Limits...
A Permutation Test for Unbalanced Paired Comparisons of Global Field Power.

Files BT, Lawhern VJ, Ries AJ, Marathe AR - Brain Topogr (2016)

Bottom Line: Global field power is a valuable summary of multi-channel electroencephalography data.The results show that the proposed test finds global field power differences in the classical P3 range; the other tests find differences in that range but also at other times including at times before stimulus onset.These results are interpreted as showing that the proposed test is valid and sensitive to real within-subject differences in global field power in multi-subject unbalanced data.

View Article: PubMed Central - PubMed

Affiliation: U.S. Army Research Laboratory, Aberdeen Proving Ground, MD, USA. benjamin.t.files.civ@mail.mil.

ABSTRACT
Global field power is a valuable summary of multi-channel electroencephalography data. However, global field power is biased by the noise typical of electroencephalography experiments, so comparisons of global field power on data with unequal noise are invalid. Here, we demonstrate the relationship between the number of trials that contribute to a global field power measure and the expected value of that global field power measure. We also introduce a statistical testing procedure that can be used for multi-subject, repeated-measures (also called within-subjects) comparisons of global field power when the number of trials per condition is unequal across conditions. Simulations demonstrate the effect of unequal trial numbers on global field power comparisons and show the validity of the proposed test in contrast to conventional approaches. Finally, the proposed test and two alternative tests are applied to data collected in a rapid serial visual presentation target detection experiment. The results show that the proposed test finds global field power differences in the classical P3 range; the other tests find differences in that range but also at other times including at times before stimulus onset. These results are interpreted as showing that the proposed test is valid and sensitive to real within-subject differences in global field power in multi-subject unbalanced data.

No MeSH data available.


Related in: MedlinePlus

Flowchart showing the steps in computing global field power differences across two conditions. Data are obtained from subjects S1 through Sn. Continuous data are epoched around a stimulus event and sorted according to condition to obtain epoched single-trial data of dimensions C channels, S samples per epoch and A or B repetitions for conditions a and b, respectively. Average ERPs are obtained by averaging over repetitions, and then global field power is computed. The unbalanced paired permutation test carries out permutation at the single trial level, before any averaging is done. A conventional permutation test permutes after averaging and computation of global field power, and conventional T test would be done after computing a difference (or equivalently a paired-samples T test would be done on the GFPs before subtraction)
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4829629&req=5

Fig1: Flowchart showing the steps in computing global field power differences across two conditions. Data are obtained from subjects S1 through Sn. Continuous data are epoched around a stimulus event and sorted according to condition to obtain epoched single-trial data of dimensions C channels, S samples per epoch and A or B repetitions for conditions a and b, respectively. Average ERPs are obtained by averaging over repetitions, and then global field power is computed. The unbalanced paired permutation test carries out permutation at the single trial level, before any averaging is done. A conventional permutation test permutes after averaging and computation of global field power, and conventional T test would be done after computing a difference (or equivalently a paired-samples T test would be done on the GFPs before subtraction)

Mentions: The testing procedure described in this paper applies to a specific combination of experimental design and comparison of interest. To pinpoint this combination of design and comparison, we refer to the experiment classification scheme of Greenblatt and Pflieger (2004; their Fig. 1, p. 227). According to that scheme, the comparison for which the unbalanced paired permutation test was designed is a two condition, paired, within-group comparison. Many tests may be used for that general comparison, but only with experimental designs that produce balanced data and/or when using summary statistics that are not biased. The unbalanced paired permutation test we describe applies to a paired, within subjects design in which trial counts in the paired conditions are imbalanced leading to a biased summary statistic. A concrete example of a design and comparison for which this test is appropriate is the typical P3/oddball design in which rare targets are embedded in a series of common distracters and the comparison of interest is the GFP evoked by targets against the GFP evoked by distracters. This design has two conditions (target, distracter), both conditions apply to every subject (the conditions are paired) and the difference in the two conditions is of interest (the comparison is within-group). This example produces unbalanced data (rare targets and frequent distracters) and is using a biased summary statistic (GFP).Fig. 1


A Permutation Test for Unbalanced Paired Comparisons of Global Field Power.

