×

Identifying treatment effects in multi-channel measurements in electroencephalographic studies: multivariate permutation tests and multiple comparisons. (English) Zbl 1521.62208

Summary: A versatile procedure is described comprising an application of statistical techniques to the analysis of the large, multi-dimensional data arrays produced by electroencephalographic (EEG) measurements of human brain function. Previous analytical methods have been unable to identify objectively the precise times at which statistically significant experimental effects occur, owing to the large number of variables (electrodes) and small number of subjects, or have been restricted to two-treatment experimental designs. Many time-points are sampled in each experimental trial, making adjustment for multiple comparisons mandatory. Given the typically large number of comparisons and the clear dependence structure among time-points, simple Bonferroni-type adjustments are far too conservative. A three-step approach is proposed: (i) summing univariate statistics across variables; (ii) using permutation tests for treatment effects at each time-point; and (iii) adjusting for multiple comparisons using permutation distributions to control family-wise error across the whole set of time-points. Our approach provides an exact test of the individual hypotheses while asymptotically controlling family-wise error in the strong sense, and can provide tests of interaction and main effects in factorial designs. An application to two experimental data sets from EEG studies is described, but the approach has application to the analysis of spatio-temporal multivariate data gathered in many other contexts.

MSC:

62P10 Applications of statistics to biology and medical sciences; meta analysis
62K15 Factorial statistical designs
62J15 Paired and multiple comparisons; multiple testing

Software:

R
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] Anderson, Permutation tests for univariate or multivariate analysis and regression, Can. J. Fish Aquat. Sci. 58 pp 626– (2001)
[2] Anderson, Permutation tests for linear models., Aust. NZ J. Statist. 43 pp 75– (2001) · Zbl 0992.62043
[3] Anderson, Permutation tests for multi-factorial analysis of variance., J. Statist. Comput. Simul. 73 pp 85– (2003) · Zbl 1014.62051
[4] Benjamini, Controlling the false discovery rate: A practical and powerful approach to multiple testing., J. Roy. Statist. Soc. Ser. B 57 pp 289– (1995) · Zbl 0809.62014
[5] Berger, Über das Elektroenkephalogramm des Menschen., Arch. Psychiat. Nerven 87 pp 527– (1929)
[6] Blair, An alternative method for significance testing of waveform difference potentials., Psychophysiology 30 pp 518– (1993)
[7] Chung, Randomization tests for a multivariate two-sample problem., J. Amer. Statist. Assoc. 53 pp 729– (1958) · Zbl 0087.14701
[8] Dougherty, Dichotic pitch: A new stimulus distinguishes normal and dyslexic auditory function., Neuroreport 9 pp 3001– (1998)
[9] Edgington, Randomization Tests (1995)
[10] Fisher, Design of Experiments (1935)
[11] Freedman, A nonstochastic interpretation of reported significance levels., J. Bus. Econ. Statist. 1 pp 292– (1983)
[12] Galán, Testing topographic differences between event related brain potentials by using non-parametric combinations of permutation tests., Electroen. Clin. Neuro. 102 pp 240– (1997)
[13] Hautus, Object-related potentials associated with the perceptual segregation of dichotically-embedded pitch., J. Acoust. Soc. Amer. 117 pp 275– (2004)
[14] Hemmelmann, Multivariate tests for the evaluation of high-dimensional EEG data., J. Neurosci. Meth. 139 pp 111– (2004)
[15] Hemmelmann, New concepts of multiple tests and their use for evaluating high-dimensional EEG data., J. Neurosci. Meth. 142 pp 209– (2005)
[16] Hochberg, A sharper Bonferroni procedure for multiple tests of significance., Biometrika 75 pp 800– (1988) · Zbl 0661.62067
[17] Hochberg, Multiple Comparison Procedures (1987)
[18] Holm, A simple sequentially rejective multiple test procedure., Scand. J. of Statist. 6 pp 65– (1979) · Zbl 0402.62058
[19] Hope, A simplified Monte Carlo significance test procedure., J. Roy. Statist. Soc. Ser. B 30 pp 582– (1968)
[20] Johnson, Neural activity associated with binaural processes or the perceptual segregation of pitch., Clin. Neurophysiol. 114 pp 2245– (2003)
[21] Johnson, Sequential processing of interaural timing differences for sound source segregation and spatial localization: Evidence from cortical event-related potentials, Psychophysiology 44 pp 541– (2007)
[22] Kempthorne, The randomization theory of experimental inference., J. Amer. Statist. Assoc. 50 pp 946– (1955)
[23] Legendre, Distance-based redundancy analysis: Testing multispecies responses in multifactorial ecological experiments., Ecological Monographs 69 pp 1– (1999)
[24] Lehmann, Reference-free identification of components of checkerboard-evoked multi-channel potential fields., Electroen. Clin. Neuro. 48 pp 609– (1980)
[25] Manly, Randomization, Bootstrap and Monte Carlo Methods in Biology (1997)
[26] Manly, A randomization procedure for comparing group means on multiple measurements., Brit. J. Math. Statist. Psy. 39 pp 183– (1986) · doi:10.1111/j.2044-8317.1986.tb00855.x
[27] Michel, EEG source imaging., Clin. Neurophysiol. 115 pp 2195– (2004)
[28] R Development Core Team, R: A Language and Environment for Statistical Computing. (2005)
[29] Venables, Modern Applied Statistics with S (2002) · doi:10.1007/978-0-387-21706-2
[30] Verdonschott, An experimental manipulation of oligochaete communities in mesocosms treated with chlorpyrifos or nutrient additions: Multivariate analyses with Monte Carlo permutation tests., Hydrobiologia 278 pp 251– (1994)
[31] Westfall, Resampling-Based Multiple Testing: Examples and Methods for p-Value Adjustment (1993)
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.