Prohibiting which of the following techniques is MOST helpful in preventing users from obtaining confidential data by using statistical queries?

Sequences of queries repeatedly targeting the same small population enable an inference attack: by running slightly varied statistical queries (e.g., SUM, COUNT, AVG) against an overlapping subset, an attacker can mathematically isolate and deduce individual confidential values. Prohibiting such sequences - a technique known as query set size control or query overlap restriction - directly mitigates this statistical inference threat by preventing the attacker from narrowing results to a single record.

Accessing multiple databases in repeated queries describes a distributed query pattern, not a statistical inference technique; it does not specifically enable deduction of individual records through aggregate statistics.

Selecting all records and displaying all columns is a direct data exposure issue (a SELECT * violation), not a statistical query inference attack, and is addressed by access controls and column-level permissions rather than query restriction policies.

Running queries that access sensitive data is a broad activity that normal authorized users must perform; restricting it entirely would prevent legitimate use and does not specifically address the statistical aggregation technique used in inference attacks.