Longitudinal evidence for decreasing statistical learning abilities across childhood

What the study found

Statistical learning declined between ages 7 and 14 in the same participants. Higher executive functions were associated with the decline. This confirms longitudinally what cross-sectional work had suggested: implicit pattern learning is strongest in early childhood.

Limitations

One cohort and one laboratory task; the stimuli are abstract sequences, not rooms or routines. Sample size and attrition not extracted from full text.

Neurodesign interpretation

A 7-year-old picks up the patterns of a classroom faster than a 14-year-old does, and without trying. The study never tested space, so the link to room design is our inference. It does give a developmental reason why consistency in early-years environments could matter more than it does later.

Possible design implications

Younger children are likely to absorb the regularities of a setting (layout, routine, where things live) with less effort than older children. Consistent spatial organisation in early-years settings is the hypothesis to test.

Why it stays in the corpus

Retain as the strongest developmental finding in the September 2026 intake, because it follows the same children rather than comparing age groups.

Study design

Longitudinal repeated-measures study with linear mixed models and latent class analysis

Population

The same children tested at ages 7, 8, 11 and 14

Setting

Laboratory sequence-learning task

Field

Developmental cognitive neuroscience

Mechanism

Implicit statistical learning, executive functions

Spatial variable

Environmental regularity (inferred, not manipulated)

Keywords

statistical learning, implicit learning, child development, executive function, longitudinal