Simple dissociations for a higher-powered neuropsychology

Area of Science:

• Cognitive Neuropsychology
• Psychometrics

Background:

• Dissociations are key to understanding cognitive neuropsychology, often studied via single-case analysis.
• Current criteria for 'classical' and 'strong' dissociations require frank deficits on one or both tasks, plus significant between-task differences.

Purpose of the Study:

• To propose a simplified, logically necessary criterion for identifying dissociations.
• To evaluate the statistical power and error control of the proposed criterion.
• To reassess the utility of classifying dissociations as 'classical' or 'strong'.

Main Methods:

• Monte Carlo simulations were employed to test the proposed dissociation criterion.
• The simulations assessed the power to detect dissociations and control Type I errors under various conditions.
• Statistical significance of between-task differences was the primary focus.

Main Results:

• The simplified criterion (significant between-task difference) increases statistical power for dissociation detection.
• The simplified criterion maintains excellent control over Type I errors.
• Classifying dissociations as 'classical' or 'strong' based on frank deficits may be unreliable for theoretical inference.

Conclusions:

• A significant between-task difference is the logically necessary and sufficient condition for identifying a dissociation.
• The strength of a dissociation can be more reliably characterized using a continuous index, like effect size.
• Relying solely on frank deficits for categorizing dissociations may introduce uncertainty into theoretical conclusions.