Shellfie
Philosophy and Approach
Why we built a read-aloud buddy, the research behind it, how it helps without taking over, how it keeps parents informed about hard and interesting words, success, reading volume and interests, and how we choose, check and revise every book on the shelf.
1Why Shellfie exists
Children learn to read by reading: out loud, often, to someone who is listening. Most children who are learning to read spend far more time with a book when an adult sits beside them, follows each word, waits while they work out a hard one, and steps in at the right moment. Families don't always have that time every night, and a second language at home can make it harder still.
Shellfie is a turtle who listens. A child reads a story out loud; Shellfie follows along word by word, notices when the reader is stuck, and offers the smallest amount of help that will get them moving again. It is not a game with reading inside it. It is the reading, with a patient friend sitting next to it.
Our goals, in order:
- More minutes of real reading aloud: the practice that builds fluency.
- The child does the work. Help is brief and fades as soon as the reader is moving again.
- Parents know how their child is doing. Each reading day leaves a short, readable record: hard words, words the child fixed alone, books finished and quiz scores, minutes and pages read, and the books the child chose.
- Good books for every reader, in English and Spanish, at every grade from kindergarten to fifth.
- Free, private, and calm. No ads, accounts, tracking or stored recordings.
2What the research says
Shellfie's design follows the research on how children learn to read, what computer and AI reading tutors have and haven't shown, and why informed, involved parents and teachers matter so much. Numbers in brackets refer to the references at the end. We try to state the strength of each finding plainly: some are backed by dozens of studies, others by a few small or vendor-linked ones.
Reading aloud with feedback builds readers
The U.S. National Reading Panel reviewed 77 studies of guided repeated oral reading and found a moderate overall effect on reading achievement (weighted effect size 0.41), with gains in word recognition, fluency and comprehension when a teacher, peer or parent listened and gave feedback [1]. Kuhn and Stahl's review of fluency research reached a similar conclusion: assisted reading generally beats reading alone [2]. The What Works Clearinghouse guide for K–3 recommends that every child read connected text every day, with adults modelling, scaffolding and giving feedback during oral reading [3].
For Shellfie: the core activity is reading whole stories aloud with immediate, word-level feedback, every day.
Volume matters: minutes add up
In a classic study of fifth graders, minutes spent reading books outside school was among the best predictors of reading growth [4]. A meta-analysis of 99 studies (7,669 participants) found moderate to strong links between how much children read and their comprehension, word reading and spelling, links that grow stronger with age [5].
For Shellfie: we count honest minutes of real reading and show them to parents, because time on task is the lever families control most directly.
Scaffold, then fade
“Scaffolding” describes an expert giving just enough support for a learner to finish a task they couldn't yet do alone, then withdrawing it [6]. Feedback helps most when it is specific, timely and aimed at the task and the strategy, not at the child as a person [7]. Children's self-corrections during oral reading are a sign that they are monitoring their own reading, a skill worth protecting [8].
For Shellfie: the three-step help ladder (prompt, syllables, model) starts with the least help, waits for self-correction, and names strategies rather than judging the reader.
Computer and AI reading tutors: promising, modest, and best with adults
- Listening tutors can work. Carnegie Mellon's Project LISTEN Reading Tutor, the ancestor of today's speech-recognition tutors, outperformed sustained silent reading in a randomized trial of 178 struggling readers in grades 1–4 (effect size +0.64 on a composite reading test) [9]. The What Works Clearinghouse found gains in word reading but judged effects on fluency and comprehension uncertain [10]. The tutor's own speech data predicted children's fluency test scores well (correlations of about 0.78–0.83) [11].
- Average effects of reading technology are small. Large syntheses by Cheung and Slavin found small average effects of reading software (about +0.14 to +0.16), with the strongest results when technology was integrated with teachers' instruction and supported by professional development, and near-zero effects for some large stand-alone programs [12, 13]. Evidence for ESSA rates the AI tutor Amira as having “moderate” evidence, averaging +0.15 across two large studies [14].
