What Teaches a Child to Read? What the Research Supports, and What to Skip

Read the shelves, or show one kind of thing at a time.

18 things on the shelves and a free shelf · as of November 2026

How much research stands behind each thing you could buy or do?

Each shelf holds one kind of research. Open a card to see what was tested and by whom.

2of 18

2 of the 18 things on these shelves have been tested themselves in controlled studies that a group other than the maker rated or pooled. One is Lexia Core5 Reading, software that schools buy. Another is tutoring by a person, from any provider.

Show one kind of thingEverythingAllBooks and kitsBooksApps and websitesAppsTutoringTutors

Showing all 18 things on the shelves. The free shelf always stays.

A shelf is a kind of research, never a score. Things on the same shelf are in alphabetical order and are not ranked against each other.

Shelf 1 · 2 things

The thing itself was tested

Controlled studies tested the thing itself. A group other than its maker rated or pooled them.

• Lexia Core5 ReadingReading software that schools buyWe found no price for families.

  Evidence for ESSA, a rating group at Johns Hopkins, rates it Moderate. Its 4 studies covered 17,281 students and found an average effect of +0.16, compared with students who did not use it. The largest compared more than 6,000 students in grades 3 to 5 with matched students in similar schools, on a reading test. Lexia published one of the 4. These were studies of use in school.

  Source: Evidence for ESSA

• Tutoring by a personOne-on-one or small-group teaching, from any providerPrices vary. We did not collect them.

  A 2024 review pooled randomized trials of tutoring for children from pre-kindergarten to grade 12, in reading and math. Tutoring raised learning by about 0.29 of a standard deviation on average, compared with students who were not tutored. Effects tended to be larger with teachers or school aides as tutors, in earlier grades, and at 3 or more days a week. Those are patterns across studies that differed in many ways.

  Sources: Nickow, Oreopoulos and Quan, 2024, Stanford summary

Shelf 2 · 4 things

The teaching method is tested. The product is not.

The maker’s page says it teaches phonics, the method with strong research. We found no study of the product itself.

• All About Reading, Level 1A phonics kit with lightly scripted lessons$159.95, plus $25.95 for letter tiles and $17.95 for a review box

  We found no study of this program. The maker’s main page says it covers phonics, vocabulary and comprehension. A customer quote there calls it Orton-Gillingham. A 2021 review of Orton-Gillingham teaching for children who struggle to read found a gain of 0.22 on basic reading skills over the comparison groups. The review calls that not statistically significant (p = .40).

  Sources: Maker’s page for Level 1, Maker’s main page, Stevens and others, 2021

• Bob Books, Set 112 short books a beginner can sound out$17.99 for the set

  We found no study of Bob Books. The federal practice guide recommends books a beginner can sound out for practice. A 2018 research review supports books like these “in the very early stages” of reading. It says their limits are likely to outweigh their benefits after that.

  Sources: Maker’s page, Federal practice guide, Castles, Rastle and Nation, 2018

• Duolingo ABCA learn-to-read appFree

  The maker’s page calls it a comprehensive phonics curriculum and says it is based on recommendations by the National Reading Panel. We could not find a study of the app that we could open.

  Source: Maker’s page

• Logic of English, Foundations AA program with its own readers$97.96 regular price for the Foundations A bundle. The maker says Core Materials are also required, and the Foundations A and Core Materials bundle lists at $251.86.

  We found no study of this program. The maker’s page says Foundations “uses a structured approach to help children understand the relationship between letters and sounds”.

  Source: Maker’s page

Shelf 3 · 9 things

Small or maker-linked studies, or none we found

The studies are small, short, not randomized, paid for or done with the maker, or about skills that come before reading. Or we found no study, and the product is not sold mainly as phonics.

• ABCmouseA learning app for young children$14.99 a month after 30 free days, until canceled

  The research its page points to is the maker’s own. We found no What Works Clearinghouse report on it. In 2020 the maker agreed to pay $10 million to settle federal charges about renewals and cancellations. That case was about billing and made no finding about learning.

  Sources: Maker’s page, Federal Trade Commission, 2020

• HomerAn early reading appThe maker’s page shows a 30-day free trial and no price.

  One randomized study covered 82 children in 7 Head Start classrooms. They used the app 15 minutes a day for 6 weeks, and a comparison group used a math app. It measured skills that come before reading, such as hearing the sounds in words. A New York University professor wrote the summary, and the company hosts it. The summary does not say who paid. We found no journal article.

  Sources: Maker’s page, Study summary, hosted by the company

• Khan Academy KidsA learning app for young childrenFree

  A University of Massachusetts team ran one randomized trial at the maker’s request. The Overdeck Family Foundation helped pay for it. It covered 49 children aged 4 and 5, mostly from low-income families, for 10 weeks. The children using the app went from the 34th percentile to about the 47th on pre-reading skills. They gained more than a comparison group that used other apps. It did not measure reading. The 2020 news release called the results preliminary.

  Sources: Maker’s page, University of Massachusetts Amherst

• Kumon readingCenter visits twice a week, with worksheets at home on the other five daysKumon’s page shows no price. Ask the center.

