Dataset Information

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

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.

SUBMITTER: Gorgen K

PROVIDER: S-EPMC6021230 | biostudies-literature | 2018 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

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

Görgen Kai K Hebart Martin N MN Allefeld Carsten C Haynes John-Dylan JD

NeuroImage 20171227 Pt A

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, analysi ...[more]

PMID: 29288130

Similar Datasets

Project description:Since the appearance of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the question regarding the efficacy of various hygiene measures and the use of personal protective equipment (PPE) has become the focus of scientific and above all public discussion. To compare respirators, medical face masks, and cloth masks and determine if it is recommendable to wear face masks to protect the individual wearer of the mask from inhaling airborne particles, we challenged 29 different masks with aerosols and tested the pressure drop as a surrogate for breathing resistance owing to the mask material. We found that Type II medical face masks showed the lowest pressure drop (12.9±6.8 Pa/cm2) and therefore additional breathing resistance, whereas respirators such as the KN95 (32.3±7.0 Pa/cm2) and FFP2 (26.8±7.4 Pa/cm2) showed the highest pressure drops among the tested masks. The filtration efficacy of the mask material was the lowest for cloth masks (28±25%) followed by non-certified face masks (63±19%) and certified medical face masks (70±10%). The materials of the different respirators showed very high aerosol retentions (KN95 [94±4%] and FFP2 [98±1%]). For evaluating the as-worn filtration performance simulating real live conditions each mask type was also tested on a standardized dummy head. Cloth masks and non-EN-certified face masks had the worst as-worn filtration efficacies among the tested masks, filtering less than 20% of the test aerosol. Remarkably, certified type II medical face masks showed similar (p>0.5) as-worn filtration results (47±20%) than KN95 masks (41±4%) and FFP2 masks (65±27%), despite having a lower pressure drop. Face shields did not show any significant retention function against aerosols in our experiment. Our results indicate that it seems recommendable to wear face masks for providing base protection and risk reduction against inhaling airborne particles, in low-risk situations. In our study, especially EN 14683 type II certified medical face masks showed protective effectiveness against aerosols accompanied by minimal additional breathing resistance. FFP2 Respirators, on the other hand, could be useful in high-risk situations but require greater breathing effort and therefore physical stress for users.

Dataset Information

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

Publications

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

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets