The Automaticity of Visual Statistical Learning.

Abstract

Recent studies of visual statistical learning (VSL) have demonstrated that statistical regularities in sequences of visual stimuli can be automatically extracted, even without intent or awareness. Despite much work on this topic, however, several fundamental questions remain about the nature of VSL. In particular, previous experiments have not explored the underlying units over which VSL operates. In a sequence of colored shapes, for example, does VSL operate over each feature dimension independently, or over multidimensional objects in which color and shape are bound together? The studies reported here demonstrate that VSL can be both object-based and feature-based, in systematic ways based on how different feature dimensions covary. For example, when each shape covaried perfectly with a particular color, VSL was object-based: Observers expressed robust VSL for colored-shape sub-sequences at test but failed when the test items consisted of monochromatic shapes or color patches. When shape and color pairs were partially decoupled during learning, however, VSL operated over features: Observers expressed robust VSL when the feature dimensions were tested separately. These results suggest that VSL is object-based, but that sensitivity to feature correlations in multidimensional sequences (possibly another form of VSL) may in turn help define what counts as an object.

The impact of attention load on the use of statistical information and coarticulation as speech segmentation cues

“…Indeed, both segmentation cues are available throughout linguistic development (i.e., since an early age-e.g., Johnson & Jusczyk, 2001;Kirkham, Slemmer & Johnson, 2002;-and in adulthood-e.g., Mattys, 2004, but only coarticulation is speech-specific, and hence may be processed by an encapsulated system that may operate quite independently of attention resources. On the contrary, TP computation relies on a domain-general learning mechanism (e.g., Fiser & Aslin, 2001;Perruchet & Pacton, 2006;Turk-Browne et al, 2005) that may be more sensitive to attention resources and hence to cognitive load. A recent study by Astheimer and Sanders (2009) also suggests that domain specificity could play a role in the allocation of selective attention to word onsets.…”

“…They were informed that the text they would read later on had no direct relation to the auditory and visual stimuli presented in the first phase. These instructions aimed at forcing participants to allocate their attention to both the visual and auditory streams, in a nonetheless incidental passive learning situation (see Turk-Browne, Jungé, & Scholl, 2005).…”

An attention-based associative account of adjacent and nonadjacent dependency learning.

“…Those studies led to the emergence of models centered on the need for learners' attentional involvement (e.g., Mackintosh, 1975). Since then, empirical data suggesting that learning does not occur without a minimal level of attention have accumulated, especially in the area of implicit, nonintentional forms of learning, which has 1 Turk-Browne et al (2005) reported successful learning of relationships between visual shapes of a given color when the related items are interrupted by irrelevant items from another color stream. Turk-Browne et al's study was not directly aimed at studying nonadjacent dependencies, and as discussed by the authors, the results are only suggestive of the possibility of what they called "interrupted statistical learning."…”

Functional differences between statistical learning with and without explicit training

“…Statistical learning has been demonstrated both when stimuli are presented passively without any explicit task (e.g., Saffran et al 1999;Aslin 2001, 2002;Toro et al 2005), and when participants are engaged in a cover task unrelated to the underlying pattern (Saffran et al 1997, Turk-Browne et al 2005). In addition, statistical learning seems to be unaffected by the precise instructions given to participants (Arcuili et al 2014;Batterink et al 2015).…”