Data Models into Database Designs Questions and Answers

Answer: A

Correct answer: True

Three actions on the parent - insert, update and delete - and the same three on the child give six in all, and each must be considered separately when the referential integrity rules are worked out. Some will be permitted, some cascaded and some prohibited, and the choice follows from the minimum cardinalities.

Answer: A

Correct answer: either table without specifying parent and child tables.

In a one-to-one relationship each row on either side matches at most one on the other, so a single foreign key column serves whichever table it is placed in. There is no many side to force the choice, and the decision is made on practical grounds - which table is more often queried, or which participation is optional.

Answer: C

Correct answer: the child table.

The foreign key always goes on the many side. Each child row points to one parent, so a single foreign key column suffices; putting it in the parent would require room for an unknown number of child references in one row, which a relation cannot provide. That asymmetry is what makes parent and child meaningful in a 1:N relationship.

Answer: A

Correct answer: True

Update actions on a required parent exist to handle a change in its primary key, and a surrogate key never changes. With nothing to propagate, cascading updates have no work to do and the case can be disregarded. Delete actions still need thought, since removing a parent would orphan its children whatever kind of key it has.

Answer: A

Correct answer: True

If the parent is required, no child may exist without one, so the design must ensure that the parent cannot be removed or its key changed while children refer to it. Either the change is cascaded to the children, keeping the references valid, or it is prohibited outright. Doing neither leaves orphaned children, which the requirement forbids.

Answer: D

Correct answer: All of the above use ID dependent entities

An ID-dependent entity cannot exist without its parent and takes the parent’s identifier as part of its own. All three cases have that character: an association relationship carries data of its own on an intersection between two parents, a multivalued attribute becomes a table dependent on the entity it describes, and an archetype and its instances stand in the same relation.

Answer: A

Correct answer: True

With one row on each side matching at most one on the other, a single foreign key serves whichever table holds it, so either may be treated as the parent. The choice is made on practical grounds - which side is optional, or which table is queried more often - rather than being forced by the cardinality.

Answer: B

Correct answer: False

If the parent is required, a child row may only be inserted when a parent already exists for it to refer to, so the insertion has to be checked or the parent created first. The action that is always permitted in that case is inserting a new parent, since a parent needs no child.

Answer: B

Correct answer: table.

The first step of the transformation is that each entity becomes a table, with its attributes as columns and its identifier as the primary key. A tuple is one row of that table, corresponding to one instance of the entity rather than to the entity type itself.

Answer: C

Correct answer: A primary key

Every table must have a primary key, since without one its rows cannot be identified and duplicates cannot be prevented. Foreign keys appear only where the table participates as a child in a relationship, alternate keys only where more than one candidate exists, and a surrogate key only where one has been introduced deliberately.