An explorative multiverse study for extracting differences in P3 latencies between young and old adults

Area of Science:

• Neuroscience
• Cognitive Psychology
• Psychophysiology

Background:

• Age-related increases in P3 latencies are well-documented.
• Investigating these differences involves numerous methodological choices, leading to varied analytical pipelines.
• The impact of these pipeline variations on age effects in P3 latencies is not fully understood.

Purpose of the Study:

• To investigate the effects of different analytical pipelines on age-related differences in P3 latencies.
• To explore the influence of varying methodological decisions within EEG data analysis.
• To identify optimal analytical strategies for detecting and characterizing age effects on P3 latencies.

Main Methods:

• An explorative multiverse study design was employed.
• Varied parameters included low-pass filter (4-32 Hz, none), latency type (area vs. peak), analysis level (single participant vs. jackknifing), and extraction method (manual vs. automated).
• EEG data from 30 young and 30 older adults performing cognitive tasks were analyzed.

Main Results:

• Analysis strategies profoundly influenced the detection and magnitude of age effects, explaining 0% to 88% of variance.
• Reliabilities ranged from .20 to 1.00, and homogeneities from -.12 to .90.
• Manual extraction, particularly on single-participant data, generally yielded the most effective pipelines.

Conclusions:

• Different analytical pipelines can drastically alter findings regarding age-related P3 latency differences.
• Manual extraction of peak latencies is effective across most filters, while area latencies benefit from filters above 8 Hz.
• Jackknifing with peak latencies may produce inconclusive results, highlighting the importance of methodological choices in aging research.