Bootstrapping the P300 in diagnostic psychophysiology: How many iterations are needed?

Autor: Rosenfeld JP; Department of Psychology, Northwestern University, Evanston, Illinois, USA., Ward A; Department of Psychology, Northwestern University, Evanston, Illinois, USA., Meijer EH; Faculty of Psychology and Neuroscience, Maastricht University, Maastricht, The Netherlands., Yukhnenko D; Faculty of Psychology and Neuroscience, Maastricht University, Maastricht, The Netherlands.
Jazyk: angličtina
Zdroj: Psychophysiology [Psychophysiology] 2017 Mar; Vol. 54 (3), pp. 366-373. Date of Electronic Publication: 2016 Dec 07.
DOI: 10.1111/psyp.12789
Abstrakt: In psychophysiological research, bootstrapping procedures are often used to classify individual participants. How many iterations are required for a reliable bootstrap test is not universally agreed upon. To investigate the number of iterations needed for a stable bootstrap estimate, we reanalyzed P300 data collected in concealed information test paradigms. We also distinguished between the bootstrap and permutations approaches. We compared results in several studies using 100 versus 1,000 versus 10,000 iterations in the bootstrap, and we concluded that 100 iterations were adequate as results from all three iteration numbers correlated highly.
(© 2016 Society for Psychophysiological Research.)
Databáze: MEDLINE
Externí odkaz:
Nepřihlášeným uživatelům se plný text nezobrazuje