The term statistical inference refers to techniques for obtaining information about a statistical model's parameters based on data from the model. There are two different but related types of question about the parameter (or parameters) that we might ask.
Uncertainty and strength of evidence
A distribution's parameter cannot be determined exactly from a single random sample — there is a 5% chance that a 95% confidence interval will not include the true parameter value.
In a similar way, a single random sample can rarely provide enough information about a parameter to allow us to be sure whether or not any statement about it will be true. The best we can hope for is an indication of the strength of the evidence against the statement.