With a higher number of nocturnal tornado events, I surveyed residents of Alabama, Arkansas, Kentucky, Mississippi, and Tennessee (N = 487 for each sample) to determine the time, in minutes, it took to reach a decision on shelter-seeking. I utilized latent class analysis (LCA) to create class memberships, based on diurnal and nocturnal scenarios, to associate with time intervals. Four actors were identified for each scenario: Tech Users, Typical Actors, Non-Reactors, and Social Actors for the day sample, Tech Users, Typical Actors, Passive Actors, and Non-Reactors for the night sample. Time intervals were created and applied to each class. All class assignments except one, Traditional Actors in the night sample, used more time than allotted in an average tornado warning lead time (~15 minutes). Future studies may be necessary to determine a reduction in time needed for decision-making, such as establishing the most impactful warning sources.