Basic Information Processing Effects from Perceptual Learning in Complex, Real-World Domains Khanh-Phuong Thai (kpthai@ucla.edu) Everett Mettler (mettler@ucla.edu) Philip J. Kellman (kellman@psych.ucla.edu) Department of Psychology, 405 Hilgard Avenue, Los Angeles, CA 90095 USA Abstract mathematics and science (e.g., Goldstone, Landy & Son, 2008; Kellman et al., 2008). In recent research, Kellman and colleagues have shown that PL can be systematically engineered and accelerated using appropriate computer-based technology (e.g., Kellman, Massey & Son, 2010). Their approach to PL methods takes the form of perceptual learning modules (PLMs). Rather than focusing on memorization of instances, PLMs employ unique instances and systematic variations in the learning set to promote the learning of invariant or diagnostic structures characterizing a category or concept. Learners engage in short, interactive episodes focused on discrimination or classification. Because specific instances seldom or never repeat in PLMs, learners pick up structural invariance and can generalize it to new instances (Kellman et al., 2010). Recent work suggests that relatively brief PLM interventions can produce dramatic learning gains in challenging mathematical domains, such as fraction learning and algebra problem solving (Kellman et al., 2008; Kellman et al., 2010). Recent research indicates that perceptual learning (PL) interventions in real-world domains (i.e., mathematics, science) can produce strong learning gains, transfer, and fluency. Although results on domain-relevant assessments suggest characteristic PL effects, seldom have real-world PL interventions been explicitly tested for their effects on basic information extraction. We trained participants to classify Chinese characters, based on either (1) overall configurations (structures), (2) featural relations (components), or (3) non- relational information (stroke-count control). Before and after training, we tested for changes in information extraction using a visual search task. Search displays contained all novel exemplars, involved manipulations of target-distractor similarity using structures and components, and included heterogeneous and homogeneous distractors. We found robust improvements in visual search for structure and component PL training relative to the control. High-level PL interventions produce changes in basic information extraction, and sensitivity induced by PL for both relational structure and specific components transfers to novel structural categories. Purpose of Current Work Keywords: perceptual learning; educational technology; visual search; categorization. In applying PL to complex, symbolic, real-world learning domains, a critical question arises - how do we tell that the driver in these effects is really PL? Kellman, Massey & Son (2010) set out characteristic design features of perceptual learning interventions and some signature effects that implicate PL. Yet, realistic learning domains are complex and involve synergies between conceptual knowledge and perception of structure. Here we sought evidence of PL effects in a high-level, realistic learning domain, by explicitly testing after PLM use for basic changes in information extraction. We trained PL for complex patterns in Chinese characters using a paradigm similar to that used to train PL in math and science learning (Kellman et al., 2010). Since Chinese characters are logographic and have both local and global structure, we were able to train participants to recognize characters at 3 different levels of hierarchical organization: stroke, component, and structure. Participants in two PL conditions matched characters by component (featural relations) or overall structure (global configuration). Importantly, in the case of matching by structure, local components were free to vary. Other studies have shown that an expert’s ability to use relevant ‘chunks’ based on components and configural structure has to be nourished by literacy development and cannot be obtained solely through Introduction Research on expertise has shown that experts effortlessly attend to relevant features and relations (Gibson, 1969), that experts extract larger “chunks” of information, discover higher-order invariance, and do so with low attentional load (Gibson, 1969; Schneider & Shiffrin, 1977). Such changes in information extraction as a result of experience constitute perceptual learning (Gibson, 1969; for a recent review, see Kellman & Garrigan, 2009). Much contemporary research on perceptual learning (PL) has focused on basic sensory discriminations; however, PL effects are not confined to low-level tasks (Garrigan & Kellman, 2008; Kellman & Garrigan, 2009). In fact, the natural function of PL is to improve the extraction of information from complex objects and events (Kellman & Garrigan, 2009). PL also likely involves discovery of abstract relational structures. Such high-level PL is a crucial component of expertise in many domains including reading (Baron, 1978; Yeh et al., 2003), chess (Chase & Simon, 1973), and X-ray interpretation (Chi, Feltovich & Glaser, 1981). In addition, recent research indicates an important role for PL in high-level symbolic domains, such as
Read full abstract