The previous will SELECT the first_name FROM grade_database WHERE the grade of the individual is greater than of equal to 80. This will give us 'Anthony', 'Anna', and 'Andrea' using th ecomparison operator >=. We could also use =, <, >, <=, >=, and <>(not equal).
This will give us 'Anthony', 'Anna', and 'Andrea', this time because it is asking for the first_name of those who scored between 80-90.
'Peter' was sick and was unable to take the test. Asking for IS NULL will return those with no value in grade.
IS NOT NULL would give us the opposite of IS NULL
Now we are asking for the last_name of individuals whose first_name is LIKE or had 'An' somwhere in it. This will give us 'Ferragano', 'Marie', and 'Barnoon'.
The % is used to identify the missing text surrounding the desired pattern. It could be used before/after your desire pattern. _ is used to identify only one missing character in a pattern.
group by specific characteristics
having by specific characteristics
placing in order
is the default and will order from small to large
is the default and will order from large to small