It depends on what your original research question was. If a correlation is significant, the basic information that it can supply is whether the direction of the relationship was negative or positive.

You should always visualize the relation in a bivariate plot. Check for outliers, variances of the measures, and see if the relation is linear. If not, your correlation may yield insignificant results and you will need a better model (if your n is sufficient).

Not enough and we have to test and not just value@

A bi-variate plot is sometime known as a scatter plot, which can show you both non-linear relationships and outliers.