Files BT, Lawhern VJ, Ries AJ, Marathe AR - Brain Topogr (2016)

Flowchart showing the steps in computing global field power differences across two conditions. Data are obtained from subjects S1 through Sn. Continuous data are epoched around a stimulus event and sorted according to condition to obtain epoched single-trial data of dimensions C channels, S samples per epoch and A or B repetitions for conditions a and b, respectively. Average ERPs are obtained by averaging over repetitions, and then global field power is computed. The unbalanced paired permutation test carries out permutation at the single trial level, before any averaging is done. A conventional permutation test permutes after averaging and computation of global field power, and conventional T test would be done after computing a difference (or equivalently a paired-samples T test would be done on the GFPs before subtraction)
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4829629&req=5

Fig1: Flowchart showing the steps in computing global field power differences across two conditions. Data are obtained from subjects S1 through Sn. Continuous data are epoched around a stimulus event and sorted according to condition to obtain epoched single-trial data of dimensions C channels, S samples per epoch and A or B repetitions for conditions a and b, respectively. Average ERPs are obtained by averaging over repetitions, and then global field power is computed. The unbalanced paired permutation test carries out permutation at the single trial level, before any averaging is done. A conventional permutation test permutes after averaging and computation of global field power, and conventional T test would be done after computing a difference (or equivalently a paired-samples T test would be done on the GFPs before subtraction)
Mentions: The testing procedure described in this paper applies to a specific combination of experimental design and comparison of interest. To pinpoint this combination of design and comparison, we refer to the experiment classification scheme of Greenblatt and Pflieger (2004; their Fig. 1, p. 227). According to that scheme, the comparison for which the unbalanced paired permutation test was designed is a two condition, paired, within-group comparison. Many tests may be used for that general comparison, but only with experimental designs that produce balanced data and/or when using summary statistics that are not biased. The unbalanced paired permutation test we describe applies to a paired, within subjects design in which trial counts in the paired conditions are imbalanced leading to a biased summary statistic. A concrete example of a design and comparison for which this test is appropriate is the typical P3/oddball design in which rare targets are embedded in a series of common distracters and the comparison of interest is the GFP evoked by targets against the GFP evoked by distracters. This design has two conditions (target, distracter), both conditions apply to every subject (the conditions are paired) and the difference in the two conditions is of interest (the comparison is within-group). This example produces unbalanced data (rare targets and frequent distracters) and is using a biased summary statistic (GFP).Fig. 1

Bottom Line: Global field power is a valuable summary of multi-channel electroencephalography data.The results show that the proposed test finds global field power differences in the classical P3 range; the other tests find differences in that range but also at other times including at times before stimulus onset.These results are interpreted as showing that the proposed test is valid and sensitive to real within-subject differences in global field power in multi-subject unbalanced data.

View Article: PubMed Central - PubMed

Affiliation: U.S. Army Research Laboratory, Aberdeen Proving Ground, MD, USA. benjamin.t.files.civ@mail.mil.

ABSTRACT
Global field power is a valuable summary of multi-channel electroencephalography data. However, global field power is biased by the noise typical of electroencephalography experiments, so comparisons of global field power on data with unequal noise are invalid. Here, we demonstrate the relationship between the number of trials that contribute to a global field power measure and the expected value of that global field power measure. We also introduce a statistical testing procedure that can be used for multi-subject, repeated-measures (also called within-subjects) comparisons of global field power when the number of trials per condition is unequal across conditions. Simulations demonstrate the effect of unequal trial numbers on global field power comparisons and show the validity of the proposed test in contrast to conventional approaches. Finally, the proposed test and two alternative tests are applied to data collected in a rapid serial visual presentation target detection experiment. The results show that the proposed test finds global field power differences in the classical P3 range; the other tests find differences in that range but also at other times including at times before stimulus onset. These results are interpreted as showing that the proposed test is valid and sensitive to real within-subject differences in global field power in multi-subject unbalanced data.

No MeSH data available.


Related in: MedlinePlus