- Newer AI apps are under-studied. Evaluations of Google's Read Along are mostly company-linked pilots or small studies without control groups [15]. Shellfie itself has not been tested in a controlled study; our tests are simulations and real-device checks, which we document openly.
- Speech recognition is imperfect, and not equally so. Commercial recognizers made almost twice as many errors for Black speakers as for white speakers in a 2020 study [16], and children's speech is harder to recognize than adults'. A low word-error rate also doesn't guarantee that real reading mistakes are caught [17].
For Shellfie: treat Shellfie as practice and as a window for parents, not as a test score. Because recognition is imperfect, Shellfie leans toward accepting a child's reading, accepts accent and dialect patterns, and gives grown-ups control over those settings.
Parents and teachers who know what's happening make the difference
- Parent involvement in reading works. A meta-analysis of 16 family-literacy studies (1,340 children, K–3) found a moderately large average effect (about 0.65); parents listening to their children read and practicing specific skills with them produced bigger gains than reading to children alone [18]. A broad research synthesis found that family involvement is linked to higher achievement across backgrounds, especially when it is tied to learning [19].
- Small, specific information changes what families do. Three short texts a week with literacy tips increased parents' home literacy activities (0.22–0.34 SD) and children's early literacy scores [20]. Weekly one-sentence teacher-to-parent messages cut the share of students failing to earn credit from 15.8% to 9.3%, and messages about what to improve worked best [21]. Weekly automated alerts to parents reduced course failures by 27% [22].
- Data helps only when adults act on it. Progress monitoring raises achievement when teachers use the data to make decisions; collecting it isn't enough [23].
For Shellfie: the Parents area turns each reading day into a short, specific summary: which words were hard, which the child fixed on their own, what they read and how much. It's built to be acted on in two minutes and to share with a teacher.
Interest and choice keep children reading
Letting children choose what they read and connecting reading to their interests is associated with more motivation and achievement. Guthrie and Humenick's synthesis of 22 studies reported large effects for choice [24], though a later meta-analysis of reading-motivation interventions found smaller average effects (g = 0.30 on motivation, 0.20 on achievement) [25].
For Shellfie: children browse by grade or by genre and pick their own books. The books they choose, shown day by day in the Parents area, give parents a picture of their child's interests.
Screens, features and comfort
A meta-analysis of 43 studies of digital storybooks found small benefits from multimedia that matches the story (such as animated pictures), while interactive extras like hotspots, games and pop-up dictionaries tended to distract, especially for children from less advantaged homes [28]. Comprehension tends to be a little better on paper than on screens, mostly for informational texts read under time pressure [29]. Colored overlays and tinted lenses have not been shown to help reading difficulties [27].
For Shellfie: no games or hotspots inside stories, a calm mode, no timers, and honest wording about reading colors. Shellfie is one part of a reading life that should also include paper books, especially nonfiction.
A word about fluency numbers
National oral reading fluency norms (for example, about 100 words correct per minute for a typical second grader by spring) are measured on unpracticed grade-level passages under test conditions [26]. Shellfie's words-correct-per-minute comes from stories a child chose, sometimes reread, with help available, so it is best used to watch your own child's trend over months, not to compare with a norm chart.
From research to the Parents view
| Parents want to know | What Shellfie shows | Why it matters |
|---|---|---|
| Hard words | “Challenging” words (helped twice or more, or Shellfie had to say them), the word quiz, and the long-term Words to practice list | Targeted practice of specific words with a parent is one of the strongest family-literacy effects [18] |
| Interesting words | Tricky words the child fixed on their own | Self-correction shows the child is monitoring their reading [8]; praise effort and strategy [7] |
| Success | Books finished, best and last accuracy, comprehension quiz scores, word-quiz results | Understanding is the goal of reading; accuracy and comprehension together show real progress [1, 3] |
| Throughput | Minutes of real reading per day and per range, pages, words correct per minute, streak, reading days | Reading volume predicts growth [4, 5] |
| Reading interests | The books the child picks each day, and “Kinds of books” ranked by minutes of reading | Choice and interest sustain motivation [24, 25] |
3Our educational practice
Shellfie is built on a handful of well-established ideas about how early reading develops. We use them as design rules.
