Treat Loyalty: Examining Stability in Preference for Food Items Over Time in Owned Dogs

Abstract

Identifying valuable reinforcers is an essential aspect of producing and maintaining behavior change in animals. Relatively more valuable reinforcers can increase the amount of behavior maintained, shorten the latency to respond, and compete with other environmental reinforcers. Preference assessments are frequently used to identify relative reinforcer value. Prior studies have shown that the results of preference assessments typically align with relative reinforcer efficacy, indicating that preference assessments are a useful tool to identify effective reinforcers. However, there is little research on the stability of dogs' preferences for certain reinforcers across time. The present study assessed the stability of dogs' preferences for food items over time. To evaluate this, we enrolled eight owned dogs, and for each dog we conducted repeated paired stimulus preference assessment with eight food items over a 6-month period. Dogs' preferences were measured based on the number of times the food item was chosen in a session. We used Cohen's kappa statistic to measure stability based on the consistency in which food items remained in the top three, middle two, and bottom three ranks across time. Our results revealed stability in the top three and bottom three ranked food items throughout the assessments for most participants. The middle-ranked food items were consistently preferred at this medium level, but their rank did vary slightly across time. Additionally, multiple dogs ranked the same food items as higher in preference and lower in preference. Compared to Session 1, food preference stability was greater when rankings in Session 3 were used to assess agreement, suggesting that dogs may need experience with the procedure and food items to establish a preference. Finally, several dogs showed an increase or decrease in stability following intervals between assessments, suggesting that there are potential factors that the dog can interact with in between time intervals that can impact their preference stability. Altogether, several dogs showed a preference for certain types of treats compared to others. Overall preference patterns were seen to be more stable with increased exposure, indicating that a single assessment might not accurately reflect dogs' preferences, and repeated assessments should be conducted over time. Analysis of rank-order stability revealed that the bins remained relatively stable over time, with variation in the actual rank in repeat assessments.