Human likes and dislikes can be established or changed in numerous ways. Three of the most well-studied procedures involve exposing people to regularities in the environment (evaluative conditioning, approach-avoidance, mere exposure), to verbal information about upcoming regularities (evaluative conditioning, approach-avoidance, or mere exposure information), or to verbal information about the evaluative properties of an attitude object (persuasive messages). In the present study, we investigated the relation between, on the one hand, different types of experiment-related beliefs (regularity, influence, and hypothesis awareness) and demand reactions (demand compliance and reactance) and, on the other hand, evaluative learning about novel food brands (Experiments 1 and 2) and well-known food brands (Experiment 2) via persuasive messages, experienced regularities, and verbal information about regularities. Participants were first exposed to an evaluative learning phase and then completed self-reported evaluation ratings, an Implicit Association Test, and a behavioral intention measure. Results indicate that regularity awareness was a necessary condition for most evaluative learning effects. Influence awareness was also a strong moderator of evaluative effects but more so for effects on self-reported ratings. Hypothesis awareness and reactance only weakly moderated evaluative learning, and demand compliance only played an important role for well-known brands. The theoretical and practical implications of our findings are discussed. (PsycInfo Database Record (c) 2025 APA, all rights reserved).