To err or to guess: an ERP study on the source of errors

Area of Science:

• Cognitive Neuroscience
• Human Factors Engineering
• Psychology

Background:

• Human error significantly contributes to complex system failures.
• Understanding error sources is crucial for system safety and performance.
• Response biases in speeded situations may lead to 'fast guess' errors.

Purpose of the Study:

• To investigate the sources of errors in both biased and unbiased conditions.
• To differentiate between fast guess errors and response selection errors.
• To explore the role of response selection processes in error generation.

Main Methods:

• Utilized a response-locked event-related potential (N-40) to detect response selection processes.
• Compared N-40 presence in biased versus unbiased experimental conditions.
• Employed a priming paradigm to induce response biases.

Main Results:

• The N-40 was detected in unbiased conditions, indicating response selection.
• The N-40 was absent in biased conditions, suggesting a lack of response selection.
• Errors in biased conditions are likely fast guesses, while errors in unbiased conditions arise from inappropriate response selections.

Conclusions:

• Different experimental conditions lead to distinct error sources.
• Fast guess errors dominate in biased situations, whereas inappropriate response selections characterize errors in unbiased situations.
• Event-related potential (N-40) provides physiological evidence for different error mechanisms.