Statistical learning in reading: variability in irrelevant letters helps children learn phonics skills

Area of Science:

• Cognitive psychology
• Developmental psychology
• Educational psychology

Background:

• Early reading acquisition relies on statistical learning of grapheme-phoneme correspondences (GPCs).
• Classroom-based evidence on how statistical learning principles impact reading acquisition is limited.
• Understanding optimal statistical learning strategies is crucial for improving reading instruction.

Purpose of the Study:

• To investigate two competing principles of statistical learning for vowel GPCs in a classroom setting.
• To determine whether variability or similarity in word stimuli better supports learning and generalization of GPCs.
• To examine the impact of these principles on first-grade students' reading abilities.

Main Methods:

• Trained 224 first-grade students on vowel GPCs using words with high or low consonant similarity.
• Assessed generalization of learned GPCs to novel words and tasks.
• Compared the effectiveness of learning under variability versus similarity conditions.

Main Results:

• Variability in learning materials consistently benefited generalization of GPCs.
• This benefit was observed for both trained and novel words and tasks.
• Similarity between words did not enhance, and potentially hindered, GPC learning and generalization.

Conclusions:

• Variability in learning materials is a more effective principle than similarity for teaching vowel GPCs in early reading.
• Classroom interventions should consider leveraging variability to enhance children's statistical learning of reading skills.
• Findings suggest practical implications for designing effective reading instruction.