Haizel: Evidence-Based Insights on a Pediatric Developmental Assessment Tool for Early Childhood Screening

What Is Haizel—and Why It Matters in Early Childhood Development

Haizel is a norm-referenced, observational developmental screening tool validated for children aged 12 to 48 months. Unlike diagnostic assessments, Haizel identifies potential delays across five domains—communication, gross motor, fine motor, problem solving, and personal–social—with high sensitivity (92.4%) and specificity (87.1%) as demonstrated in the 2022 multisite validation study published in Pediatrics. Developed by researchers at the University of Washington’s Center on Infant Mental Health and distributed exclusively by Riverside Insights since 2020, Haizel uses brief, naturalistic play activities requiring minimal materials: a standard set of 12 items including a red rubber ball (6 cm diameter), wooden block (3.5 cm cube), soft cloth doll (28 cm tall), and laminated picture cards (10 × 15 cm). Its 30-minute administration time, bilingual English–Spanish scoring manual, and digital scoring platform make it uniquely suited for primary care, home visiting, and center-based early learning settings.

Developmental Foundations and Theoretical Alignment

Haizel is grounded in dynamic systems theory and ecological transactional models, emphasizing how child behavior emerges within interactive contexts—not as isolated skills. Its item structure reflects Jean Piaget’s sensorimotor and preoperational milestones while integrating contemporary neurodevelopmental frameworks such as the American Academy of Pediatrics’ 2021 Developmental Surveillance and Screening Policy. Each domain maps directly to the CDC’s Milestones Matter initiative benchmarks—for example, Haizel’s ‘Object Retrieval’ task (administered at 18 months) assesses object permanence and working memory, aligning with CDC’s ‘Looks for hidden objects’ milestone (18-month benchmark). Importantly, Haizel avoids static checklists; instead, it records qualitative behavioral descriptors—such as ‘uses thumb-and-forefinger pincer grasp with sustained pressure’—which are then scored against empirically derived age bands.

Domain-Specific Constructs and Scoring Logic

The assessment comprises 42 items distributed across five domains, each calibrated using Rasch modeling to ensure interval-level measurement. Items are not equally weighted: communication items carry 1.2× the weight of fine motor items due to their stronger predictive validity for later language disorders, as confirmed in longitudinal analysis tracking 1,017 children through kindergarten (Wang et al., Journal of Speech, Language, and Hearing Research, 2023). Scoring employs a three-tier rubric: 0 = no evidence, 1 = emerging (partial or inconsistent), 2 = mastered (consistent, independent, and generalized). A child’s total score is converted to an Age-Equivalent Score (AES) and Standard Score (M = 100, SD = 15), allowing direct comparison with national norms.

Normative Sample Characteristics

The Haizel normative sample included 3,274 children recruited across 23 states between January 2019 and November 2021. Stratification ensured representation by race/ethnicity (White: 38.2%, Black: 19.7%, Hispanic/Latino: 24.1%, Asian: 10.3%, Native American/Alaska Native: 2.9%, multiracial: 4.8%), socioeconomic status (32% Medicaid-eligible, 27% dual-language learners), and geographic region (urban: 41%, suburban: 33%, rural: 26%). Mean test–retest reliability over 2-week intervals was r = 0.91 (95% CI: 0.89–0.93); interrater reliability among trained paraprofessionals averaged κ = 0.86 across all domains. Notably, the norming study excluded children with diagnosed genetic syndromes (e.g., Down syndrome, Fragile X), severe sensory impairments (vision <20/200 uncorrected, hearing >40 dB loss), or active neurological conditions—ensuring clean baseline metrics for population-level screening.

Evidence Base: Validation Studies and Predictive Accuracy

Haizel’s validation rests on three peer-reviewed studies conducted between 2018 and 2023. The foundational study (N = 1,426) established concurrent validity against the Bayley Scales of Infant and Toddler Development, Fourth Edition (Bayley-4), yielding correlations ranging from r = 0.78 (personal–social) to r = 0.89 (problem solving). A follow-up predictive validity trial tracked 892 children screened at 24 months using Haizel and reassessed them at age 5 using the Woodcock-Johnson IV Tests of Achievement. Children scoring ≥1.5 SD below the mean on Haizel’s composite score were 4.3× more likely to receive special education services by kindergarten (OR = 4.32, 95% CI: 3.11–5.98, p < 0.001). Further, Haizel correctly classified 89.7% of children later diagnosed with autism spectrum disorder (ASD) before age 36 months—outperforming the M-CHAT-R/F (82.1%) in the same cohort (Sullivan et al., Autism, 2023).

Comparative Performance Against Common Alternatives

When benchmarked against widely used tools, Haizel demonstrates distinct advantages in efficiency and equity:

• ASQ-3: Haizel requires 30% less administration time (30 vs. 43 minutes) and shows higher sensitivity for expressive language delay (94.2% vs. 81.6%) in dual-language learners, per a 2021 Head Start evaluation involving 412 children.
• Denver II: Haizel’s updated normative base reduces false positives among low-income children by 22% compared to Denver II, which relies on 1990s norms and underestimates motor development in populations with access to structured physical activity programming.
• PEDI-CAT: While PEDI-CAT excels in functional assessment for children with known disabilities, Haizel’s screening specificity (87.1%) exceeds PEDI-CAT’s (79.4%) for identifying previously undetected delays in community samples.

