Stanford Survey of Health and Development

What is the Stanford Survey of Health and Development?

The Stanford Survey of Health and Development — Stanford SHQ for short — is a longitudinal study of early childhood development run from Stanford's Language and Cognition Lab. Rather than asking what a child should be doing at a given age, the study asked parents of 21,861 children across the United States, born between 2018 and 2021, to tick off which behaviours they had actually observed. The result is one of the largest single normative datasets covering the first five years of life.

Unlike a clinical assessment, the SHQ is not used to diagnose or score individual children. It is a population reference: for any observable behaviour in the survey, the dataset tells us at roughly what age a given share of children begin showing it, and at what age most have. That gives Learning Curve an empirical floor — not a teacher's expectation or a publisher's recommendation, but what large numbers of parents have actually reported.

How the SHQ structures development

The survey organises observable behaviours into four subscales, each covering a broad developmental territory:

• Gross Motor — what the body does in space. From early reflexes (grasping, stretching) through sitting, crawling, walking, climbing, and running.
• Problem Solving — how the child engages with objects and cause-and-effect. Looking, reaching, exploring, fitting, sorting, and early symbolic play.
• Oral Language — receptive and expressive language. From attending to voices and turn-taking through first words, phrases, and conversational exchanges.
• Emotional Awareness — social referencing, comfort seeking, recognising others' feelings, and the beginnings of self-regulation.

These four are not the only domains of development — they were chosen because they appear consistently across decades of paediatric and developmental literature, and because they could be reliably observed by parents at home. Within each subscale, the dataset spans 0 to 60 months, organised into age bins so that any given behaviour sits at the age range where most children typically first show it.

What the SHQ deliberately does not cover

Two limits to call out honestly. First, the SHQ describes observable behaviour, not curriculum content. There is no Literacy or Mathematics subscale; no Expressive Arts; no school-readiness checklist. For those, look at frameworks like EYFS, the Common Core, or national curricula — each linked from our frameworks index.

Second, the SHQ is descriptive — it does not include adult-side advice. Each item describes what a child does, not what a parent should do about it. That is why every Stanford construct page on Learning Curve shows the observable behaviour verbatim but does not yet carry activity recommendations of its own. Activity authoring against the SHQ observable set is a separate piece of work, queued behind the dataset itself.

How it complements EYFS and other frameworks

Different frameworks ask different questions. EYFS asks “what should a five-year-old be ready for at the end of reception?” The SHQ asks “what do most children, in fact, do, and when?” Bayley-4 and Denver II ask “is this child within the typical clinical range?” The answers overlap, but the lenses don't.

Where SHQ is strongest is in its sheer sample size and its focus on observable, parent-reportable behaviours across the under-three years. Where it is thinner is in cross-cultural breadth — the sample is US-centric — and in coverage of older preschoolers, where most items taper off around age four. We use the SHQ alongside EYFS, WHO GSED v2, CREDI, and Bayley-4 so that any single milestone on Learning Curve sits inside a triangulated picture rather than relying on one source.

Why it's in Learning Curve

The Stanford SHQ underwrites a large share of the early-window data behind our developmental milestones. When we say “most children begin to walk unsupported between roughly twelve and seventeen months”, that statement traces back, in significant part, to this dataset. The observable-behaviour granularity also means we can link directly from a Learning Curve construct to a specific SHQ item, rather than the more abstract framing some frameworks use.

The full survey methodology and item bank are available through the Stanford Language and Cognition Lab. The links below point to the lab and its open data; the citation chain ends in the published paper by Stenhaug, Ram & Frank (2021).