Reading aloud with feedback
Repeated, supported reading out loud is one of the most reliable ways to build fluency. Shellfie gives feedback while the child is reading, not afterward.
Wait before helping
A pause is often a reader thinking. Shellfie waits, gives the reader a chance to fix their own mistake, and celebrates when they do (“Nice fix!”).
Least help first
Help moves from a hint, to a strategy (syllables or sounding out), to modelling the word. The child always gets the last turn to say it.
Reading is understanding
Each chapter ends with three short questions about who, what and why, so decoding and meaning grow together.
Grade-tuned pages
Words per page, type size and the definition of a “hard” word change by grade, following the California Common Core reading standards for text complexity.
Accent is not an error
Dialect, accent and speech sounds a child is still developing are accepted as correct reading, so children are never corrected for how they talk.
Motivation is kept simple and tied to reading itself: a shell for each page read, a special shell for finishing a whole book, and a gentle daily streak. There are no timers that pressure a child, no leaderboards, and no rewards for skipping ahead.
4Scaffolding: how Shellfie helps
When a reader stops on a word for longer than is usual for their grade, Shellfie starts a three-step help ladder. The whole ladder usually takes under ten seconds, and it ends the moment the child reads the word.
Prompt
The word is underlined and blinks. Shellfie waits two to three seconds so the reader can solve it alone. A self-correction earns a “Nice fix!”.
Scaffold
The word is enlarged and broken into syllables, read slowly one piece at a time. Spanish uses its regular syllable rules; irregular English words are said whole; older readers who miss a long word again hear it spelled.
Model + “Your turn!”
Shellfie says the word, then gives the child about eight seconds (more for careful readers, extended each time they sound out a piece) to say it themselves.
Smaller supports around the ladder
- Waiting is grade-tuned. Shellfie waits about 4 seconds before helping a kindergartner and about 2.5 seconds for a fifth grader, and longer for a reader whose profile says they read carefully.
- Reading on beats stopping. If the child keeps reading past a word, Shellfie treats that as a decision, not a stall, and steps aside.
- Names and repeat helps are quick. Character names, and any word after two full ladders on the same page, are simply said, so a page never becomes a drill.
- Hard words are marked but capped. In K–2 a dotted underline previews the few hardest words on a page; the number counted as “hard” is limited so pages stay readable.
- Tap for help. Any word can be tapped at any time to jump straight to the scaffold, and Shellfie works fully in tap mode when there's no microphone.
- Lost your place? If a reader skips a paragraph, Shellfie points to where to pick back up (“Oops! We skipped a part. Start here!”).
- Reading comfort. A child-friendly reading typeface (Andika), soft page and text colors for children who find them more comfortable, and a calm mode with fewer animations. (Research has not shown that colored backgrounds treat dyslexia [27], so we offer them as a comfort choice, not a treatment.)
5Time on task
The most important number in early reading is how many minutes a child actually spends reading aloud. Shellfie is designed to protect that time and to measure it honestly.
Only real reading counts
“Minutes read” grows only while the child is moving through the words. Gaps of 15 seconds or more without progress are not counted.
Pause, not penalty
After about 15 seconds of silence the book pauses itself (“Ready when you are”), so no one is scolded for stepping away.
Keep the flow
Help is short, Shellfie turns the page for the reader, and the only sound while reading is Shellfie saying a word when helping.
A reading streak counts days with any reading, and the 365 Bedtime Stories book opens on tonight's story, which makes a short nightly habit easy to keep. We suggest 10–20 minutes a day for most K–5 readers; a little every day matters more than a long session once a week.
