In data modeling, what does a network schema allow that a hierarchical schema does not—specifically with respect to relationship cardinalities?

Introduction / Context:
Logical database models—hierarchical, network, and relational—differ in how they represent relationships. Understanding these differences is crucial when selecting a model for complex domains or when translating legacy schemas into modern relational designs.

Given Data / Assumptions:

• Hierarchical models form strict trees (one parent per child).
• Network models generalize to graphs using owner/member sets.
• Relational models represent data in tables with keys and can express many-to-many via junction tables.

Concept / Approach:

Network schemas (CODASYL model) allow records to participate in multiple sets, enabling many-to-many relationships directly. This contrasts with hierarchical schemas, which restrict child records to a single parent, and with relational schemas that use tables, not sets, to encode relationships. Thus, the salient capability of the network model is direct support for many-to-many via set membership.

Step-by-Step Solution:

Verification / Alternative check:

Legacy CODASYL documentation and textbooks describe owner–member sets enabling non-tree relationships, confirming the answer.

Why Other Options Are Wrong:

Restricts to one-to-many: That is hierarchical, not network.

Stores data in tables: That is relational, not network.

All of the above / None: Both contradict the specific capability of the network model.

Common Pitfalls:

Conflating logical models (hierarchical/network/relational) with physical storage or modern SQL implementations.

Final Answer:

Permits many-to-many relationships via set membership