Autor: |
Senoussi, Mehdi, Vinding, Mikkel C., Algermissen, Johannes, Trübutschek, Darinka, Pascarella, Annalisa, Marshall, Tom, Yang, Yu-Fang, Puoliväli, Tuomas, Yeaton, Jeremy, Busch, Niko, Fischer, Nastassja L., Gianelli, Claudia, Cesnaite, Elena, Navid, Muhammad Samran, Nilsonne, Gustav, Koen, Joshua, Vitale, Andrea |
Rok vydání: |
2022 |
Předmět: |
|
DOI: |
10.17605/osf.io/xfrbe |
Popis: |
In psychology, a growing number of scientists have raised concerns about how reproducible and replicable the results described by different research groups are (Open Science Collaboration, 2015, Klein et al., 2018). These concerns have also reached the neuroscience community, and some have argued that the complexity of neuroimaging analyses and lack of consensus on optimal analysis processing steps (i.e., pipelines) of the data leads to many degrees of freedom and a high risk of “overfitting”, causing variability of results (Poldrack et al., 2017, Pavlov et al., 2021). A recent empirical investigation where different research teams analysed the same fMRI dataset confirmed considerable variability in outcomes due to different analytic choices along the process (Botvinik-Nezer et al. 2020). Regarding EEG research, studies have investigated the extent to which the change of one or more preprocessing steps may alter the final results (Robbins et al. 2020) and how experimenters’ degrees of freedom can lead to unreliable outcomes (Luck and Gaspelin, 2017). However, it is still unknown to which extent different EEG analysis practices by different research groups “in the wild” affect final results given the same dataset and same hypotheses. To tackle this highly pressing question the EEGManyPipelines project (https://www.eegmanypipelines.org/) was founded, which aims to investigate how different analysts (i.e. a broad spectrum of research groups) approach the same research question using the same dataset. The project primarily and foremost will investigate different practices and the effect these different analysis strategies have on the result and provide suggestions for future analysis practices. |
Databáze: |
OpenAIRE |
Externí odkaz: |
|