Posttraumatic Stress in Young Children: Causal Network Model from LONGSCAN

Glenn N. Saxe applied Causal Data Science methods to an existing observational, longitudinal dataset from the Consortium for Longitudinal Studies in Child Abuse and Neglect (LONGSCAN). The dataset contains a sample of 1,354 children identified in infancy to early childhood as being maltreated or at risk. The resulting causal network model includes 251 variables and 818 bivariate relations, revealing direct causes and effects of Posttraumatic Stress at age 8.

Use Cases

• Causal inference analysis based on the discovered network of 251 variables and 818 bivariate relations.
• Longitudinal study of child development based on data collected from infancy to age 8.
• Modeling risk factors for Posttraumatic Stress based on variables like Emotional Maltreatment and Witnessing Violence.
• Evaluating Causal Data Science algorithms as an alternative to experimental methods in human research.

Strengths

• Dataset is derived from a substantial longitudinal study of 1,354 children.
• The causal model is explicitly defined with 251 nodes and 818 edges.
• Data is shared under a permissive CC-BY-4.0 license.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is very small (19.7 KB), indicating it likely contains summary or model results rather than raw observational data.

Provenance

Source

Consortium for Longitudinal Studies in Child Abuse and Neglect (LONGSCAN).

Collection Method

Observational, longitudinal data collection from a sample identified as maltreated or at risk.

Time Range

Data collected from infancy to age 8.

Freshness

Last updated 2026-04-22 05:49:30; freshness should be verified.