Coding responses and the generalization of matching to sample in children

Area of Science:

• Behavioral psychology
• Cognitive development
• Learning sciences

Background:

• Generalization is a key aspect of learning, allowing individuals to apply learned information to new situations.
• Relational matching to sample tasks assess the ability to identify relationships between stimuli.
• Understanding the conditions for stimulus control is vital for explaining generalization.

Purpose of the Study:

• To investigate the specific conditions of stimulus control required for the generalization of relational matching to sample.
• To examine how the stability of component responses influences generalization performance.
• To identify factors that facilitate or hinder the transfer of matching skills to novel stimuli.

Main Methods:

• Two experiments were conducted involving participants learning a matching to sample task with a 90-degree rotation.
• Experiment 1 trained participants on coding sample orientations and transforming these codes.
• Experiment 2 involved retraining participants to modify stimulus control over coding responses.

Main Results:

• High generalization occurred when stable sample-coding responses were initially available.
• Low generalization was observed when stable sample-coding responses were not initially available, but improved with training.
• Modifying stimulus control over coding responses led to immediate improvements in generalized matching.

Conclusions:

• The generalization of matching to sample is significantly dependent on the structure of stimulus control.
• Stable and appropriately controlled component responses are necessary for effective generalization.
• Interventions targeting the modification of stimulus control can enhance generalized learning.