What question did this study set out to answer?

To examine how routine methods for quantifying and analyzing event-related potential datasets produce false-positive results and to establish best practices for reliable inference.

December 20, 2016PsychophysiologyOpen Access

How to get statistically significant effects in any ERP experiment (and why you shouldn't)

View Full Paper

Ask AI

Bookmark

Authors

SLSteven J. LuckUniversity of California, Riverside NGNicholas GaspelinUniversity of Missouri

Discussion

Member takes

Implication

Simulation study reveals false-positive rates exceeding 50% in event-related potential analyses, highlighting the need for robust analytical strategies.

Key Points

To examine how routine methods for quantifying and analyzing event-related potential datasets produce false-positive results and to establish best practices for reliable inference.
Simulated typical event-related potential (ERP) experimental designs generating multi-thousand data point profiles per participant.
Reanalyzed empirical ERP datasets to evaluate component amplitude and latency quantification methods.
Tested Type I error inflation arising from selecting time windows and electrode sites on grand-averaged data alongside multifactor statistical models.
Common data-contingent quantification practices yielded at least one statistically significant bogus effect in more than 50% of typical experimental setups.
Selecting time windows and electrode sites post-hoc using grand-averaged data substantially increased false-positive discoveries.
Combining flexible component selection with multifactor statistical designs compounded error rates, underscoring the necessity of preplanned, hypothesis-driven analytical choices.

Cite This Study

Luck et al. (2016) studied this question.

synapsesocial.com/papers/699bddaa9762345f166ddeb0 https://doi.org/10.1111/psyp.12639

View Full Paper

Ask AI

Bookmark

Implication

Simulation study reveals false-positive rates exceeding 50% in event-related potential analyses, highlighting the need for robust analytical strategies.

How to get statistically significant effects in any ERP experiment (and why you shouldn't)

Discussion

Member takes

Implication

Key Points

Cite This Study

How to get statistically significant effects in any ERP experiment (and why you shouldn't)

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider