The Science of Reading: Seeking Research-Based Answers to Critical Questions

    by Elfrieda (Freddy) H. Hiebert, TextProject | November 30, 2022

    wooden building blocks featuring a green block with a question mark

    I wrote a series of five blogs in late 2020 on the relationship of research to reading acquisition and instruction:

    1. An overview of the contributions of research to understandings of reading acquisition and instruction
    2. Why orthography is so critical in learning to read English
    3. What we know from science that supports effective reading curriculum
    4. What we know from science that supports effective reading instruction, and
    5. Instruction that does NOT support independent word recognition.

    In the first blog, I promised a final blog that would raise questions that require the attention of researchers. I am finally following through on that promise. This blog presents questions about the curriculum, instruction, and texts of reading instruction (especially in the early stages) for which I have been looking for answers in the ensuing two-year gap.

    I offer these questions in the spirit of inquiry. I am honestly looking for answers to questions that I have as a teacher of reading, teacher educator, and researcher. I have spent many hours reviewing the literature and talking to colleagues, looking for answers to these questions. If you have suggestions for research-based answers, I request that you contact me with citations and the names of researchers who are pursuing these questions. And I promise that, when you share research-based answers with me, there won’t be a two-year lag in my sharing the responses with you.

    Before I list the questions (which I will be presenting as part of a session at the annual meeting of the Literacy Research Association on December 1, 2022), I want to underscore that there are clear-cut answers from research about the experiences and instruction that children need to become independent readers of written English. One of these is that, in an alphabetic language like English, even when all the relationships between letters and sounds are not consistent, guidance about the code is essential. At the same time, a fundamental premise of science is that the knowledge base is not static. Here are some of the questions for which I am hopeful we have answers….and if we don’t, I am hopeful that the knowledge of scientists can continue to grow.

    Looking for Answers

    Questions Related to Curriculum & Instruction

    1. Lessons for vowel grapheme-phoneme correspondences typically begin in kindergarten-—first short vowels, followed by long vowels. What evidence validates this pace of introduction for the approximately 75% of a cohort who don’t recognize any words on kindergarten entry?
    2. Typical phonics curricula concentrate on short and long vowels as well as consonants in kindergarten and first grade. Lessons on these same letter-sound correspondences are repeated in subsequent grades: 40% of second grade, 25% of third grade, and 10% of fourth and fifth grades. For example, words in a Grade 3 text of a best-selling core reading program (Ada et al., 2020) linked to a lesson on short vowels include: unique, individuality, feature, personality, counselor, and competitive. On what evidence is the typical phonics curricula of reading programs based?
    3. How does the degree of variability of a letter-sound correspondence influence the amount of exposure required to become automatic with a grapheme-phoneme unit? Example: Among the 1,000 most-frequent words, readers are as likely to encounter words with variant phonemes associated with “ea” (e.g., bread, great) as they are words with the “consistent” phoneme (e.g., team, eat).
    4. Multisyllabic words are frequent in texts for young children, including decodable ones (Kearns & Hiebert, 2022). Are beginning readers able to identify multisyllabic words when the first syllable consists of letter-sound correspondences they have been taught? For example, once correspondences for short vowels and consonants have been taught, can they decode words such as puppet, hotdog, and suntan?
    5. Within the Lesson-to-text-match model, words can be counted toward the decodability of a text if they are taught as sight words. Example: If letter-sound correspondences for m, t, a, n, s, and p have been taught, the words and and can are sight words (since d and c have not been taught). Which words do children learn—sight words or words they are expected to decode (e.g., man, pan, ant)?
    6. At what point in reading acquisition should the variability in written English, especially vowels, be explicitly discussed and demonstrated in lessons? That is, how early can children be supported in development a set for variability?

    Questions Related to Texts

    1. What ratio of instructional/learning time should be spent on reading individual words (words on cards or lists) and text reading? How does/should this ratio vary at different stages of reading proficiency?
    2. In the lesson-to-text match (LTTM), a word is counted as decodable if a lesson has been provided for each letter-sound correspondence in it. Example: If lessons appear in the teacher’s guide for short vowel sounds associated with a, t, g, h, m, p, s and the has been taught as a sight word, the following text is evaluated as decodable: “Pam has 3 hams. Pam has 3 tags. Pam tags the hams.” What evidence supports the assumption that “once taught, then learned”—especially for students who enter kindergarten not reading?
    3. The current model of decodability (in evaluating texts) emphasizes the letter-sound correspondence. At what point should the unit of instruction/learning move to a larger “grain size,” such as rimes or syllables?
    4. Which type of decodable text better supports young readers’ learning of a
      letter-sound pattern: (a) multiple words with the pattern, most of which appear a single time in target texts (e.g., tee, glee, steed, sheet, cheek, seek, eel, steel, sweep, teen, fleet) or (b) a handful of words with the pattern, one or two of which appear numerous times in texts (e.g., see, feel, tree, street, sleep)?
    5. In Language at the Speed of Sight, Seidenberg (2017) does not mention decodable text but does describe the need for large amounts of data to learn to read: “Readers become orthographic experts by absorbing a lot of data, which is one reason why the sheer amount and variety of texts that children read is important. (p. 92). How much “data” do students get in typical reading programs? Is this amount sufficient to support automatic word recognition for the students whose reading experiences occur primarily in school?
    6. Do decodable texts foster knowledge acquisition? For example: A decodable text that describes a treacherous storm includes the words mates, skipper, fleet, ketch, raft, hull, and swells. Do young readers retain knowledge about sailing and storms at sea from this exposure?