| Chapter 2: | Literature Review |
Effects of Ad Variation
Past research on ad variation has focused on two theoretical propositions in explaining why varied ads are more effective than non-varied ads: namely, the encoding variability and the differential attention hypotheses. The encoding variability hypothesis holds that when people are exposed to a stimulus more than once, memory performance can be improved if the stimulus is presented differently (Glenberg, 1979; Young & Bellezza, 1982; Burnkrant, & Unnava, 1987, 1991). In other words, memory performance will be better if stimuli are presented and processed in different contexts since different contexts can increase the number of possible retrieval cues. If people are exposed to two different ads under the same brand name, these ads will leave two different memory traces instead of one, and the information stored may provide more contextual cues for later retrieval. Burnkrant and Unnava (1987) argue that “the likelihood of retrieving the stimulus is believed to be directly related to the number of traces involving that stimulus” (p. 173). Thus, the encoding variability hypothesis predicts that the probability of recall is directly related to the number of cues relevant to the target information (Glenberg, 1979). Therefore, the encoding variability hypothesis suggests superior memory in varied ad conditions than in non-varied ad conditions.
In the advertising literature, several studies have shown evidence for the encoding variability explanation (Unnava & Burnkrant, 1991; Burnkrant & Unnava, 1987; Singh, Linville, & Sukhdial, 1995). For example, Burnkrant and Unnava (1987) exposed participants to a series of three whisky ads that were either identical or varied. They found significant effects for ad variation on recall but found no significant differences in self-reported attention. Thus, the researchers concluded that the significant effects of variation were due to encoding variability rather than differential attention.
A follow-up study by Unnava and Burnkrant (1991) produced supporting evidence of encoding variability for ad variation. These researchers