Validity of a test or assessment is important to researchers and to the participants taking them. Much of what we read today in magazines and newspapers are not valid. An example of an item that has face validity, but no real validity is an opinion poll in a popular magazine. To the person taking it, it looks real and official (face validity). In actuality, the poll is unscientific and could be completely biased.

Construct validity concerns how well or accurately an assessment measures a construct (the thing being measured). For example, how well an anxiety assessment would measure anxiety.

Content validity is how well the content or items of an assessment make sense in terms of the construct. A question asking your favorite number would not be valid on an assessment measuring depression.

Criterion validity is whether or not a test can accurately predict some future event. An example if this is that the GRE test should predict how well a student will perform in grad school.

Convergent validity is simply scores on a certain test should be related to scores on another test measuring the same thing.

Discriminant validity is just the opposite of convergent validity. Scores on a certain assessment should not be related to scores on another assessment measuring something different.