6Reader profiles
No two children read the same way. Shellfie keeps a small reader profile so its patience and help fit the child in front of it rather than a grade-level average.
How a profile is learned
During the first stretch of reading aloud (about 50 words in kindergarten, rising to about 120 words from third grade), Shellfie listens with extra-gentle timing and measures pace, typical pause between words, and slips per 100 words. It repeats this on the first reading of each day and blends the new reading into the old one (70% history, 30% today), so one tired evening doesn't redefine a reader.
Reads quickly and smoothly. Shellfie waits a little less before helping.
Reads at a steady pace. Shellfie uses its normal timing.
Takes time with each word. Shellfie waits much longer, so sounding out is never interrupted.
Speech patterns
The profile also records the ways of speaking Shellfie should accept as correct: th sounds (the → de), r or l as w (rabbit → wabbit), dropped endings (walked → walk), dropped s in Caribbean and coastal Spanish, and others. They switch on when Shellfie hears them often. On everyday words every pattern is accepted; on the hard words a page is teaching, only the patterns in the child's profile are, so a real decoding slip is still noticed.
Grown-ups stay in charge
- Choose a reading style yourself instead of “Automatic”.
- Switch any speech pattern on or off.
- Lock the profile so it doesn't update each day, or ask Shellfie to re-check during the next book.
The profile lives only in this browser on this device. Shellfie keeps one reader per browser, so siblings who share a device can each use their own browser profile or device. Progress can be exported and imported from the grown-ups panel.
7Grown-ups in the loop: word review
Shellfie is a practice partner, not a replacement for a caring adult. The most powerful thing a parent can do is look at the words a child needed help with and practice a few of them together. Shellfie records exactly that, behind a press-and-hold Parents button so children don't wander in.
Reading days & word review
The grown-ups panel lists every reading day with the minutes read, the books read, the number of helps and the quizzes taken. Days when your child read fluently and needed no help are listed too. It opens on the latest day; choose another day, a custom date range, or 7, 14 or 30 days. The words from that range are sorted into three groups:
Needed help two or more times, or Shellfie had to say or spell the word.
Shellfie helped sound it out once.
Tricky words the reader fixed on their own. Worth celebrating.
Press Start word quiz and the selected words become a short spoken review. The child reads each word aloud. A word read correctly gets praise and the quiz moves on. If the child is silent, the word is split into syllables, Shellfie says them, and then it's “Your turn!”. Help me and Skip are always there, and without a microphone the child taps “I said it”. At the end, the review shows which words were read on their own, which with help, and which were skipped. The list can also be copied, printed or emailed, for the fridge, a tutor, or a teacher.
The longer-term Words to practice table shows the 25 words that have needed the most help overall and which support worked best (syllables, sounding out, spelling, hearing the word, or a self-fix). Here is the routine we recommend:
Latest day review
- After reading, open Parents; the latest day is already selected.
- Run the word quiz on the Challenging and Helped words.
- Praise effort and self-fixes (the Interesting words), not speed.
7-day patterns
- Choose 7 days and scan the groups for word families: -tion, silent e, ll/y.
- Quiz the words that keep coming back.
- Check comprehension scores and reread a favorite.
30-day check-in
- Choose 30 days; print or email the list to share with a teacher.
- Note minutes, books finished and words correct per minute. Export progress to keep a dated snapshot.
- Move up a grade shelf, or stay put, based on comfort, not age.
For any day or range, the panel also shows minutes of reading and kinds of books (genres ranked by reading time), a simple picture of your child's interests. Under the Books table, a short guide explains accuracy: books read at about 95% or higher usually feel comfortable, and a book that stays well below 90% is a good one to reread together or swap for an easier shelf. Because accuracy comes from speech recognition, treat it as a rough guide.