  We found no independent study of its reading program, and its page cites none. Its preschool program starts with “look, listen and repeat” word work before phonics.

  Source: Kumon’s page

• Reading EggsA learn-to-read app$13.99 a month for the reading and math plan for up to 4 children, free for the first 30 days

  The maker says it worked with a partner firm on a study of “the relationship between program usage and students’ achievement”. That kind of study cannot show the app caused the scores. We found no randomized trial.

  Sources: Maker’s page, Maker’s account of its study

• StarfallA learn-to-read websiteThe learn-to-read lessons are free. A home membership is $35.00 a year.

  The maker hired Metis Associates, an evaluation firm, to study its kindergarten classroom program in one Colorado school district. In spring, 58.8% of the 267 Title 1 kindergartners who had the program scored proficient or higher on a reading assessment, against 31.1% of the 164 who did not. It was not a randomized trial, and the evaluators asked for a larger randomized study. We found no study of home use.

  Sources: Maker’s page, Metis Associates report, Price

• SylvanTutoring at a center$50 or more a session. Prices vary by location, and rates could be higher or lower.

  We found no independent study of its reading program.

  Source: Sylvan’s pricing page

• Teach Your Child to Read in 100 Easy LessonsOne book of lessons that a parent teaches$29.99 list price

  We read two small studies. In 1990, the parents of 40 first-graders were randomly assigned to different ways of helping at home. Children whose parents were trained in a method derived from the book progressed faster than children whose parents only heard them read. In 2016, three kindergartners at risk were taught by their parents for 30 of the lessons. They read more words correctly than before, and there was no comparison group. The school method behind the book has a large research base, and that research is about school programs.

  Sources: Maker’s page, Leach and Siddall, 1990, Review that names the book, McConnell and Kubina, 2016, Stockard and others, 2018

• UFLI Foundations manualA classroom phonics program written for teachers$70.00 for the manual. The lesson slides are free.

  Two studies were done with the developer, and neither was randomized. A 2023 report from WestEd compared kindergartners and first-graders in one Florida district with matched students from the year before. All were below the middle score at the start. At the end of the year, UFLI kindergartners were 1.20 and first-graders 1.42 standard deviations higher on early reading tests. A 2025 journal article found a gain of more than 1.0 over matched students. Its authors are the program’s director, two assistant directors and a WestEd researcher, and a gift from Sue and James Patterson paid for it. The two authors who wrote the program say they do not benefit financially from its distribution. We found no randomized trial.

  Sources: Maker’s page, WestEd report, 2023, Lane and others, 2025, Price

Shelf 4 · 3 things

Research is against it, or a regulator acted

A research review went against the method, or the Federal Trade Commission charged that the maker’s claims about teaching reading had no proof.

• Hooked on PhonicsAn app with phonics materials mailed each month$23.96 a month

  In 1995 the Federal Trade Commission settled charges that Gateway Educational Products had no reasonable basis for broad ad claims that the program teaches reading. The commission did not challenge phonics teaching itself, and the order does not stop anyone buying the product. A federal review found no study that met its standards as of July 2007. We found no independent study of today’s product.

  Sources: Maker’s page, Federal Trade Commission, 1995, What Works Clearinghouse

• Sight-word card sets, as the main methodCards for memorizing whole wordsWe did not collect prices.

  The National Reading Panel found that children taught phonics systematically did better than children taught whole-word programs. From mid-childhood, children learn about 3,000 new words a year. A 2018 research review calls teaching each one as a whole word an “insurmountable task”. The federal guide supports a ring of three to five common words at a time.

  Sources: National Reading Panel, Castles, Rastle and Nation, 2018, Federal practice guide

• Your Baby Can ReadA kit sold for children from 3 months to 5 years oldAbout $200 a kit, as sold from at least January 2008

  In 2012 the Federal Trade Commission charged that its ads were deceptive. It said the sellers had no reliable scientific proof that babies can learn to read. The company settled in 2012 and the creator in 2014. A settlement is not a court ruling.

  Sources: Federal Trade Commission, 2012, Federal Trade Commission, 2014

The free shelf · 9 things

Free things to do

Habits from the federal tips for home, and free materials to read or print.

• Family activities from Florida State UniversityPrintable activities with step-by-step instructionsFree

  The Regional Educational Laboratory—Southeast made them from the federal practice guide for kindergarten to grade 3. We did not test any activity.

  Source: Florida Center for Reading Research

• Play with the sounds in wordsCount the words in a sentence, clap syllables, take “cow” out of “cowboy”Free

  The federal practice guide rates teaching sounds and how they link to letters “Strong Evidence”. Those studies were in schools.

  Sources: Federal tips for home, Federal practice guide

• Public library: Every Child Ready to ReadA parent program some libraries run, for children from birth to age 5Free

  We did not check which California libraries run it, or any study of its results. Ask your branch.

