Positional and temporal clustering in serial order memory.

Area of Science:

• Cognitive Psychology
• Neuroscience
• Memory Research

Background:

• Serial recall performance exhibits clustering around target item positions, supporting positional-coding theories.
• Existing findings on serial recall clustering are potentially confounded by temporal clustering effects.

Purpose of the Study:

• To investigate and disentangle positional and temporal clustering in serial recall.
• To evaluate the explanatory power of associative chaining and positional coding models for serial recall data.

Main Methods:

• Analyzing serial recall errors, conditioning on the positional accuracy of the previously recalled item to isolate clustering effects.
• Developing and testing a computational model of associative chaining with specific parameters (neighboring, remote associations, primacy gradient).

Main Results:

• Revised analyses, controlling for prior recall accuracy, demonstrate that temporal clustering is significantly more prevalent than positional clustering.
• The associative chaining model successfully accounts for the observed clustering patterns and serial position curves.
• A leading positional coding model fails to explain the nuanced clustering patterns identified in the study.

Conclusions:

• Temporal clustering plays a more dominant role in serial recall than previously recognized.
• Associative chaining models offer a more comprehensive explanation for serial recall phenomena than prominent positional coding models.