Pause, prompt, praise
When your child reads to you, wait about five seconds before helping, since many children fix the word themselves. If they're still stuck, give a prompt (“What's the first sound?”, “Does that make sense?”) before saying the word. Then praise the effort or the fix, not just getting it right. Parent tutoring like this has a solid research record [30].
A word that moves from Challenging to Interesting, or drops off the list, is the clearest sign of progress. Words that stay Challenging for weeks are the ones to share with a teacher.
8How we pick books
The library holds about 355 readable books (roughly 217 in English and 138 in Spanish) across kindergarten to fifth grade, plus titles marked “Soon” that are still being prepared. Books come from three kinds of sources:
| Source | What it is |
|---|---|
| Public-domain classics | Fairy tales, fables, poetry, Beatrix Potter, Mother Goose, Aesop, Quiroga, Rafael Pombo and more, from Project Gutenberg, Wikisource and national library collections. We check each edition's rights; an old story does not make a modern translation free to use. |
| Retellings & adaptations | Short, faithful-in-spirit retellings for younger readers when the original is too long, too old-fashioned or not suitable as written. Every adaptation is labelled as one in its credit. |
| Original Shellfie books | “Meet the Animals” nonfiction, “When I Grow Up” jobs books, seasonal and holiday stories, and our school's own McKinley Bulldogs shelf, written in both languages. |
What we look for
- Grade fit by three measures, following the California Common Core standard for text complexity (RL.10): qualitative features (plot, themes, vocabulary, background knowledge), quantitative measures, and the reader and the task, including the quiz. No single readability score decides a grade.
- A grade is a shelf suggestion, not a limit. Children can read any shelf; reading ability is never inferred from age.
- Variety. Fairy and folk tales, fables, animal stories, everyday life, adventure, poetry, nonfiction and seasonal books, browsable by grade or by genre.
- Real Spanish. Spanish books are chosen and written as Spanish, not machine-translated, with modernized spelling and Mexican Spanish voices for help. Spanish texts are not placed by English readability formulas.
- Many cultures, told respectfully. Legends from Mexican, Nahua, Mazatec, Colombian and other traditions name their sources and note that versions vary.
9Monitoring content
Every book and every quiz goes through the same layered review. Automated tools help us look; people decide.
Automated flags
A bilingual scan flags passages for review: graphic violence, weapons, slurs and dehumanizing language, self-harm, substances, abuse and more. A flag prompts a reading in context; it is never treated as a verdict or a safety certificate.
Whole-text review
A book's row in our grade review is marked reviewed only after the entire text and every quiz question has been read. A keyword scan does not count as a review.
Questions and distractors
The question, the evidence, the right answer and every wrong option are checked. A wrong answer may not introduce a scary scene that isn't in the story.
Automated checks
Tests confirm that approved edits are applied, quiz answers appear in the revised text, sources and credits are present, and pages flow in order.
Our content guidance by grade
| Grades | Guidance |
|---|---|
| All grades | No graphic injury, torture or killing; no sexual content; no instructions for dangerous acts; no slurs or stereotypes. Including people of different races, religions, abilities, families and traditions is not a concern; respectful representation is welcome. |
| K–1 | Reassuring endings and mild, brief conflict. No realistic threats or frightening detail. Loss, if present, is gentle and non-graphic. |
| 2–3 | Mild peril with context and resolution. Bullying, fear and historical injustice reviewed carefully; cruelty never normalized. |
| 4–5 | More complex history and social conflict when it has a clear educational purpose, handled without sensational or graphic detail. |
This is Shellfie's editorial policy, informed by the California standards. It is not an official rating, and it does not claim every book is approved by every school or district.