  Source: Every Child Ready to Read

• Read aloud to your child and talk about the bookAsk open questions before, during and after the storyFree

  A 2008 federal review of 19 studies of reading together with young children found spoken language 0.73 of a standard deviation higher than for children not read to that way. Print knowledge was 0.50 higher, in 4 studies. It found no effect on letter names or hearing sounds in words, with only 2 studies of each. The practice guide rates the matching recommendation “Minimal Evidence”.

  Sources: National Early Literacy Panel, Federal tips for home, Federal practice guide

• Reading 101: A Guide for ParentsA guide by grade and by skill from Reading Rockets, a public television projectFree

  Reading Rockets made it with the National Education Association and Colorín Colorado.

  Source: Reading Rockets

• Sound out words, left to rightSay each letter sound, then join the sounds into the wordFree

  The federal practice guide rates teaching children to decode words “Strong Evidence”. Those studies were in schools.

  Sources: Federal tips for home, Federal practice guide

• Take turns reading aloudRead in turns, read together, or read a passage and have your child repeat itFree

  The federal practice guide rates daily reading of connected text “Moderate Evidence”. Those studies were in schools.

  Sources: Federal tips for home, Federal practice guide

• Tips for Supporting Reading Skills at HomeFour tips for kindergarten to grade 3 from the U.S. Department of EducationFree

  The department says the studies behind the tips were in schools. It did not review studies of parents using them at home.

  Source: What Works Clearinghouse

• UFLI Foundations ToolboxFree lesson slides, passages a beginner can sound out, and gamesFree

  These go with a classroom program. The lesson plans are in the paid manual, and the studies of the program are on the shelves.

  Source: University of Florida Literacy Institute

Prices are the regular prices we read on November 11, 2026, and sale prices are left out. A number such as 0.29 is an effect size. It is the gap between a group that got the teaching and a group that did not, counted in standard deviations, the usual spread of scores. “We found no study” describes our search, and a product with no study may still teach well.

See every card as a table

Every thing on the shelves: its shelf, what it is, its price on November 11, 2026, what the research is, and the sources.

