ParentsPick

Methodology

How ParentsPick analyzes children's books and produces its content classifications.

What ParentsPick analyzes

ParentsPick provides factual content analysis for 8,990 children's and young adult books. Each book is classified across 9 content themes. Classifications report what is present in a book — not whether it is appropriate. ParentsPick does not rate, recommend, or warn against any title.

How books are analyzed

For each book, an AI classification system searches trusted web sources — including Common Sense Media — for information about the book's content. It performs targeted searches for each of the 9 themes, collecting results from established review sites and educational resources.

The search results are then analyzed by an AI model (GPT-4o-mini) using a detailed classification prompt. The model determines whether each theme is present in the book's narrative and assigns a confidence level based on the quality of evidence. Only content that appears in the book itself is flagged — not film adaptations, academic interpretations, author biography, or absence of representation.

Age-appropriate thresholds are applied: a stomachache in a picture book is not flagged as violence, and mild suspense is not flagged as scary content. Thresholds scale with the target age range of the book.

The 9 content themes

ViolencePhysical conflict, fighting, or depictions of harm.
Scary contentFrightening scenes, suspense, or horror elements.
Religious themesReferences to faith, prayer, religious figures or practices.
Racial/cultural contentRacial identity, cultural heritage, or diversity themes.
ProfanityMild to strong language or crude humour.
Climate changeEnvironmental themes, climate crisis, or conservation.
Sexual identitySexual content or references to sexual identity.
Gender rolesTraditional or non-traditional gender expectations.
LGBTQ+ themesLGBTQ+ identities or same-sex relationships.

Confidence levels

Each classification includes a confidence rating:

  • High — Multiple trusted sources explicitly confirm the content is present in the book. This is the threshold used for all public-facing statistics.
  • Medium — Evidence suggests the content is present, but sources are limited or less specific.
  • Low — Weak or indirect evidence. The content may be present but cannot be confirmed with certainty.

Books classified as “unclear” are not counted in any published statistics.

Current database statistics

ThemeBooks% of 8,990
Violence2,08923.2%
Scary content6817.6%
Religious themes3744.2%
Racial/cultural content7458.3%
Profanity2793.1%
Climate change670.7%
Sexual identity7358.2%
Gender roles3884.3%
LGBTQ+ themes8059.0%

Counts reflect high-confidence classifications only. Last updated March 2026.

Data quality

When a title exists as multiple editions — translations, large print, coloring books — ParentsPick identifies the most representative edition and uses it as the canonical record. This prevents duplicate titles from inflating statistics.

Limitations: The classification system relies on publicly available information about each book. Books with limited online coverage may receive lower confidence ratings or incomplete analysis. The system does not read the full text of each book.

How to use this data

ParentsPick reports what is in a book. It does not tell you whether a book is right for your child. Every family has different values, and what matters to one parent may not matter to another. Use this data as one input in your decision — not the final word.