Another important issue is the question of what factors induce individuals to process particular features in either a shallow or a deeper fashion. In both experiments 1 and 2, it appeared that the amount of semantic content possessed by a particular feature tended to determine which type of processing was employed. Hence, it may be reasonable to assess features based on their semantic content and classify them as either likely to be processed in a shallow or deeper fashion. At the same time, particular conditions under which an individual processes a specific feature are also likely to influence the type of processing he or she employs. In particular, if an individual's motivation or ability to process a particular feature is limited, he or she may process it consistently in a perceptual manner, regardless of that feature's semantic content. On the other hand, with sufficient repetition and consistent associations, any feature can become invested with semantic content over time. Therefore, from a managerial perspective, knowledge about the amount of semantic content a specific feature possesses, as well as the particular conditions under which this feature has been processed, will presumably enhance marketers' ability to anticipate the onset of wearout of repeatedly viewed stimuli.
Finally, although these results as well as those of other studies involving extremely short exposure durations (e.g., Bornstein & D'Agostino, 1992) suggest that when features are processed in a shallow fashion, no downturn in affective response occurs even at high exposure levels, this question merits further consideration. Whether or not a downturn in affective response will eventually occur under perceptual processing conditions when the exposure level is extremely high remains an empirical question.