Evidence

What each exercise is based on

Each one is based on a published method, and names it. Adapted for home practice, not copied.

Read this first. The findings below come from other people’s research into these methods. They are not results for Sayo, and we did not take part in any of it.

Script TrainingAphasia treatment literature; used for the phone-call and counter setsWhole conversations rehearsed as scripts, then produced with the support removed a layer at a time.IN THE LITERATUREConsistent gains in the fluency and accuracy of the scripts that were trained. Generalisation to untrained conversation is limited — which is why the scripts here are the ones you actually need.
Semantic Feature AnalysisWord-finding treatment literatureWord finding through meaning first — what it is for, where it lives — and sound last.IN THE LITERATUREImproved naming of the treated words, with some carry-over to untreated words in the same category.
Phonological Components AnalysisWord-finding treatment literatureThe same target approached through its sounds: first sound, rhyme, number of beats.IN THE LITERATUREImproved naming of treated words. Works for people who get little from the meaning route, so the two are offered as alternatives, not a sequence.
Verb Network Strengthening (VNeST)Sentence-level aphasia treatment literatureSentences built around the verb — who did it, what they did, what it was done to.IN THE LITERATUREGains at sentence level rather than single words, including on sentences that were not trained.
Oral Reading for Language in Aphasia (ORLA)Reading-aloud treatment literatureReading in unison with a voice that keeps time, then reading alone.IN THE LITERATUREImproved reading aloud and, in several studies, spoken language beyond reading. Designed for repeated, high-dose practice — the kind an app can actually deliver.
Goal Management TrainingCicerone practice standard for executive functionStopping to plan before starting. The pause is the skill being trained.IN THE LITERATUREBetter performance on everyday goal-directed tasks. Rated a practice standard, the strongest tier in the Cicerone reviews.
Spaced retrieval and external cueingINCOG 2.0 memory recommendationsPractice distributed across days, and reminders put where they will catch you.IN THE LITERATURESpacing beats massed repetition for retention. External aids carry INCOG 2.0’s strongest recommendation tier for memory — stronger than internal-strategy training.
Contextualised attention trainingINCOG 2.0 attention recommendationsAttention practiced inside real tasks. The guidelines advise against abstract drill, so there is none.IN THE LITERATUREAttention training helps when it is embedded in real tasks. INCOG 2.0 specifically advises against decontextualised computer drill, which is why there are no coloured squares here.
Errorless learningLearning-method literature for moderate to severe impairmentShown before asked, so a wrong guess is never rehearsed alongside the right answer.IN THE LITERATUREBetter learning than trial-and-error for people with more severe memory impairment. For milder impairment, generating and correcting your own errors can be better — so the app switches by level.
Stroke rehabilitation guidanceCSBPR 7th editionSession length, frequency and the practice-between-appointments model.IN THE LITERATUREMore practice is associated with better outcomes, and practice between sessions is explicitly recommended.
Read those findings carefully

They describe what the method achieved in supervised research, usually with a therapist in the room. They are not results for Sayo, and they are not a prediction about you.

Word-level difficulty is graded on published norms — SUBTLEX-US frequency, Kuperman age-of-acquisition and Brysbaert concreteness ratings — rather than on our own judgement of what looks hard.

The full reference list sits in our clinical evidence review. Email clinical@sayo.app and we will send it.

What we do not claim

No trial of our own

The methods are published and studied. Sayo itself has not been through a clinical trial, and we will say so until it has.

No predicted recovery

We will not tell you where you will be in six weeks. Level ladders describe how an exercise gets harder — they are not a forecast about you.

No diagnosis, no treatment claim

Sayo does not diagnose, treat or cure anything, and it is not a medical device.