The science

What the science of reading actually says about phonics apps

By Ryan King · Updated June 2026 · Sourced.

Most apps that say they teach reading do not. A working knowledge of the published research lets you tell the few that do from the many that do not, in about five minutes per app.

There was a reading war, and it is over

For thirty years a public debate ran in schools and at kitchen tables about how children learn to read. The two camps were systematic phonics, which teaches letter-sound relationships explicitly and in a deliberate order, and whole language, which exposes children to text and trusts them to absorb the system.

The debate is settled. The National Reading Panel concluded in 2000 that systematic phonics instruction produces significantly better outcomes than unsystematic or no-phonics approaches. The most comprehensive review of the modern research, Castles, Rastle, and Nation's Ending the Reading Wars (2018), reaches the same conclusion with twenty years of additional evidence. England's Department for Education has required systematic synthetic phonics in primary schools since 2007.

What a real phonics app does

Five things, every time:

• Letters arrive in a deliberate order. The classic order is SATPIN: s, a, t, p, i, n. Those six letters combine into many real words almost immediately, which is why reading programs and apps that take phonics seriously use that order.
• Children learn the sound a letter makes, anchored to a picture and a key word. The letter S makes the sss sound, like sun.
• Sounds are blended into words explicitly. The child is asked to push the sounds together, not just to hear them said and repeat.
• Text is decodable. The child is never asked to read a word containing letters and patterns the program has not taught yet.
• Comprehension is checked. Reading a sentence is not the same as understanding it. The good apps ask the child to prove they read the sentence, usually by picking a matching picture.

What phonics theater looks like

Alphabet songs, letter recognition flashcards, animated stories that the child cannot decode themselves, and reward systems for tapping the right thing without ever sounding it out. These activities are not harmful and many children enjoy them. They are not phonics. They will not by themselves teach a child to read.

A useful test, free and unscientific, is to open an app and ask three questions. Does it teach letter sounds explicitly. Does it ask the child to blend. Are the sentences the child is asked to read decodable. If the answer to any of those is no, the app is alphabet entertainment, not reading instruction.

What we built into Cairn Read

Cairn Read is systematic synthetic phonics wrapped in a block-world adventure. Letters arrive in SATPIN order. Every word the child reads is built from letters already taught. The Word Builder stage gives the child a beat to attempt the blend before the narrator reveals it. The Sentence Quest stage ends with a picture comprehension check. The pedagogy note inside the app cites the studies behind each game.

The science is not new. The discipline of building games around it instead of around engagement metrics is what is rare.

Sources: National Reading Panel report (2000); Castles, Rastle, Nation (2018), Ending the Reading Wars, Psychological Science in the Public Interest; Letters and Sounds, England Department for Education.