TY - UNPB

T1 - Addressing the Theory Crisis in Statistical Learning Research

AU - Conway, Christopher M.

AU - Jenkins, Holly

AU - Milne, Alice E.

AU - Singh, Sonia

AU - Wilson, Benjamin

PY - 2025/3/18

Y1 - 2025/3/18

N2 - Research into the construct of statistical learning, the ability to learn structured patterns in the environment, has been exploding in recent years. Despite the plethora of studies, critical reviews have highlighted challenges facing this area of research, including confusion about what statistical learning is, what cognitive and neural mechanisms support it, how these mechanisms map onto different laboratory tasks, and how it relates to other forms of learning, memory, and cognition. These challenges are not unique to statistical learning research but in fact are difficulties confronting the psychological sciences more broadly. In this paper, we use recent conceptual advances regarding the nature of the “theory crisis” in psychology to provide a roadmap for developing a theoretical framework that can provide meaningful and valid conceptualizations of the various phenomena that statistical learning touches upon. First, we describe three important challenges to statistical learning research: a lack of robust phenomena to constrain theories, issues with construct validity, and challenges with establishing causality. Next, we discuss each issue in relation to several research phenomena that have been examined in some detail in the statistical learning research. Finally, we offer recommendations that we believe can help address the theory crisis in statistical learning research to help move the field forward in fruitful directions.

AB - Research into the construct of statistical learning, the ability to learn structured patterns in the environment, has been exploding in recent years. Despite the plethora of studies, critical reviews have highlighted challenges facing this area of research, including confusion about what statistical learning is, what cognitive and neural mechanisms support it, how these mechanisms map onto different laboratory tasks, and how it relates to other forms of learning, memory, and cognition. These challenges are not unique to statistical learning research but in fact are difficulties confronting the psychological sciences more broadly. In this paper, we use recent conceptual advances regarding the nature of the “theory crisis” in psychology to provide a roadmap for developing a theoretical framework that can provide meaningful and valid conceptualizations of the various phenomena that statistical learning touches upon. First, we describe three important challenges to statistical learning research: a lack of robust phenomena to constrain theories, issues with construct validity, and challenges with establishing causality. Next, we discuss each issue in relation to several research phenomena that have been examined in some detail in the statistical learning research. Finally, we offer recommendations that we believe can help address the theory crisis in statistical learning research to help move the field forward in fruitful directions.

U2 - 10.31234/osf.io/t5f3h_v1

DO - 10.31234/osf.io/t5f3h_v1

M3 - Preprint

T3 - PsyArXiv

BT - Addressing the Theory Crisis in Statistical Learning Research

PB - PsyArXiv

ER -