In a study that is in some ways even more compelling, we demonstrated a double dissociation between explicit and implicit measures, in which performance on explicit and implicit tests responded differentially, and in predictable ways, to experimental manipulations (Law & Braun, 2000). In this study, we examined the effectiveness of product placement by having participants watch one of two excerpts of Seinfeld (TV program); containing at least six product placements under the guise of collecting their evaluations of the show.
A subset of the placements was central to the plot (e.g., a discussion between Kramer and Elaine’s mother about a chocolate ´eclair), whereas others were more peripheral (such as a box of Tide in a garbage pail). Furthermore, a few of the products were both seen and mentioned, whereas others were either simply seen or only spoken about. Thus, plot centrality and placement modality were the two independent variables in the study. After viewing the Seinfeld excerpt, participants completed an implicit choice task and an explicit recognition test.
In the implicit choice task, participants were asked to choose a brand from a set of two brand names (where one was a brand present in the Seinfeld episode and the other was not present). No reference to the earlier viewing was made. The construction of the explicit recognition test was similar, consisting of previously seen and new brand names, except that in this case participants were instructed to think back to the viewing episode and identify which brands had been present in the video.
We expected placement centrality and placement modality would dissociate between explicit and implicit measures of memory. The results confirmed these expectations: placements that were central to the program were best recalled and recognized though least likely to be chosen. In contrast, placement modality showed an opposite effect on the two measures: seen-only placements showed lower recall and recognition compared with heard only placements but were chosen most frequently.
In other words, the experimental manipulations produced strong double dissociations: Although centrality affected explicit measures and had no impact on the implicit measure, placement modality showed the reverse effect. One limitation of the Law and Braun (2000) study is that, like other placement research, it relied on existing footage to investigate product placement effects (see Russell, 2002, for an exception).