Name	Shelf	What it is	Price on November 11, 2026	What the research is	Sources
Lexia Core5 Reading	1. The thing itself was tested	Reading software that schools buy	We found no price for families.	Evidence for ESSA, a rating group at Johns Hopkins, rates it Moderate. Its 4 studies covered 17,281 students and found an average effect of +0.16, compared with students who did not use it. The largest compared more than 6,000 students in grades 3 to 5 with matched students in similar schools, on a reading test. Lexia published one of the 4. These were studies of use in school.	Evidence for ESSA
Tutoring by a person	1. The thing itself was tested	One-on-one or small-group teaching, from any provider	Prices vary. We did not collect them.	A 2024 review pooled randomized trials of tutoring for children from pre-kindergarten to grade 12, in reading and math. Tutoring raised learning by about 0.29 of a standard deviation on average, compared with students who were not tutored. Effects tended to be larger with teachers or school aides as tutors, in earlier grades, and at 3 or more days a week. Those are patterns across studies that differed in many ways.	Nickow, Oreopoulos and Quan, 2024, Stanford summary
All About Reading, Level 1	2. The teaching method is tested. The product is not.	A phonics kit with lightly scripted lessons	$159.95, plus $25.95 for letter tiles and $17.95 for a review box	We found no study of this program. The maker’s main page says it covers phonics, vocabulary and comprehension. A customer quote there calls it Orton-Gillingham. A 2021 review of Orton-Gillingham teaching for children who struggle to read found a gain of 0.22 on basic reading skills over the comparison groups. The review calls that not statistically significant (p = .40).	Maker’s page for Level 1, Maker’s main page, Stevens and others, 2021
Bob Books, Set 1	2. The teaching method is tested. The product is not.	12 short books a beginner can sound out	$17.99 for the set	We found no study of Bob Books. The federal practice guide recommends books a beginner can sound out for practice. A 2018 research review supports books like these “in the very early stages” of reading. It says their limits are likely to outweigh their benefits after that.	Maker’s page, Federal practice guide, Castles, Rastle and Nation, 2018
Duolingo ABC	2. The teaching method is tested. The product is not.	A learn-to-read app	Free	The maker’s page calls it a comprehensive phonics curriculum and says it is based on recommendations by the National Reading Panel. We could not find a study of the app that we could open.	Maker’s page
Logic of English, Foundations A	2. The teaching method is tested. The product is not.	A program with its own readers	$97.96 regular price for the Foundations A bundle. The maker says Core Materials are also required, and the Foundations A and Core Materials bundle lists at $251.86.	We found no study of this program. The maker’s page says Foundations “uses a structured approach to help children understand the relationship between letters and sounds”.	Maker’s page
ABCmouse	3. Small or maker-linked studies, or none we found	A learning app for young children	$14.99 a month after 30 free days, until canceled	The research its page points to is the maker’s own. We found no What Works Clearinghouse report on it. In 2020 the maker agreed to pay $10 million to settle federal charges about renewals and cancellations. That case was about billing and made no finding about learning.	Maker’s page, Federal Trade Commission, 2020
Homer	3. Small or maker-linked studies, or none we found	An early reading app	The maker’s page shows a 30-day free trial and no price.	One randomized study covered 82 children in 7 Head Start classrooms. They used the app 15 minutes a day for 6 weeks, and a comparison group used a math app. It measured skills that come before reading, such as hearing the sounds in words. A New York University professor wrote the summary, and the company hosts it. The summary does not say who paid. We found no journal article.	Study summary, hosted by the company
Khan Academy Kids	3. Small or maker-linked studies, or none we found	A learning app for young children	Free	A University of Massachusetts team ran one randomized trial at the maker’s request. The Overdeck Family Foundation helped pay for it. It covered 49 children aged 4 and 5, mostly from low-income families, for 10 weeks. The children using the app went from the 34th percentile to about the 47th on pre-reading skills. They gained more than a comparison group that used other apps. It did not measure reading. The 2020 news release called the results preliminary.	University of Massachusetts Amherst
Kumon reading	3. Small or maker-linked studies, or none we found	Center visits twice a week, with worksheets at home on the other five days	Kumon’s page shows no price. Ask the center.	We found no independent study of its reading program, and its page cites none. Its preschool program starts with “look, listen and repeat” word work before phonics.	Kumon’s page
Reading Eggs	3. Small or maker-linked studies, or none we found	A learn-to-read app	$13.99 a month for the reading and math plan for up to 4 children, free for the first 30 days	The maker says it worked with a partner firm on a study of “the relationship between program usage and students’ achievement”. That kind of study cannot show the app caused the scores. We found no randomized trial.	Maker’s account of its study
Starfall	3. Small or maker-linked studies, or none we found	A learn-to-read website	The learn-to-read lessons are free. A home membership is $35.00 a year.	The maker hired Metis Associates, an evaluation firm, to study its kindergarten classroom program in one Colorado school district. In spring, 58.8% of the 267 Title 1 kindergartners who had the program scored proficient or higher on a reading assessment, against 31.1% of the 164 who did not. It was not a randomized trial, and the evaluators asked for a larger randomized study. We found no study of home use.	Metis Associates report, Price
Sylvan	3. Small or maker-linked studies, or none we found	Tutoring at a center	$50 or more a session. Prices vary by location, and rates could be higher or lower.	We found no independent study of its reading program.	Sylvan’s pricing page
Teach Your Child to Read in 100 Easy Lessons	3. Small or maker-linked studies, or none we found	One book of lessons that a parent teaches	$29.99 list price	We read two small studies. In 1990, the parents of 40 first-graders were randomly assigned to different ways of helping at home. Children whose parents were trained in a method derived from the book progressed faster than children whose parents only heard them read. In 2016, three kindergartners at risk were taught by their parents for 30 of the lessons. They read more words correctly than before, and there was no comparison group. The school method behind the book has a large research base, and that research is about school programs.	Leach and Siddall, 1990, Review that names the book, McConnell and Kubina, 2016, Stockard and others, 2018
UFLI Foundations manual	3. Small or maker-linked studies, or none we found	A classroom phonics program written for teachers	$70.00 for the manual. The lesson slides are free.	Two studies were done with the developer, and neither was randomized. A 2023 report from WestEd compared kindergartners and first-graders in one Florida district with matched students from the year before. All were below the middle score at the start. At the end of the year, UFLI kindergartners were 1.20 and first-graders 1.42 standard deviations higher on early reading tests. A 2025 journal article found a gain of more than 1.0 over matched students. Its authors are the program’s director, two assistant directors and a WestEd researcher, and a gift from Sue and James Patterson paid for it. The two authors who wrote the program say they do not benefit financially from its distribution. We found no randomized trial.	WestEd report, 2023, Lane and others, 2025, Price
Hooked on Phonics	4. Research is against it, or a regulator acted	An app with phonics materials mailed each month	$23.96 a month	In 1995 the Federal Trade Commission settled charges that Gateway Educational Products had no reasonable basis for broad ad claims that the program teaches reading. The commission did not challenge phonics teaching itself, and the order does not stop anyone buying the product. A federal review found no study that met its standards as of July 2007. We found no independent study of today’s product.	Federal Trade Commission, 1995, What Works Clearinghouse
Sight-word card sets, as the main method	4. Research is against it, or a regulator acted	Cards for memorizing whole words	We did not collect prices.	The National Reading Panel found that children taught phonics systematically did better than children taught whole-word programs. From mid-childhood, children learn about 3,000 new words a year. A 2018 research review calls teaching each one as a whole word an “insurmountable task”. The federal guide supports a ring of three to five common words at a time.	National Reading Panel, Castles, Rastle and Nation, 2018, Federal practice guide
Your Baby Can Read	4. Research is against it, or a regulator acted	A kit sold for children from 3 months to 5 years old	About $200 a kit, as sold from at least January 2008	In 2012 the Federal Trade Commission charged that its ads were deceptive. It said the sellers had no reliable scientific proof that babies can learn to read. The company settled in 2012 and the creator in 2014. A settlement is not a court ruling.	Federal Trade Commission, 2012, Federal Trade Commission, 2014
Family activities from Florida State University	The free shelf	Printable activities with step-by-step instructions	Free	The Regional Educational Laboratory—Southeast made them from the federal practice guide for kindergarten to grade 3. We did not test any activity.	Florida Center for Reading Research
Play with the sounds in words	The free shelf	Count the words in a sentence, clap syllables, take “cow” out of “cowboy”	Free	The federal practice guide rates teaching sounds and how they link to letters “Strong Evidence”. Those studies were in schools.	Federal tips for home, Federal practice guide
Public library: Every Child Ready to Read	The free shelf	A parent program some libraries run, for children from birth to age 5	Free	We did not check which California libraries run it, or any study of its results. Ask your branch.	Every Child Ready to Read
Read aloud to your child and talk about the book	The free shelf	Ask open questions before, during and after the story	Free	A 2008 federal review of 19 studies of reading together with young children found spoken language 0.73 of a standard deviation higher than for children not read to that way. Print knowledge was 0.50 higher, in 4 studies. It found no effect on letter names or hearing sounds in words, with only 2 studies of each. The practice guide rates the matching recommendation “Minimal Evidence”.	National Early Literacy Panel, Federal tips for home, Federal practice guide
Reading 101: A Guide for Parents	The free shelf	A guide by grade and by skill from Reading Rockets, a public television project	Free	Reading Rockets made it with the National Education Association and Colorín Colorado.	Reading Rockets
Sound out words, left to right	The free shelf	Say each letter sound, then join the sounds into the word	Free	The federal practice guide rates teaching children to decode words “Strong Evidence”. Those studies were in schools.	Federal tips for home, Federal practice guide
Take turns reading aloud	The free shelf	Read in turns, read together, or read a passage and have your child repeat it	Free	The federal practice guide rates daily reading of connected text “Moderate Evidence”. Those studies were in schools.	Federal tips for home, Federal practice guide
Tips for Supporting Reading Skills at Home	The free shelf	Four tips for kindergarten to grade 3 from the U.S. Department of Education	Free	The department says the studies behind the tips were in schools. It did not review studies of parents using them at home.	What Works Clearinghouse
UFLI Foundations Toolbox	The free shelf	Free lesson slides, passages a beginner can sound out, and games	Free	These go with a classroom program. The lesson plans are in the paid manual, and the studies of the program are on the shelves.	University of Florida Literacy Institute