10The revisions we make
When a story doesn't meet the guidance, we change it openly. The original stays in version history, the book's credit says what changed, and every change is written up in a public revision log with the source, location, reason, guidance used, and tests.
| Type | What we do | Example |
|---|---|---|
| Removal | Remove a sentence, paragraph or whole chapter with dated or hurtful language, and note it in the credit. | Old fairy-tale collections with offensive period stereotypes. |
| Adaptation | Replace violence or cruelty with a safe, kind resolution while keeping the heart of the story. Labelled as an adaptation. | Quiroga's jungle tales retold without hunting or revenge, with flamingos' stockings made of fabric. |
| Selection | Use part of a long work, clearly marked as a selection. | The first four stanzas of Pombo's Mirringa Mirronga. |
| Modernizing | Update old Spanish spelling, and remove front matter so the story starts on page one. | Source notes removed from the start of Jemima and El ratón Pérez. |
| Context | Add a short, separate note with facts or safety advice, kept distinct from the story. | Why flamingos are pink; ask an expert before helping an injured wild animal; fire safety. |
| Quiz fixes | Rewrite questions that are unclear, unsupported by the revised text, or that introduce upsetting options. | A replacement Mother Goose quiz. |
| Re-shelving | Move a book to a better-fitting grade, recording the old grade and the evidence. | Tracked in the grade review worklist. |
We are honest about what is still pending. Where an educator, a native Spanish speaker or a member of a cultural community has not yet reviewed a book, the log says so, and we don't describe a story as community-approved when it is not.
11Privacy and honesty
- No ads, no tracking, no accounts. Progress is saved only in this browser.
- Recordings are never stored. Some browsers send audio to their maker's speech service to turn it into text; where on-device speech recognition is available, audio stays on the device. The grown-ups panel tells you which is in use.
- Logs stay on the device until a grown-up chooses to copy or send them for troubleshooting.
- We report what we know. Speech recognition on some phones and tablets is still being verified, and our test notes say so. If something seems off, the grown-ups panel can copy the last 30 minutes of notes to share.
Shellfie is a free bilingual read-aloud buddy at sawyertobin.com/shellfie. Questions or a book you'd like to see? Open the Parents panel and use Troubleshooting, or talk to your child's teacher about how Shellfie fits their classroom reading.
12References
- National Reading Panel (2000). Teaching Children to Read, Ch. 3: Fluency. NICHD. www.nichd.nih.gov
- Kuhn, M. R., & Stahl, S. A. (2003). Fluency: A review of developmental and remedial practices. Journal of Educational Psychology, 95(1), 3–21.
- Foorman, B., et al. (2016). Foundational Skills to Support Reading for Understanding in Kindergarten Through 3rd Grade (NCEE 2016-4008). What Works Clearinghouse, IES. ies.ed.gov
- Anderson, R. C., Wilson, P. T., & Fielding, L. G. (1988). Growth in reading and how children spend their time outside of school. Reading Research Quarterly, 23(3), 285–303.
- Mol, S. E., & Bus, A. G. (2011). To read or not to read: A meta-analysis of print exposure from infancy to early adulthood. Psychological Bulletin, 137(2), 267–296. doi.org
- Wood, D., Bruner, J. S., & Ross, G. (1976). The role of tutoring in problem solving. Journal of Child Psychology and Psychiatry, 17(2), 89–100.
- Hattie, J., & Timperley, H. (2007). The power of feedback. Review of Educational Research, 77(1), 81–112.
- Clay, M. M. (1969). Reading errors and self-correction behaviour. British Journal of Educational Psychology, 39(1), 47–56.
- Mostow, J., Nelson-Taylor, J., & Beck, J. E. (2013). Computer-guided oral reading versus independent practice: Comparison of sustained silent reading to an automated reading tutor that listens. Journal of Educational Computing Research, 49(2), 249–276. www.istation.com
- What Works Clearinghouse. Intervention report: Project LISTEN Reading Tutor. IES. Summarized by Evidence for ESSA (see 14).
