A., Vadasy, P. F. (1998). Ladders to literacy: A kindergarten activity book. Baltimore, MD: Paul H. Brookes.
Oelwein, P. L. (1995). Teaching reading to children with Down syndrome: A guide for parents and teachers. Bethesda, MD: Woodbine House.
Perney, J., Morris, D., & Carter, S. (1997). Factorial and predictive validity of first graders’ scores on the early reading screening instrument. Psychological Reports, 81, 207-210.
Stahl, S. A., Duffy-Hester, A. M., Stahl, K. A. D. (1998). Everything you wanted to know about phonics (but were afraid to ask). Reading Research Quarterly, 33, 338-355.
Tyner, B. (2004). Small-group reading instruction: A differentiated teaching model for beginning and struggling readers. Newark, DE: International Reading Association.
U.S. Department of Education, Ladders to literacy for kindergarten students. Institute of Education Sciences, National Center for Education Evaluation and Regional Assistance, What Works Clearinghouse. Retrieved from http://ies.ed.gov/ncee/wwc/interventionreport.aspx?sid=259#