Automaticity and cognitive control in the learned predictiveness effect.

Area of Science:

• Cognitive Psychology
• Learning and Memory
• Decision Making

Background:

• Learning is often biased by prior experiences, favoring cues previously found to be predictive.
• Learned predictiveness describes this bias, where past predictive cues are prioritized in new learning contexts.
• Recent research indicates that explicit instructions can modulate learned predictiveness, suggesting volitional control.

Purpose of the Study:

• To investigate the influence of instructional manipulations on learned predictiveness.
• To explore the interplay between explicit instructions and prior cue predictiveness in learning.
• To differentiate between inferential and automatic contributions to learned predictiveness.

Main Methods:

• Three experiments were conducted involving instructional manipulations of cue causal status.
• Experiment 1 examined how instructions affect associative memory and causal reasoning.
• Experiments 2 and 3 used a procedure to disentangle inferential and automatic learning biases by manipulating prior cue predictiveness alongside instructions.

Main Results:

• Instructions significantly influence learned predictiveness, affecting both causal reasoning and associative memory.
• Prior cue predictiveness independently impacts subsequent learning, even when explicit instructions are present.
• The effects of explicit instruction and prior predictive history on learning are interactive, not simply additive.

Conclusions:

• Both explicit instructions and past predictive experience shape learning biases.
• Learned predictiveness is influenced by a complex interaction between top-down (instructional) and bottom-up (experience-based) factors.
• Understanding this interaction is crucial for explaining how prior knowledge and explicit guidance jointly guide new learning.