What does the research say teaches a child to read?

Teaching children how letters stand for sounds, in a planned order, has strong research behind it.

That teaching is called phonics. “Systematic” phonics means the letter sounds are taught in a set order, and the child practices each one.

In 2000 the National Reading Panel pooled the studies it could find. Researchers report results as an effect size. It is the gap between a group that got the teaching and a group that did not. The gap is counted in standard deviations, a usual spread of scores. An effect size of 1.0 means the teaching group’s average was one standard deviation above the other group’s. Systematic phonics had an effect of 0.44 compared with unsystematic phonics or none. The panel called that moderate.

Timing mattered. The effect was 0.56 in kindergarten and 0.54 in first grade. It was 0.27 in grades 2 through 6. The panel found that no named phonics program did reliably better than another. In its words, the programs “do not appear to differ significantly from each other”. Most programs had only three or four comparisons, so the panel said more evidence was needed.

Sound games have support too. Phonemic awareness is hearing and changing the separate sounds in a spoken word. Teaching it had an effect of 0.53 on reading compared with no such teaching. The panel found it worked better when taught with letters.

The U.S. Department of Education’s practice guide for kindergarten through grade 3 rates the research behind each of its four recommendations. Linking sounds to letters and teaching children to decode words are both rated “Strong Evidence”. Decoding means turning letters into sounds and joining them into a word. Daily reading of connected text is rated “Moderate Evidence”. Teaching vocabulary and school language is rated “Minimal Evidence”. That label means the panel could not point to a body of studies showing the teaching raises achievement. Seven studies on this recommendation qualified, and they did not mostly find gains. The guide still recommends the teaching.

England’s Education Endowment Foundation reaches a similar result. It credits phonics with about 5 months of extra progress on average. It rates that research as high in security. It took off one mark because many of the 228 studies were not independently evaluated.

Phonics teaches one of the two parts of reading. A 1986 model that a 2018 research review calls highly influential says reading is decoding multiplied by understanding language. A child who can sound out every word still needs to know what the words mean. The panel put it this way: “Phonics teaching is a means to an end.”

So the method has support. A product that uses the method is a separate question.

Which books and programs are worth buying?

Choose books and kits by method and price, because studies of them are few, small or done with the maker.

The National Reading Panel found no named phonics program did reliably better than another. It said more evidence was needed. The panel did not test price. A higher price does not by itself show a better program.

Books and kits

Bob Books are short books a beginner can sound out. These are called decodable books. Set 1 has 12 books and its regular price was $17.99 on November 11, 2026. We found no study of Bob Books. The federal guide recommends decodable text for practice, and a 2018 research review supports it “in the very early stages”. The same review says the limits of these books are likely to outweigh the benefits after those stages.

