How we score

Every product is evaluated against 9 developmental literacies derived from synthesized research. Not safety ratings. Not feature lists. What does this actually build in your child?


The 9 Literacies

Nine literacies organized into three domains. Not skills with a half-life — literacies that become more important as AI gets more capable.

Doing

  • Agency

    Can I start?

  • Persistence

    Can I keep going?

  • Adaptability

    Can I change course?

Thinking

  • Curiosity

    Can I wonder?

  • Creativity

    Can I make something new?

  • Judgment

    Can I decide well?

Being

  • Connection

    Can I relate?

  • Self-Regulation

    Can I manage what I feel?

  • Purpose

    Can I find meaning?

Each literacy is backed by a deep research synthesis. Explore the framework


Understanding Tiers

Every scored product receives a headline tier based on how many literacies it builds strongly.

Exceptional

6+ Strong literacies, zero Limited

A developmental powerhouse. Builds broadly and deeply across domains.

Recommended

3+ Strong, or at least 1 Strong literacy

Demonstrates clear developmental strengths. Products with even one Strong literacy earn a "Best For" badge.

Reviewed

Scored, no Strong literacies

Some developmental value, but no standout strengths in our framework.

Caution

Design patterns actively undermine 2+ literacies

Rare. Reserved for products with anti-developmental mechanics. Requires editorial justification.

Important: The NL Score measures developmental literacy-building, not product quality. A Reviewed product is not a bad product. A times-table app can be excellent at teaching math while having no standout developmental strengths in our framework.


How We Score

Every score follows a structured process. The credibility comes from the research-derived framework, not the scorer.

1

Evidence gathering

We build a review corpus from academic research, editorial reviews, parent communities, and product documentation. Every source is cited with a URL.

2

Rubric application

Each literacy has 4-6 evaluation questions derived from research. These structure the analysis — they are considered holistically, not scored individually.

3

AI-assisted analysis

AI agents apply the rubric to the evidence corpus, producing draft ratings (Strong, Moderate, Limited) and a 2-4 sentence rationale per literacy.

4

Editorial review

Every profile is reviewed by an editor who adjusts ratings, checks reasoning, adds context, and discloses any conflicts of interest. The editorial layer is the credibility layer.

Scored by our research-derived framework, with AI-assisted analysis and editorial review.


The Literacy Signature

Every product gets a visual fingerprint — a compact representation of its developmental profile across all three domains.

Example: A product with 5 Strong literacies

5 Strong
Doing Thinking Being

Filled dot = Strong. The product demonstrably builds this literacy based on evidence.

Empty dot = Not Strong. The product may still support this literacy at a Moderate or Limited level.

Dots are colored by domain: blue for Doing, violet for Thinking, emerald for Being. The shape of the dots tells you the shape of the product's developmental impact at a glance.


They tell you it's safe. We tell you what it builds.

Most product guides for parents answer one of two questions: "Is it safe?" or "Is it worth the money?" Those are useful questions. We ask a different one: what does this actually build in your child?

Common Sense Media focuses on safety, age-appropriateness, and content flags. Essential — but it doesn't tell you whether Minecraft develops persistence or just entertains.

Wirecutter focuses on product quality and value. Excellent for buying decisions — but structurally blind to curricula, AI tools, and anything you can't buy on Amazon.

We score against a developmental framework backed by 50,000+ words of synthesized research. The question isn't just "is this good?" — it's "what specific literacies does this develop, and how strongly?"