Relational Model and SQL: 10 practice questions

Question 1 of 10

In PostgreSQL terminology, what does the word relation refer to?

1. AA guaranteed ordering of a table's rows
2. BA table, the structure that stores the rows
3. CA grouping of databases on a single server
4. DA declared link between two tables of a database

Question 2 of 10

Which arrangement is fixed inside a row but left unguaranteed for the rows of a table?

1. AThe data type of every column
2. BThe order of the columns
3. CThe name given to each column
4. DThe number of columns per row

Question 3 of 10

Which part of a query states which values the output should contain?

1. AThe client application
2. BThe select list
3. CThe table expression
4. DThe sort specification

Question 4 of 10

What is true of every row belonging to one given table?

1. AIt carries whichever columns were supplied when that row was written
2. BIt carries at least one column whose data type is decided per row
3. CIt carries the same set of named columns, each with a specific data type
4. DIt carries a stored position that fixes where it appears in results

Question 5 of 10

A table already carries three separate unique constraints. How many primary keys may it also declare?

1. AOne for each unique constraint that is also not null
2. BNone, because the unique constraints already identify rows
3. CAt most one, whatever the number of unique constraints
4. DAny number, provided every listed column is not null

Question 6 of 10

Which columns can a foreign key constraint legitimately point at?

1. AColumns carrying any index that speeds up the lookup
2. BColumns declared not null in the referenced table
3. CA primary key, a unique constraint, or a non-partial unique index
4. DColumns whose data types match those of the referencing columns exactly

Question 7 of 10

What does a SELECT statement written without any table expression do?

1. AIt reports a syntax error until a source table is supplied in a FROM clause
2. BIt reads every table the current database contains
3. CIt returns the previous result the session produced
4. DIt evaluates the expressions in its select list and returns them

Question 8 of 10

A table has two columns that are each unique and never null. Other tables must be able to reference this table without naming a column list, and client tooling must be able to identify a row from one declared identity. Which design meets both needs?

1. ADeclare one of the two columns the primary key and leave the other as a unique constraint
2. BDeclare both columns as primary keys so either can be referenced without a column list
3. CDeclare the pair of columns together as one primary key covering both columns
4. DDeclare both columns as unique constraints and let each reference name the column it wants

Question 9 of 10

A report must show a value computed from two numeric columns of one table, must not change which rows come back, and must expose that value as an output column. Where does the calculation belong?

1. AIn the table expression, as an extra source combined with the base table
2. BIn the sort specification, so the value is produced while ordering the rows
3. CIn the select list, as a computed entry beside the column references
4. DIn the row qualification, so only rows with the wanted value are produced

Question 10 of 10

An application lists stored records and users report that two runs of the same unsorted query presented the records in different sequences. No data changed between the runs. What accounts for this and what fixes it?

1. AThe rows were regrouped into another database, so the query has to name it
2. BColumn order shifted between runs, so the select list has to name its columns
3. CThe result set is cached per session, so the session has to be reopened
4. DRow order in a table is not guaranteed, so an explicit sort has to be requested