Maximal Elements Under Reference-Dependent Preferences with Applictions to Behavioral Traps and Games Artículo académico