Research foundation
Built on approaches studied around the world.
SayBloom draws on approaches studied in the UK, Sweden, Australia, India, Pakistan, the US and elsewhere, such as parent-mediated communication, video modeling and AAC. Here is what we build on, and how each idea shows up in the app.
Please read: what this page is and isn't
SayBloom is an educational and communication tool. It is not a medical device and is not a medical or clinical treatment. It doesn't diagnose, treat or prevent any condition, and it isn't a substitute for a speech-language pathologist, behavior analyst, occupational therapist or doctor.
The studies below informed our design. They evaluated the original approaches, usually delivered by trained people, not SayBloom itself. Citations are summarized from the clinical literature and are being verified by our clinical advisors; please check the original sources before relying on them. SayBloom's own pilot results are coming 2027.
Approaches and how SayBloom uses them
Scroll the table sideways on small screens.
| Approach | Where it was studied & key evidence | How SayBloom uses it |
|---|---|---|
| Video modeling | US, UK, Australia. Classed as evidence-based by the US National Clearinghouse on Autism Evidence & Practice.Steinbrenner et al., 2020 | Short clips of peers greeting, ordering food or asking for help, then the learner practices. |
| AAC & PECS | Global. Meta-analyses show AAC does not stop speech and often increases it.Millar, Light & Schlosser, 2006 | Tap-to-speak I Need board, always one tap away, so speech practice never blocks communication. |
| Parent-mediated communication (PACT) | UK randomized trial; gains still present about six years later in follow-up.Green et al., Lancet 2010; Pickles et al., Lancet 2016 | Coach mode shows parents how to respond to and expand their child's attempts. |
| PASS (adapted PACT) | India & Pakistan randomized trial; delivered by non-specialists.Rahman et al., Lancet Psychiatry 2016 | Supports the idea that well-designed practice can happen at home, not only in clinics. |
| Social skills groups (KONTAKT) | Sweden randomized trial.Choque Olsson et al., JAMA Psychiatry 2017 | Structured conversation scripts in Talk Trails: greet, ask, answer, follow up, close. |
| JASPER / ESDM | US and Italy trials.Kasari; Dawson et al., 2010; Colombi et al., 2018 | Joint-attention games and play-based turn taking. |
| Rapid Syllable Transition (ReST) | Australia randomized trial for childhood apraxia of speech.Murray, McCabe & Ballard, 2015 | Syllable drills in Say It Clear with feedback on timing and stress. |
| DTTC | US (Strand); studies report gains in children with severe apraxia. | Slow, cued imitation: watch the mouth, say it together, then alone. |
| Nuffield Dyspraxia Programme (NDP3) | UK; compared with ReST in a randomized trial.Murray et al., 2015 | Picture-based sound sequences built from simple to complex. |
| Auditory-Motor Mapping Training | US study with minimally verbal autistic children.Wan et al., 2011 | Sing-and-tap: words sung on two pitches while tapping drum pads on screen. |
| Music therapy | Cochrane review found improved social interaction; the large TIME-A trial found no change in overall symptom severity.Geretsegger et al., 2014; Bieleninik et al., JAMA 2017 | Music supports calm and engagement. SayBloom does not claim music treats autism. |
| Social Stories | UK and US; Carol Gray method. | Illustrated stories in My Day for new events: a haircut, a fire drill, a doctor visit. |
| Zones-style regulation & Positive Behavior Support | Widely used in US, UK and Australian schools. | Color-coded feelings check-in with a matching calm-down plan. |
| Functional Communication Training | Widely replicated since the original study.Carr & Durand, 1985 | Quick messages like “break please” get the same result as a challenging behavior, so the behavior is no longer needed. |
| Slow-paced breathing | Heart rate variability biofeedback studies.Lehrer & Gevirtz, 2014 | Breathing flower at about six breaths per minute, synced with music. |
| Robot / avatar practice | Japan and Europe.e.g. Kumazaki et al.; DREAM project (EU) | Bloo, a friendly animated guide, is less socially demanding than a human face at first. |
| Easy-to-Read | Inclusion Europe standards. | Short sentences, one idea per line, a picture beside every key word. |
| Repeated food exposure | Repeated-exposure and sensory feeding research; feeding therapy reviews.e.g. Sharp et al. reviews | Food Explorer's five calm steps: look, touch, smell, kiss, tiny taste. |
Language
The home language comes first.
Studies of bilingual autistic children find no harm from using the family's language (Hambly & Fombonne, 2012). Families should never be told to drop their language.
SayBloom launches in English and Spanish. Spanish, with five pure vowels and mostly open syllables, is one of the easier major languages to pronounce. Inside each language, words are ordered by how easy they are to say, not by the alphabet: lip sounds and open vowels first (“ma,” “more,” “bye,” “up”), then simple two-syllable words.
Masking & framing
Skills, not pressure to hide who you are.
Research with autistic adults links constant “masking” (hiding autistic traits) to exhaustion, anxiety and depression (Hull et al., 2017; Cassidy et al., 2018). That's why SayBloom frames everything around independence, clear communication, confidence and self-regulation. The app teaches skills; the learner keeps the choice of when to use them.
Nutrition & wellness evidence
Every wellness suggestion shows its evidence grade.
SayBloom suggests topics to discuss with a doctor or registered dietitian. It doesn't prescribe supplements, doses or restrictive diets.
| Topic | Evidence | What SayBloom suggests |
|---|---|---|
| Constipation & GI pain | Strong GI problems are common in autism and linked to irritability and challenging behavior.Buie et al., Pediatrics 2010; Chaidez et al., 2014 | Screen everyone; prompt a doctor visit; track fiber, water and stools. |
| Sleep & melatonin | Strong A European randomized trial of pediatric prolonged-release melatonin improved sleep in autistic children.Gringras et al., 2017 | Track sleep; suggest discussing it with a doctor if sleep is poor. |
| Selective eating (ARFID) | Strong Feeding therapy and gradual food exposure help.Sharp et al. reviews | Food Explorer with gentle steps and rewards. |
| Iron & vitamin D | Moderate Deficiency is more common in selective eaters; low iron is linked to restless sleep. | Suggest asking the doctor about a blood test, not self-supplementing. |
| Artificial food colors | Moderate UK Southampton trial found increased hyperactivity in general-population children.McCann et al., Lancet 2007 | Optional 4-week trial with tracking. |
| Omega-3 | Weak / mixed Some meta-analyses show small effects on hyperactivity, others none. | Discuss with a doctor; measure with a one-change trial. |
| Probiotics | Early / mixed Small trials, inconsistent results. | Discuss with a doctor; measure if tried. |
| Gluten-free, casein-free diet | Not supported Controlled trials show no consistent benefit.Hyman et al., 2016; Piwowarczyk et al., 2018 | Only with dietitian supervision, as a measured trial. |
| Sugar and hyperactivity | Not supported Meta-analysis of blinded trials.Wolraich et al., JAMA 1995 | Focus on regular meals instead. |
Advisory board
Guided by clinicians, families and autistic adults.
We're forming an advisory board of a speech-language pathologist, a BCBA, a special education teacher, an occupational therapist, two parents and two autistic adults. We'll also seek an independent university research partner.