Frequentist statistics include the methods that most people are first exposed to. If you don’t know what you’re doing, you’re probably a frequentist (the converse is not true).
More technically, frequentist statistics in statistical inference refers to the choice to model the unknown parameter as fixed and unknown (as opposed to random and unknown like in Bayesian statistics). Frequentist results do not include any user-specified priors. For instance, confidence intervals are considered frequentist, whereas credible intervals are considered Bayesian. When we evaluate frequentist methods, we look at what happens if we were to repeatedly run the experiment.
Philosophically it is justified by the frequentist interpretation of probability.