Teach Your Child to Read in 100 Easy Lessons is one book that a parent teaches from. Its list price is $29.99. We read two small studies of it. In a 1990 study, 40 first-graders’ parents were randomly assigned to one of four ways of helping at home. Children whose parents were trained in a method derived from the book progressed faster. The comparison children had parents who only heard them read. In a 2016 study, parents taught three kindergartners at risk for 15 minutes a night, five nights a week. The parents covered 30 of the 100 lessons. The children read more words correctly than before, and the study had no comparison group. The book is built on Direct Instruction, a school method with a large research base of 328 studies. That research is about school programs.

UFLI Foundations is a classroom program from the University of Florida. The lesson slides, passages and games are free, and the teacher’s manual costs $70.00. It was written for teachers, and we found no research on parents using it. You may hear that it has large effects. Two studies were done with its developer, and neither was randomized. A 2023 report came from WestEd, an education research group. It compared kindergartners and first-graders in one Florida district with matched students from the year before. All of them started below the middle score on a reading test. At the end of the year, UFLI kindergartners were 1.20 standard deviations higher on early reading tests. First-graders were 1.42 higher. The report measures each gap against the spread of scores at the start of the year. A 2025 journal article found a gain of more than 1.0 over matched students. Its authors are the program’s director, two of its assistant directors and a WestEd researcher. A gift from Sue and James Patterson paid for it. The two authors who wrote the program say they do not benefit financially from its distribution. We found no randomized trial.

All About Reading Level 1 costs $159.95, plus $25.95 for letter tiles and $17.95 for a review box. Logic of English sells its Foundations A bundle at a regular price of $97.96. The publisher says Core Materials are also required, and the Foundations A and Core Materials bundle lists at $251.86. We found no study of either program. A customer quote on All About Reading’s main page calls it an Orton-Gillingham program. Orton-Gillingham is a way of teaching reading that many programs say they follow. The next section covers what that name is worth.

Hooked on Phonics costs $23.96 a month for an app plus mailed materials. We found no independent study of the current product. The federal What Works Clearinghouse found no study that met its standards as of July 2007. In 1995 the Federal Trade Commission settled charges against Gateway Educational Products. The charges were that it had no reasonable basis for broad advertising claims that the program teaches reading. The commission said it “did not challenge the educational value of phonics instruction itself.”

Apps

These studies measured different things. The Khan Academy Kids and Homer studies measured skills that come before reading, in preschoolers, over 6 to 10 weeks. The Starfall and Lexia studies measured reading in school classes.

• Khan Academy Kids is free. A University of Massachusetts team ran a randomized trial at the maker’s request. The Overdeck Family Foundation helped pay for it. It covered 49 children aged 4 and 5, mostly from low-income families, for 10 weeks. Children using the app went from the 34th percentile to about the 47th on pre-reading skills. They gained more than a comparison group that used other apps. The 2020 news release called the results preliminary.
• Homer has one study that we found, published by the company. A New York University professor wrote the summary. It covered 82 children in 7 Head Start classrooms for 6 weeks, and the comparison group used a math app. The maker’s page shows a free trial and no price.
• Reading Eggs costs $13.99 a month after 30 free days. That plan has reading and math for up to 4 children. Its maker says a partner firm studied “the relationship between program usage and students’ achievement”. A study like that cannot show the app caused the scores.
• Starfall has free learn-to-read lessons. The maker hired an evaluation firm for one study of its kindergarten classroom program in one Colorado district. In spring, 58.8% of the 267 Title 1 kindergartners who had the program scored proficient or higher on a reading assessment. Of the 164 who did not have it, 31.1% did. The study was not randomized, and the evaluators recommended a large randomized trial in other districts.
• ABCmouse costs $14.99 a month after 30 free days, “until canceled”. The research its page points to is its maker’s own. In 2020 the maker agreed to pay $10 million to settle federal charges about renewals and cancellations. That case was about billing, and it made no finding about learning.
• Duolingo ABC is free. Its maker says the lessons follow recommendations by the National Reading Panel. We could not find a study of the app that we could open.

Lexia Core5 Reading is software that schools buy. Evidence for ESSA, a rating group at Johns Hopkins, rates it Moderate. Its 4 studies covered 17,281 students. The average effect was +0.16 compared with students who did not use it. The largest study followed more than 6,000 students in grades 3 to 5. It compared them with matched students in similar schools, on a reading test. Lexia published one of the 4. We found no price for families.

England’s toolkit adds a caution about apps in general. It says digital phonics tends to be less successful than phonics led by a teacher or teaching assistant.

Tutoring, Kumon and Sylvan

Tutoring by a person has controlled studies behind it. A 2024 review pooled randomized trials of tutoring for children from pre-kindergarten to grade 12, in reading and math. Tutoring raised learning by about 0.29 of a standard deviation on average, compared with students who were not tutored. Effects tended to be larger with teachers or school aides as tutors and in the early grades. They also tended to be larger at 3 or more days a week. Those are patterns across studies that differed in many ways.

