Investigating lexical categorization in reading based on joint diagnostic and training approaches for language...

Area of Science:

• Neuroscience
• Cognitive Psychology
• Educational Technology

Background:

• Efficient reading is crucial for societal participation and a primary educational objective.
• Visual word recognition involves complex cognitive processes, including lexical categorization, which is vital for reading proficiency.
• The Lexical Categorization Model (LCM) describes pre-lexical orthographic processing in the brain's left-ventral occipital cortex.

Purpose of the Study:

• To investigate individualized diagnostics and training for visual word recognition processes.
• To evaluate the effectiveness of a training framework based on the Lexical Categorization Model (LCM) for improving reading skills.
• To utilize machine learning for predicting individual responses to reading training.

Main Methods:

• Developed an individualized diagnostics and training framework motivated by the LCM.
• German language learners underwent training in lexical categorization while reading speed was monitored.
• Estimated LCM-based features and assessed individual lexical categorization capabilities.
• Employed machine learning to identify optimal feature selection and regression models for predicting training benefits.

Main Results:

• Most language learners demonstrated increased reading skills after training.
• The best machine learning pipeline boosted reading speed from 23% in the unselected group to 43% in the machine-selected group.
• The predictive model's success was significantly dependent on LCM-associated parameters.

Conclusions:

• Training lexical categorization skills can enhance reading proficiency.
• Combining computational descriptions of brain functions with machine learning offers a powerful approach for individualized reading training.
• This framework effectively identifies individuals who will benefit most from targeted reading interventions.