Increasing previous but not concurrent extinction attenuates the "extinction makes acquisition context specific"...

Area of Science:

• Cognitive Psychology
• Human Predictive Learning

Background:

• The extinction makes the acquisition context specific (EMACS) effect describes how learning can become dependent on the context in which it occurs.
• Understanding factors that attenuate the EMACS effect is crucial for explaining context processing in predictive learning.

Purpose of the Study:

• To investigate whether the EMACS effect is reduced when the prediction error associated with extinction diminishes.
• To examine the influence of extinction timing and duration on context specificity in human predictive learning.

Main Methods:

• Four experiments utilized a predictive learning task where participants associated food cues with gastric illness in specific restaurant contexts.
• Participants underwent extinction training for a cue (E) either concurrently or before acquiring a new target cue (P).
• Performance on cue P was tested across two contexts (A and B) to assess context dependency.

Main Results:

• Concurrent extinction training led to context-dependent performance (EMACS effect) irrespective of the number of extinction trials (12 or 24).
• The EMACS effect was eliminated when extinction training (24 trials) preceded the acquisition of the target cue P.
• This suggests that the timing and duration of extinction influence the development of context-specific learning.

Conclusions:

• The findings indicate that the EMACS effect is not solely dependent on the disappearance of prediction error from extinction.
• Extinction conducted prior to acquisition, with sufficient trials, can prevent the emergence of context-specific learning.
• Results offer insights into attentional explanations for context processing in associative learning.