The role of causal structure in implicit evaluation.

Area of Science:

• Cognitive Psychology
• Neuroscience
• Decision Making

Background:

• Humans leverage causal relationships to interact with their environment, influencing reward and punishment.
• Explicit evaluations are known to be sensitive to causality, but sensitivity of implicit evaluations remains less understood.

Purpose of the Study:

• To rigorously test if implicit (automatic) evaluations are sensitive to causal relationships.
• To investigate the underlying representational mechanisms of this sensitivity.

Main Methods:

• Large-scale study (N=4836) across 6 experiments with varying designs and verbal support.
• Participants observed causal events where one stimulus, not another, was causally responsible for outcomes.
• Evaluated explicit (Likert, slider scales) and implicit (Implicit Association Tests) measures.

Main Results:

• Causal status consistently influenced both explicit and implicit evaluations (large Bayes Factors).
• Implicit evaluations were sensitive to direct causal information but not to flexibly derived causal relationships.
• Effect sizes: explicit (Cohen's d=0.28), implicit (Cohen's d=0.22).

Conclusions:

• Implicit evaluations are indeed sensitive to causal information, extending beyond mere stimulus co-occurrence.
• This sensitivity is likely mediated by precompiled, causally informed value representations.
• Implicit evaluation does not rely on online computations over a flexible causal model.