That research is about tutoring in general. It tells you little about one company. Kumon describes two center visits a week, with worksheets at home on the other five days. Its preschool program starts with “look, listen and repeat” word work before phonics. Stanford’s tutoring center defines high-impact tutoring as three or more sessions a week. We found no independent study of Kumon’s or Sylvan’s reading programs, so we can’t tell you whether either works. Kumon’s page shows no price. Sylvan lists tutoring at $50 or more a session and says prices vary by location.

Some things on the shelves have research or a regulator against them, and a few common claims do too.

What should we skip?

We would skip any program that promises a baby will read, teaches guessing, or makes whole-word memorizing the main method.

Three other claims need care before you pay more for them.

Baby reading kits

In 2012 the Federal Trade Commission charged the sellers of Your Baby Can Read with deceptive advertising. The kits cost about $200 each. The commission said the sellers lacked reliable scientific proof that babies can learn to read. The company settled that year and the creator settled in 2014. A settlement is not a court ruling that the law was broken.

Guessing from the picture or the first letter

Some programs teach a child to work out a word from the picture, the sentence and the first letter. Teachers call this three-cueing. The federal practice guide tells teachers to “discourage students from guessing unknown words using beginning letters or pictures.” When your child is stuck on a word, point to the letters and sound it out together. An occasional look at the picture is still typical in kindergarten and first grade, say the California Dyslexia Guidelines.

Big sight-word sets as the main method

The National Reading Panel found that children taught phonics systematically did better than children taught whole-word programs. From mid-childhood, children learn about 3,000 new words a year. The 2018 review calls teaching each one as a whole word “an insurmountable task”. A few words are fine. The federal guide suggests a ring of three to five common words at a time. New cards are added as the child learns them.

Paying more for the Orton-Gillingham name

Orton-Gillingham is a way of teaching reading that many programs say they follow. The name is not one product. A 2021 review pooled studies of Orton-Gillingham teaching for children who struggle with reading. The gain on basic reading skills was 0.22 over the comparison groups. The review calls that not statistically significant (p = .40). A p of .40 means a gap at least this size would turn up about 4 times in 10 if the teaching made no difference. In 2010 the federal clearinghouse found 31 studies of Orton-Gillingham-based teaching with no brand name, for students with learning disabilities. None met its standards. The studies are few and weak. The fair reading is that programs using the name have not been shown to work better than other step-by-step phonics.

Reading Recovery, if your school offers it

Reading Recovery is a program that gives first graders one-on-one lessons with a teacher. A large randomized trial found big gains in first grade. The effect was above 0.50 compared with similar children who had not yet started. A parallel follow-up was not randomized. It compared children just below and just above each school’s cutoff score for the program. The outcome was state tests in grades 3 and 4. Children who had the program scored 0.19 to 0.43 standard deviations below similar children who did not. The follow-up lost more than 75% of its original sample. Its authors say the results “should be taken seriously” anyway. If a school offers it, ask what comes after first grade.

Games sold to raise school performance

In 2015 the Federal Trade Commission settled with the makers of Jungle Rangers, a computer game for children. The makers must stop unsupported claims that the game improves school performance.

Some habits with research behind them cost nothing.

What can we do at home for free?

The U.S. Department of Education publishes four tips for home, and they cost nothing to use.

One limit comes with them. The department says the studies behind the tips were done in schools. It “did not assess studies of these practices involving parents or caregivers at home.”

1. Talk about the book before, during and after reading. Ask open questions such as “Why did the character do what he did?”
2. Play with the sounds in spoken words. Count the words in a sentence, or hold up a finger for each syllable. Ask what is left when you take “cow” out of “cowboy”.
3. Sound out words from left to right. Say each letter sound, join them, then ask whether that is a real word.
4. Read aloud together. Take turns, or read the same passage at the same time. You can also read a passage and have your child read it back.

Tip 4 has your child read out loud. Reading a book to a young child is a different thing. A 2008 federal review looked at 19 studies of reading together with young children. Their spoken language was 0.73 of a standard deviation higher than for children who were not read to that way. Print knowledge, a child’s understanding of what print is and how it is used, was 0.50 higher, in 4 studies. It found no effect on letter names or on hearing the sounds in words, with only 2 studies of each. It said that was too few studies to be sure. A separate federal report on dialogic reading found gains in spoken language and no clear effect on hearing the sounds in words. In dialogic reading, the adult asks the child questions during a story. We would keep reading aloud for language, and add sound and letter work beside it.

Free materials from sources we could check:

• Reading 101: A Guide for Parents from Reading Rockets, a project of the public broadcaster WETA.
• Family activities listed by the Florida Center for Reading Research at Florida State University. The Regional Educational Laboratory—Southeast made them from the federal practice guide.
• The UFLI Foundations Toolbox, which has passages a beginner can sound out.
• Your public library. Some libraries run a parent program called Every Child Ready to Read for children from birth to age 5. We did not check which California libraries run it, so ask your branch.

Some signs call for something a product cannot give.

Is my child behind, and when is it time to get an evaluation?

We would ask the school district for an evaluation if the signs below last past the typical ages.

First, what is typical.

