How can correlational study is that this one is used to describe the relationships between two variables, particularly smoking and GPA.

How can correlational study is that this one is used to describe the relationships between two variables, particularly smoking and GPA. Multiple observations are needed as one case is not enough to establish or refute a relation. Data should always be graphed as it reveals more than summary statistics? How can you accurately determine whether an anecdote is typical requires statistical evidence?