- Beck, J. E., Jia, P., & Mostow, J. (2004). Automatically assessing oral reading fluency in a computer tutor that listens. Technology, Instruction, Cognition and Learning, 2, 61–81. www.ri.cmu.edu
- Cheung, A. C. K., & Slavin, R. E. (2013). Effects of educational technology applications on reading outcomes for struggling readers: A best-evidence synthesis. Reading Research Quarterly, 48(3), 277–299. doi.org
- Cheung, A. C. K., & Slavin, R. E. (2012). How features of educational technology applications affect student reading outcomes: A meta-analysis. Educational Research Review, 7(3), 198–215.
- Evidence for ESSA. Amira (program rating). Johns Hopkins University Center for Research and Reform in Education. www.evidenceforessa.org
- Google (2019). Read Along (Bolo) pilot findings, Unnao, India; with later independent and academic evaluations of mixed rigor. services.google.com
- Koenecke, A., et al. (2020). Racial disparities in automated speech recognition. PNAS, 117(14), 7684–7689. doi.org
- Molenaar, B., Tejedor-García, C., Cucchiarini, C., & Strik, H. (2023). Automatic assessment of oral reading accuracy for reading diagnostics. Interspeech 2023. arxiv.org
- Sénéchal, M., & Young, L. (2008). The effect of family literacy interventions on children's acquisition of reading from kindergarten to grade 3: A meta-analytic review. Review of Educational Research, 78(4), 880–907.
- Henderson, A. T., & Mapp, K. L. (2002). A New Wave of Evidence: The Impact of School, Family, and Community Connections on Student Achievement. SEDL.
- York, B. N., Loeb, S., & Doss, C. (2019). One step at a time: The effects of an early literacy text-messaging program for parents of preschoolers. Journal of Human Resources, 54(3), 537–566. www.nber.org
- Kraft, M. A., & Rogers, T. (2015). The underutilized potential of teacher-to-parent communication: Evidence from a field experiment. Economics of Education Review, 47, 49–63. annenberg.brown.edu
- Bergman, P., & Chan, E. W. (2021). Leveraging parents through low-cost technology: The impact of high-frequency information on student achievement. Journal of Human Resources, 56(1), 125–158. www.povertyactionlab.org
- Stecker, P. M., Fuchs, L. S., & Fuchs, D. (2005). Using curriculum-based measurement to improve student achievement: Review of research. Psychology in the Schools, 42(8), 795–819. doi.org
- Guthrie, J. T., & Humenick, N. M. (2004). Motivating students to read: Evidence for classroom practices that increase reading motivation and achievement. In P. McCardle & V. Chhabra (Eds.), The Voice of Evidence in Reading Research (pp. 329–354). Brookes.
- McBreen, M., & Savage, R. (2021). The impact of motivational reading instruction on the reading achievement and motivation of students: A systematic review and meta-analysis. Educational Psychology Review, 33, 1125–1163.
- Hasbrouck, J., & Tindal, G. (2017). An Update to Compiled ORF Norms (Technical Report No. 1702). Behavioral Research and Teaching, University of Oregon. www.readingrockets.org
- Griffiths, P. G., Taylor, R. H., Henderson, L. M., & Barrett, B. T. (2016). The effect of coloured overlays and lenses on reading: A systematic review of the literature. Ophthalmic and Physiological Optics, 36(5), 519–544. doi.org
- Takacs, Z. K., Swart, E. K., & Bus, A. G. (2015). Benefits and pitfalls of multimedia and interactive features in technology-enhanced storybooks: A meta-analysis. Review of Educational Research, 85(4), 698–739. pmc.ncbi.nlm.nih.gov
- Delgado, P., Vargas, C., Ackerman, R., & Salmerón, L. (2018). Don't throw away your printed books: A meta-analysis on the effects of reading media on reading comprehension. Educational Research Review, 25, 23–38.
- Erion, J. (2006). Parent tutoring: A meta-analysis. Education and Treatment of Children, 29(1), 79–106.
Effect sizes are reported as in the sources. Several figures come from summaries of the original papers; see the originals before citing a specific number.