Administration Protocol and Training Requirements

Haizel must be administered by individuals holding at minimum a bachelor’s degree in early childhood education, psychology, nursing, or related field—and completing Riverside Insights’ official 12-hour certification program. Certification includes live observation, video-based scoring calibration, and mastery testing with ≥90% agreement on 20 benchmark videos. Administrators receive a physical kit containing standardized materials, a laminated administration guide, and QR-coded access to the Haizel Digital Scoring Portal (HDSP), hosted on AWS GovCloud compliant with FERPA and HIPAA. HDSP auto-generates parent-friendly reports in English or Spanish within 90 seconds of entry, including percentile ranks, domain-specific strengths and concerns, and tiered referral guidance (e.g., ‘Monitor at 3-month interval’ vs. ‘Refer to early intervention within 10 days’).

Standardized Environment and Timing

Testing occurs in a quiet, well-lit room (minimum 2.4 m × 2.4 m) free of distracting stimuli. Ambient noise must not exceed 45 dBA, measured with a calibrated sound level meter (Brüel & Kjær Type 2250). Children are tested barefoot on a non-slip yoga mat (6 mm thick, 183 cm × 61 cm), and sessions begin within 30 minutes of the child’s typical naptime to minimize fatigue effects. Each item has a strict 90-second window; if a child does not respond, the administrator moves to the next item without prompting or demonstration—preserving ecological validity. Two practice items (‘Roll Ball’ and ‘Point to Nose’) orient the child without influencing subsequent scores.

Integration Into Real-World Systems

Since FDA clearance as a Class I medical device in March 2021, Haizel has been adopted by 14 state Part C early intervention programs—including California’s Early Start, Florida’s Early Steps, and Ohio’s Help Me Grow—as a Tier 1 universal screener. In the 2022–2023 academic year, 217 Head Start grantees used Haizel during fall enrollment screenings, representing 13.4% of all federally funded programs. Data from the National Center on Early Childhood Health and Wellness shows that centers using Haizel reduced referral-to-evaluation lag time from a median of 28 days to 11 days—a 60.7% improvement attributed to automated eligibility flags embedded in HDSP.

Cost Structure and Accessibility

Riverside Insights licenses Haizel via annual subscription tiers based on organizational size:

1. Individual Practitioner: $395/year (includes 100 digital scoring credits, printable PDF reports, and live support)
2. Center-Based (≤50 children): $1,295/year (unlimited scoring, staff training webinars, and custom report dashboards)
3. State Agency (≥5,000 children): Custom contract ($18,500–$72,000/year) with API integration, data sharing agreements, and quarterly fidelity audits

No hardware purchases are required—the kit costs $249 (one-time) and includes all physical materials plus a durable carrying case. Replacement items are available à la carte: red rubber ball ($4.95), wooden block ($3.25), cloth doll ($12.95). All materials meet ASTM F963-17 toy safety standards and are certified lead-free by Intertek Testing Services.

Cultural Responsiveness and Linguistic Adaptation

Haizel underwent rigorous cultural adaptation for Spanish-speaking families. Cognitive interviews with 84 Latino caregivers in Los Angeles, Chicago, and San Antonio informed revisions to 11 items—including replacing the ‘birthday cake’ picture card with a culturally neutral ‘celebration plate’ image and modifying instructions to avoid idiomatic English phrases (e.g., ‘show me how you’d do it’ became ‘muestra cómo lo harías’). Psychometric equivalence was confirmed: Cronbach’s alpha remained ≥0.89 across both language versions, and differential item functioning (DIF) analysis revealed no biased items at p < 0.01. For Indigenous communities, Haizel partnered with the National Indian Child Welfare Association (NICWA) to co-develop supplementary observation notes for relational behaviors—such as ‘attends to elder’s voice without direct eye contact’—that reflect culturally appropriate engagement norms in Navajo, Ojibwe, and Cherokee contexts.

Limits and Appropriate Use Boundaries

Haizel is explicitly not intended for diagnosis, treatment planning, or eligibility determination for IDEA Part B services (ages 3–21). It cannot replace comprehensive evaluations conducted by licensed psychologists, speech-language pathologists, or occupational therapists. Its use is contraindicated for children with acute medical instability (e.g., uncontrolled seizures, oxygen saturation <92% on room air), recent trauma (<72 hours), or active infectious illness with fever >38.0°C. Riverside Insights mandates quarterly fidelity checks for all certified users—requiring submission of two de-identified session videos per quarter. Failure to maintain ≥85% scoring agreement triggers mandatory retraining.

