Brain-lmaging Detection of Visual Scene Encoding in Long-term Memory for TV Commercials

CONSUMER CHOICES usually occur some time after exposure to advertisements; this is the case for mass media advertisements and also for direct-response advertisements, such as web advertisements, if the purchase response is not made instantaneously. It is possible that delayed choice may result from implicit brand awareness and brand attitude enhancement, without recall of the episodic advertising event or events that produced the enhancements (Rossiter and Percy, 1997), though this is not the view of the majority of advertisers. Most advertisers believe that explicit memory retention of traces of the advertisement is instrumental in influencing subsequent brand choice, and they therefore strive to create memorable advertisements. Visual recognition tests of key frames from TV campaigns, or illustrations from print campaigns, are an important measure in advertising tracking studies (Sutherland, 1993; Sutherland and Sylvester, 2000). Consumers' recognition of visual excerpts from advertisements is regarded as proof of media plan reach and as minimally sufficient evidence that the advertisements "worked" (Aaker and Bruzzone, 1981; Rossiter and Percy, 1997). We know, from Starch studies, quite a lot about what types of visuals are recognizable from print advertisements (Rossiter and Percy, 1997). However, little is known about what types of visuals are recognizable from TV commercials. What has been missing is the ability to predict which video scenes from a commercial will "cut through" and be memorable. Presumably, these scenes should identify the brand and depict the essential message content (Rossiter and Percy, 1997). A recent advancement in brain-imaging technology, called steady-state probe topography (SSPT), offers the fast temporal resolution needed to record cortical activity during exposure to a dynamic stimulus sequence, as presented by television programs and television commercials (Silberrstein et al., 1990; Silberstein, 1995; Silberstein et al., 2000). SSPT, which measures steady-state visually evoked potential (SSVEP), a brain electrical response elicited by visual stimuli, is essentially an innovative version of the well-known electroencephalographic (EEG) technology. An earlier study of EEG and memory for TV commercials (Rothschild and Hyun, 1990) is compared with ours in the Discussion. Other brain-imaging technologies, notably Positron Emission Tomography (PET) and functional Magnetic Resonance Imaging (fMRI), measure blood flow in the brain and do not provide fast enough recording and hence the temporal RICHARD B. SILBERSTEIN The Brain Sciences Institute Swinburne University of Technology