Blocking in human causal learning is affected by outcome assumptions manipulated through causal structure

Area of Science:

• Cognitive Psychology
• Human Causal Learning

Background:

• Blocking is a phenomenon in associative learning where prior learning inhibits new causal associations.
• Additivity assumptions, typically manipulated via pretraining or instructions, are known to influence blocking.
• Existing research often relies on pretraining or explicit instructions to alter these assumptions.

Purpose of the Study:

• To investigate whether manipulating the causal structure, rather than pretraining or instructions, affects human causal learning.
• To examine how prior knowledge about outcome additivity influences the blocking effect.
• To test if assuming different causal systems (additive outcomes) versus the same system (ceiling effect) impacts blocking.

Main Methods:

• Two experiments were conducted without pretraining or explicit instructions.
• Participants were presented with cues embedded in either a 'different-system' (assumed additivity) or 'same-system' (ceiling effect) causal structure.
• Blocking was assessed by observing participants' causal judgments when compound cues were presented.

Main Results:

• Experiment 1: Blocking was observed when cues belonged to separate causal systems (expected additivity).
• Experiment 1: Blocking was not observed when cues belonged to the same causal system (ceiling effect prevented additivity).
• Experiment 2: Partially replicated findings, showing the pattern when cues were tested a second time.

Conclusions:

• Prior knowledge about the nature of causal relations significantly affects human causal learning.
• The findings challenge associative theories of learning that do not easily account for results obtained without pretraining.
• Causal structure manipulation offers a novel way to probe the influence of prior knowledge on learning.