English
 
User Manual Privacy Policy Disclaimer Contact us
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT
  The same analysis approach: Practical protection against the pitfalls of novel neuroimaging analysis methods

Görgen, K., Hebart, M. N., Allefeld, C., & Haynes, J.-D. (2017). The same analysis approach: Practical protection against the pitfalls of novel neuroimaging analysis methods. NeuroImage, 180(Part A), 19-30. doi:10.1016/j.neuroimage.2017.12.083.

Item is

Basic

show hide
Item Permalink: http://hdl.handle.net/21.11116/0000-0005-20D0-B Version Permalink: http://hdl.handle.net/21.11116/0000-0005-5894-1
Genre: Journal Article

Files

show Files

Locators

show

Creators

show
hide
 Creators:
Görgen, Kai1, Author
Hebart, Martin N.1, Author              
Allefeld, Carsten 1, Author
Haynes, John-Dylan1, Author
Affiliations:
1External Organizations, ou_persistent22              

Content

show
hide
Free keywords: Experimental design; Confounds; Multivariate pattern analysis; Cross validation; Below-chance accuracies; Unit testing
 Abstract: Standard neuroimaging data analysis based on traditional principles of experimental design, modelling, and statistical inference is increasingly complemented by novel analysis methods, driven e.g. by machine learning methods. While these novel approaches provide new insights into neuroimaging data, they often have unexpected properties, generating a growing literature on possible pitfalls. We propose to meet this challenge by adopting a habit of systematic testing of experimental design, analysis procedures, and statistical inference. Specifically, we suggest to apply the analysis method used for experimental data also to aspects of the experimental design, simulated confounds, simulated null data, and control data. We stress the importance of keeping the analysis method the same in main and test analyses, because only this way possible confounds and unexpected properties can be reliably detected and avoided. We describe and discuss this Same Analysis Approach in detail, and demonstrate it in two worked examples using multivariate decoding. With these examples, we reveal two sources of error: A mismatch between counterbalancing (crossover designs) and cross-validation which leads to systematic below-chance accuracies, and linear decoding of a nonlinear effect, a difference in variance.

Details

show
hide
Language(s):
 Dates: 2017-03-152017-12-232017-12-27
 Publication Status: Published online
 Pages: -
 Publishing info: -
 Table of Contents: -
 Rev. Method: Peer
 Identifiers: DOI: 10.1016/j.neuroimage.2017.12.083
PMID: 29288130
PMC: PMC6021230
Other: Epub 2017
 Degree: -

Event

show

Legal Case

show

Project information

show hide
Project name : -
Grant ID : GRK1589/1 ; JA945/3-1
Funding program : -
Funding organization : German Research Foundation (DFG)
Project name : -
Grant ID : 01GQ1006
Funding program : -
Funding organization : German Federal Ministry of Education and Research (BMBF)
Project name : -
Grant ID : 93-M-380170
Funding program : Intramural Research Program
Funding organization : National Institute of Mental Health (NIMH)
Project name : -
Grant ID : -
Funding program : Feodor-Lynen Fellowship
Funding organization : Alexander von Humboldt Foundation

Source 1

show
hide
Title: NeuroImage
Source Genre: Journal
 Creator(s):
Affiliations:
Publ. Info: Orlando, FL : Academic Press
Pages: - Volume / Issue: 180 (Part A) Sequence Number: - Start / End Page: 19 - 30 Identifier: ISSN: 1053-8119
CoNE: https://pure.mpg.de/cone/journals/resource/954922650166