A possible factor contributing to the preferred downward orientation may relate to the configuration of the design elements.
While Experiment 2 ruled out a number of possible confounding explanations for some of the results of Experiment 1, it may be the case that the other design elements still provide a context for specific orientation preferences.
This would suggest that the configuration of the elements and package, when taken together, constitute an important constraint in terms of optimal design.
In other words, it could be that the preferred orientation of a specific design element on a product’s packaging is not independent of the characteristics of such an element (e.g., contour and shape), as well as the characteristics of the other design elements.
An alternative interpretation for that discrepancy is potentially related to the specific product category that is used, as it can provide a semantic context for specific preferences.
In particular, wine is a special category in that several (hundred) broad product types are offered but most of them in the same bottle size (at 750 ml). While Westerman et al. (2013) did not find an interaction between product type and orientation on preferences, it is worth mentioning that they only used two product types and the stimuli used here are more realistic, and thus provide more information about product type.