Implicit Learning Modulates Selective Attention at Sensory Levels of Perceptual Processing
Couperus, Jane W., Perception and Psychophysics
Electrophysiological evidence suggests that attention can be modulated as early as 100 msec after stimulus presentation. However, it is not clear whether these changes are based primarily on stimulus properties such as perceptual load (i.e., the level of perceptual difficulty), or other properties, such as general attentional set or learned expectations concerning perceptual load. Using event-related potentials, this study examined how implicit learning of perceptual load conditions modulates selective attention at sensory levels of perceptual analysis. The results show significant differences in P1 amplitude recorded over occipital areas of the brain as a function of learned expectations of perceptual load, only when perceptual load could be reliably predicted by the preceding stimuli. Moreover, differences in processing were found when both low and high perceptual load conditions could be predicted. These findings suggest that implicit learning modulates the allocation of attention at early stages of perceptual processing.
Imagine an infant entering the world, bombarded by information. How does he or she focus attention to select out important information so as not to be overwhelmed by the vast amount that is available? Fortunately, humans have a valuable tool to help guide attention: implicit learning. Implicit learning is defined as the ability to learn without conscious awareness. Although research suggests that attention is modulated by implicit learning (see, e.g., Jiang & Chun, 2001, 2003), it is not clear how early in processing this modulation occurs. Previous research suggests that attention can modulate processing of stimuli at sensory levels of analysis (for reviews, see Mangun, 1995, and Mangun & Hillyard, 1995). Thus, it seems reasonable for one to suggest that implicit learning may indirectly act at this early stage of perceptual processing by modulating selective attention, making it easier to function in a stimulus-rich environment.
There are multiple theories that attempt to explain the mechanisms of early attentional selection (see, e.g., Broadbent, 1958; Lavie, 1995; Treisman & Gelade, 1980); however, the specifics of attentional selection are still contested. For example, Lavie's perceptual load theory of selective attention (Lavie, 1995; Lavie, Hirst, de Fockert, & Viding, 2004) implies that early attentional selection is influenced by perceptual features and the type of perceptual analysis required. Support for this idea can be found in electrophysiological studies that have shown changes in selective attention as a function of perceptual load as early as sensory-level processing in the extrastriate cortex (Handy & Mangun, 2000; Handy, Soltani, & Mangun, 2001). For example, when presented with a difficult perceptual discrimination (i.e., high perceptual load), attention is more narrowly focused, reducing the processing of a parafoveal distractor (Handy et al., 2001). This description of the narrowing of attention can be understood as a spotlight or a zoom lens (Posner & Petersen, 1990), whereby attention is more diffuse under conditions of low perceptual load than of high perceptual load. Theeuwes, Kramer, and Belopolsky (2004) refined and extended this claim, suggesting that perceptual load alone is not sufficient for early attentional selectivity. They demonstrated that selective attention was most efficient during high perceptual load conditions in a block of high-load stimuli or if the current high-load stimulus was preceded by another high-load stimulus. Selection was not efficient when a high-load stimulus was preceded by a low-load stimulus. Thus, early attentional selection may be modified by other factors in addition to perceptual load, at least in conditions of high perceptual load.
Despite the aforementioned research, questions remain concerning the mechanisms of early attentional selection. Thus, it is important that we explore potential factors that may influence this process. …