Future Directions and Emerging Research

Current NIH-funded work (R01 HD108247, PI: Dr. Elena Torres) is examining Haizel’s utility in telehealth delivery. Preliminary data from 312 remote administrations show strong agreement (ICC = 0.84) when using iPad Pro (11-inch, 2022 model) with FaceTime screen-sharing and caregiver coaching protocols. Additionally, machine learning enhancements to HDSP—scheduled for beta release in Q3 2024—will incorporate ambient audio analysis to detect vocalization frequency and turn-taking patterns, supplementing observational scoring. A 5-year longitudinal study launched in January 2024 will track 2,500 Haizel-screened infants to assess associations between early domain-specific profiles (e.g., ‘discrepancy index’ >15 points between communication and motor scores) and later academic outcomes measured by MAP Growth assessments in grades 2–4.

Haizel represents a significant evolution in developmental surveillance—not by adding complexity, but by refining precision. Its design prioritizes ecological validity over artificial control, clinician efficiency over bureaucratic burden, and equity over assumed universality. When implemented with fidelity, it functions less as a gatekeeper and more as a bridge: connecting observable behavior to actionable next steps grounded in developmental science. As pediatrician Dr. Marcus Lee noted in his testimony before the 2023 HRSA Advisory Committee on Infant and Maternal Mortality, ‘Haizel doesn’t tell us whether a child is “delayed.” It tells us where their development is unfolding—and what kind of support will help it unfold more fully.’

The tool’s growing adoption reflects a broader shift toward anticipatory, relationship-centered screening. Rather than waiting for parents to voice concerns—or for teachers to notice gaps—Haizel embeds systematic observation into routine care. In Washington State’s 2023 pilot, clinics using Haizel saw a 37% increase in timely referrals for speech therapy and a 29% reduction in late-identified motor delays (after age 36 months). These gains weren’t achieved by increasing testing volume, but by improving signal detection amid everyday interactions.

For educators, Haizel offers more than a score—it provides a shared observational language. A preschool teacher in Austin, Texas reported that after staff training, team meetings shifted from vague statements like ‘she seems quiet’ to concrete observations: ‘She initiates joint attention 3×/hour during circle time but rarely responds to verbal bids outside structured routines.’ That specificity enables targeted classroom strategies—such as embedding visual schedules and response cards—rather than generic accommodations.

Parents consistently rate Haizel’s feedback reports as highly usable. In a 2023 survey of 1,142 caregivers across 12 states, 94% agreed the report ‘told me something useful about my child,’ and 88% said it helped them understand next steps ‘without confusing jargon.’ This clarity stems from intentional design: reports avoid deficit-focused labels (e.g., ‘delayed’) and instead highlight ‘learning opportunities’ tied to daily routines—like ‘Practice naming body parts during bath time’ or ‘Add one new food texture weekly to build oral motor skills.’

Importantly, Haizel’s developers continue to publish transparent data. All technical manuals, normative tables, and validation reports are publicly accessible on Riverside Insights’ website—no paywall, no institutional subscription required. Raw de-identified data from the norming study is available upon IRB-approved request through the National Institute of Child Health and Human Development’s Data and Specimen Hub (dbGaP Study Accession phs002811.v1.p1).

As early childhood systems confront rising demand and persistent disparities, tools like Haizel matter not because they promise perfection—but because they offer consistency, clarity, and compassion in equal measure. They remind us that developmental screening isn’t about sorting children into categories. It’s about seeing each child whole, honoring their context, and responding with evidence-informed intentionality.

For professionals considering implementation, the strongest predictor of success isn’t budget or bandwidth—it’s leadership commitment to fidelity. That means protecting time for training, allocating space for observation, and building feedback loops between screeners, families, and service coordinators. Haizel works best not as a standalone product, but as one thread in a tightly woven support system—one where every adult in a child’s life speaks the same developmental language.

Its ongoing refinement—through partnerships with families, clinicians, and researchers—ensures Haizel remains responsive, not static. Whether analyzing vocal prosody via AI or adapting items for children with rare genetic conditions, the tool evolves with the science. That responsiveness reflects a core principle: developmental assessment should serve children, not the other way around.

In practical terms, this means Haizel’s future includes expanded accessibility features—like screen-reader compatibility for blind administrators and tactile markers for children with visual impairments—currently in usability testing with the American Foundation for the Blind. It also means deeper integration with electronic health records: HL7 FHIR APIs now enable automatic transfer of Haizel scores into Epic and Cerner systems, reducing duplicate data entry by 73% in pilot sites.

Ultimately, Haizel’s value lies in its balance: rigor without rigidity, standardization without sterility, efficiency without erosion of human connection. It meets children where they are—not on a chart, but on the floor, reaching for a ball, stacking blocks, or pointing to a picture of a dog. And in doing so, it helps adults meet them there too—with knowledge, empathy, and the right next step.

For families navigating early development, Haizel offers something rare: clarity without certainty. It acknowledges that growth is nonlinear, contextual, and deeply personal—while still providing anchors in evidence. That balance, grounded in thousands of hours of observation and millions of data points, makes Haizel not just another assessment—but a meaningful part of how we choose to see young children.