Reinforcement learning in children with FASD.

Area of Science:

• Neuroscience
• Developmental Psychology
• Child Psychiatry

Background:

• Children with Fetal Alcohol Spectrum Disorder (FASD) are anecdotally reported to struggle with reinforcement learning.
• Empirical evidence supporting these claims is limited, necessitating further investigation.

Purpose of the Study:

• To investigate reinforcement learning rates in children with FASD.
• To assess the influence of reinforcer concreteness on learning in children with FASD.

Main Methods:

• A reinforcement learning task with probabilistic feedback was administered to 18 children with FASD (IQ ≥ 70) and 18 controls (ages 11-17).
• The task was completed twice, using either tokens or points as reinforcers, in a counterbalanced order.

Main Results:

• Controls exhibited significantly stronger overall reinforcement learning compared to the FASD group.
• The FASD group's learning was more influenced by the most recent trial's outcome, unlike controls who integrated information over multiple trials.
• No significant differences were found between groups regarding the rate of improvement or the impact of reinforcer type (tokens vs. points).

Conclusions:

• Reinforcement learning in children with FASD appears to be slower but not functionally different from controls.
• Children with FASD may rely more on immediate feedback rather than integrating reinforcement history.
• Sufficient repetition may enable learning in children with FASD, though other factors like complexity and impulsivity require further study.