The Relational Model and Normalization Questions and Answers

Answer: C

Correct answer: The columns must be ordered.

Column order is precisely what a relation does not require: attributes are identified by name, so the columns may be written in any sequence without changing the relation. The other three are genuine requirements - single-valued cells, homogeneous columns, and no duplicate rows - and a table that breaks any of them is not a relation.

Answer: A

Correct answer: True

The vocabularies correspond directly: relation, tuple and attribute in the formal relational terminology are table, record and field in the older file-processing one. The terms are used interchangeably in practice, and being fluent in both matters because textbooks and product documentation mix them freely.

Answer: A

Correct answer: A → B.

If A determines the pair (B, C), then it determines each of them separately: given a value of A, the value of B is fixed and so is the value of C. This is the decomposition rule of Armstrong’s axioms. The converse does not hold, so nothing follows in the reverse direction, and the other three options all reverse the arrow.

Answer: A

Correct answer: True

This is the definition of a key. It may consist of a single attribute or of several, in which case it is a composite key, and it must be minimal, so that dropping any attribute from it would destroy uniqueness. A relation may have several such candidate keys, one of which is chosen as the primary key.

Answer: A

Correct answer: True

This is the definition. R → S asserts that each value of R is associated with exactly one value of S, so knowing R is enough to determine S. R is the determinant, S the dependent attribute. It is a statement about the meaning of the data, so no set of sample rows can prove it - a single counterexample refutes it.

Answer: B

Correct answer: modification anomalies.

A poorly structured relation makes some changes impossible to perform cleanly: an insertion anomaly prevents a fact being recorded until some unrelated fact is known, a deletion anomaly destroys information as a side effect, and an update anomaly requires the same value to be changed in many places. Normalisation exists to remove them.

Answer: A

Correct answer: True

Attributes are identified by name, so writing the columns in a different order produces the same relation. The same is true of rows, which form a set and so have no inherent order either. Both facts follow from the relation being defined by its content rather than by any arrangement of it.