The prevalence and importance of statistical learning in human cognition and behavior

https://doi.org/10.1016/j.cobeha.2020.01.015Get rights and content

Highlights

  • Infants use statistical learning to acquire environmental structure and regularities.

  • Statistical learning operates over multiple modalities and levels of abstraction.

  • Several aspects of human behavior and cognition rely upon or benefit from statistical learning.

  • An emerging neural understanding of statistical learning is informing behavioral hypotheses.

Statistical learning, the ability to extract regularities from the environment over time, has become a topic of burgeoning interest. Its influence on behavior, spanning infancy to adulthood, has been demonstrated across a range of tasks, both those labeled as tests of statistical learning and those from other learning domains that predated statistical learning research or that are not typically considered in the context of that literature. Given this pervasive role in human cognition, statistical learning has the potential to reconcile seemingly distinct learning phenomena and may be an under-appreciated but important contributor to a wide range of human behaviors that are studied as unrelated processes, such as episodic memory and spatial navigation.

Introduction

Although each day brings new experiences, our world does not present us with a series of novelties. Rather, our experience is highly repetitive and structured. Over the past two decades, a subfield of cognitive science has emerged on how humans acquire this information about the world via statistical learning. This research has highlighted that infants, children, adults  and in some cases non-human animals  possess the remarkable ability to detect and represent regularities from the environment in an unsupervised fashion, often without awareness. In this review, we first highlight recent findings demonstrating not only that humans have the capacity for statistical learning, but also that these learned regularities are relevant for behavior throughout the lifespan  from acquiring language to forming predictions about upcoming experiences. We then propose that these mechanisms have behavioral consequences, from facilitating cognitive processing, to shaping representations, to enabling integration over past experiences. Finally, we end by motivating future investigations of statistical learning based on an emerging understanding of its neural foundations, focusing on its reliance on the hippocampus, a brain structure conventionally implicated in episodic memory and spatial navigation.

Section snippets

Mechanisms of statistical learning

Statistical learning is a rapid, efficient means of extracting regularities from the environment. To this end, it has often been studied in the context of development, a period when it is particularly adaptive to quickly learn about the world. However, statistical learning continues to operate and play an important role in cognition throughout the lifespan in adults. Here we review these two bodies of research on statistical learning.

Behavioral consequences of statistical learning

The adaptive purpose of statistical learning in infancy is readily apparent  for learning about the structure of an unknown world. What are the consequences and benefits for adults? One possibility is that adults are robust statistical learners as a vestige of its importance in development. Alternatively, statistical learning may continue to play an important functional role throughout adulthood. Here, we highlight some of these adaptive benefits.

Behavioral implications of statistical learning in the brain

Exploring how statistical learning influences and interacts with other cognitive systems, such as attention, memory, and decision-making, helps to reveal its broad and adaptive role in behavior. However, these studies employ a wide variety of tasks and stimuli, raising the possibility that there are multiple forms of statistical learning. Here we ask whether our understanding of how statistical learning operates in the brain can be used to make novel behavioral predictions and better

Conclusions and future directions

Throughout this paper, we have explored the pervasive role of statistical learning in human cognition and behavior. We ended with the suggestion, based on a theory of the hippocampus, that one such role of statistical learning may be to influence how and when episodic memories are formed. This approach of generating novel behavioral predictions from an improved neural understanding holds additional promise because statistical learning has been linked to several brain regions beyond the

Conflict of interest statement

Nothing declared.

References and recommended reading

Papers of particular interest, published within the period of review, have been highlighted as:

  • • of special interest

  • •• of outstanding interest

Acknowledgements

This work was supported by funding from the National Institutes of Health (R01 MH069456) to NTB and the National Science Foundation (GRFP) to BES.

References (64)

  • R.Q. Yu et al.

    Object representations are biased toward each other through statistical learning

    Visual Cogn

    (2018)
  • R.Q. Yu et al.

    Implicit updating of object representation via temporal associations

    Cognition

    (2018)
  • G.E. Wimmer et al.

    Preference by association: how memory mechanisms in the hippocampus bias decisions

    Science

    (2012)
  • T.F. Brady et al.

    Compression in visual working memory: using statistical regularities to form more efficient memory representations

    J Exp Psychol: Gen

    (2009)
  • I. Momennejad et al.

    The successor representation in human reinforcement learning

    Nat Hum Behav

    (2017)
  • I.K. Brunec et al.

    Boundaries shape cognitive representations of spaces and events

    Trends Cogn Sci

    (2018)
  • N.V. Covington et al.

    The necessity of the hippocampus for statistical learning

    J Cogn Neurosci

    (2018)
  • A.C. Schapiro et al.

    Complementary learning systems within the hippocampus: a neural network modelling approach to reconciling episodic memory with statistical learning

    Philos Trans R Soc B: Biol Sci

    (2017)
  • A.J. Chanales et al.

    Overlap among spatial memories triggers repulsion of hippocampal representations

    Curr Biol

    (2017)
  • J.R. Saffran et al.

    Infant statistical learning

    Annu Rev Psychol

    (2018)
  • J. Arciuli et al.

    The promise  and challenge  of statistical learning for elucidating atypical language development

    Curr Direct Psychol Sci

    (2018)
  • E.M. Clerkin et al.

    Real-world visual statistics and infants’ first-learned object names

    Philos Trans R Soc B: Biol Sci

    (2017)
  • C.D. Monroy et al.

    The infant motor system predicts actions based on visual statistical learning

    NeuroImage

    (2019)
  • S.P. Johnson et al.

    Development of object concepts in infancy: evidence for early learning in an eye-tracking paradigm

    Proc Natl Acad Sci U S A

    (2003)
  • C.T. Ellis et al.

    Infant fMRI: a model system for cognitive neuroscience

    Trends Cogn Sci

    (2018)
  • S.D. Palmer et al.

    Statistical learning for speech segmentation: age-related changes and underlying mechanisms

    Psychol Aging

    (2018)
  • A. Shufaniya et al.

    Statistical learning is not age-invariant during childhood: performance improves with age across modality

    Cogn Sci

    (2018)
  • Y. Leung et al.

    Learning unfamiliar pitch intervals: a novel paradigm for demonstrating the learning of statistical associations between musical pitches

    PLOS ONE

    (2018)
  • R. Dotsch et al.

    Statistical learning shapes face evaluation

    Nat Hum Behav

    (2017)
  • Y. Luo et al.

    Statistical learning creates novel object associations via transitive relations

    Psychol Sci

    (2018)
  • G. Lengyel et al.

    Unimodal statistical learning produces multimodal object-like representations

    eLife

    (2019)
  • J. Jun et al.

    Visual statistical learning at basic and subordinate category levels in real-world images

    Attent Percept Psychophys

    (2018)
  • Cited by (48)

    • Weight status impacts children's incidental statistical learning

      2023, International Journal of Psychophysiology
    View all citing articles on Scopus
    View full text