California’s guidelines say formal reading teaching begins in kindergarten. Simple sounding out starts that year, and first graders read longer sentences. In one New Zealand study, state-school children began formal reading lessons at 5. Children in Steiner schools, which start formal teaching later, began at 7. The children who began earlier read better at first. By age 11 the difference was gone. The study was not an experiment. Its authors say the finding may not hold for children in less fortunate circumstances. They also warn against reading it as a reason to give no reading teaching before 7.

The California Dyslexia Guidelines list signs that students in kindergarten and first grade who have dyslexia may show. The International Dyslexia Association names slow or inaccurate reading and poor spelling among the symptoms of dyslexia. The state’s list includes:

• “Inability to create a rhyme for a simple one-syllable word like ‘hot’ or ‘cat'”
• “Inability to remember the names of the letters of the alphabet or consistent confusion over the letters of the alphabet”
• “Difficulty sounding out one-syllable words (e.g., ‘dog,’ ‘hop,’ ‘bat’)”
• “A tendency to make reading errors that are not connected to the sounds of the letters on the page”
• “Inability to separate words into their individual sounds by the end of first grade”

The same page lists what is still typical at this age. That list includes difficulty with some of these signs through the middle of kindergarten. It also includes confusing “b” and “d”, especially in writing. The association also calls it a myth that people with dyslexia “read backwards”.

A sign is a reason to ask. It does not tell you what your child has. An evaluation is meant to determine whether a child has a disability and what the child’s educational needs are.

Who can request one, and how fast

Federal rules say a parent may request an initial evaluation. In California, a written request is called a referral. If the district agrees to assess your child, it must give you a written assessment plan. The plan is due within 15 days of the referral. Days between school terms, and school vacations longer than five school days, do not count. You can agree in writing to more time. A referral in the last 10 days of a school year gets a plan within 10 days after the next year starts.

Once you consent in writing, the district has 60 days. In that time it must decide whether your child qualifies and hold an IEP meeting. Vacations and a written extension change that count too. The 60 days do not apply if a parent repeatedly does not bring the child to the assessment. They also do not apply when a child moves to a new district partway through. In that case you and the new district agree on a date.

A district can also say no. If it does, federal rules say it must give you written notice that explains why. Our guide to IEPs in California covers how to write the request and what happens next.

What it costs

Federal rules say the public agency, meaning the school district, must conduct an initial evaluation. They also say special education and related services are “provided at public expense” and “without charge”. The California Department of Education says supports and services found appropriate should be provided “at no cost to the family”. We did not find one California page that says the district’s assessment itself is free to parents.

A private evaluation is a different matter. The International Dyslexia Association’s Dyslexia Basics page says a full evaluation of this kind in the U.S. typically costs $500 to $2,000, and that more detailed testing can reach $1,500 to $5,000. The page gives no source for those figures.

If the school says to wait

The Dyslexia Guidelines say a district “cannot delay or deny assessment of a student for a suspected disability because of implementation of Response to Intervention”. That is the classroom help many schools try first. The guidelines also say a parent “may request an initial evaluation at any time”.

Schools are changing too, which affects what your child gets before you ever ask.

What is California changing in its schools?

California now screens children from kindergarten to grade 2 for reading risk, and a 2025 law covers teaching materials.

These rules are as we read them in November 2026.

Start with where children are. In spring 2026, 44.09% of California third graders scored Proficient or Advanced in English. English here means reading, writing, listening and research, so the test measures more than reading. It is one of California’s state tests (CAASPP). Proficient and Advanced are the top two of four levels. Earlier years called them Met and Exceeded, and the state says the cut scores did not change. The year before, the figure was 44.21%. About 391,000 third graders took the test in 2026, by our read of the state’s research file.

Screening

Since the 2025–26 school year, districts must screen kindergarten, first and second graders every year (Education Code 53008). The test must come from the state’s approved list. You can opt out in writing. The law has other rules for children who enroll late and for children who do not speak English well. In some cases a parent can exempt a child with written consent, such as when the child already has a diagnosis. Kindergartners are not screened before the 91st school day. The school must share results within 45 calendar days. The law says a result is “a flag for potential risk of reading difficulties, not as a diagnosis of a disability.” It also says screening cannot be used to delay a special education evaluation.

Teaching materials

Assembly Bill 1454 was signed on October 9, 2025. It tells the State Board of Education to adopt English materials for kindergarten through grade 8 by January 31, 2027. The materials must include “explicit and systematic instruction” in phonics and the other basic reading skills. A district may choose materials that are not on the state’s list. It must then certify that they meet the same criteria. The law also sets dates in 2028 for new standards in some teacher and administrator training programs.

The law’s text does not mention three-cueing. We searched it for “cueing” and found nothing, so we would not say California has banned the method.

EdReports, a nonprofit that reviews teaching materials, has reviewed two school reading programs. It found that Fountas & Pinnell Classroom for kindergarten through grade 2 did “not meet the expectations”. It found the 2018 kindergarten Units of Study lacked “systematic and explicit instruction in all foundational skills standards”. A review like this reads the materials. It is not a trial of how children learned with them, and we did not check the newer editions.