The impact of phantom decoys on choices in cats.

Context-dependent choice is an important aspect of decision making. The paper examines context-dependent choice in cats (Felis catus), with particular reference to the effect of local context, on the basis of hypotheses developed in the field of human decision making. Cats were initially confronted with two different feeding options. This binary choice set was later manipulated incorporating a decoy that was better than the available options but ultimately unavailable (a phantom). By means of a within-subjects manipulation of phantom location in the attribute space, the author compared the effects of close and distant phantoms on the final choices. The main finding is that close phantom decoys affected choice behavior of cats by altering the overall share of the available options, leading some animals to reject even some of the available feeding options, and by causing the animals to favor the available option that was more similar to the phantom decoy. No significant effects emerged for phantoms that were far from the alternatives in the attribute space. The strengths of this paper lie in its novel approach and high originality. No other study has used dominating decoys with animals or decoys that are unattainable. This paper provides strong links to the human decision making literature, the presentation of the predictions of a range of different choice models, and the novelty of the application to animals. The use of a phantom decoy is particularly interesting because the phantom cannot actually be chosen, and thus the binary and trinary choice sets both have the very same choices available. Overall, the effect of phantoms is real, interesting and new.