This paper by Maris and Oostenveld (2007) discusses the statistical analysis of ElectroEncephaloGraphic (EEG) and MagnetoEncephaloGraphic (MEG) data using nonparametric techniques. The authors highlight the advantages of nonparametric tests, particularly in solving the multiple comparisons problem (MCP) and incorporating biophysically motivated constraints to enhance the sensitivity of the statistical test. The paper is divided into two main sections: methods and justification. In the methods section, the authors provide a tutorial on how to apply nonparametric statistical tests to EEG and MEG data, focusing on single-sensor and multi-sensor analyses. They use an example dataset from a study on semantic processing of sentences to illustrate the application of these tests. The example dataset involves comparing two experimental conditions based on the semantic congruity of sentence endings. The authors demonstrate how to calculate sample-specific \( t \)-values and \( p \)-values, and how to apply Bonferroni correction to control the family-wise error rate (FWER). They also introduce the cluster-based nonparametric test, which is more sensitive and robust to the MCP. In the justification section, the authors provide a theoretical foundation for the nonparametric tests. They define the null hypothesis and explain the concept of exchangeability, which is crucial for the validity of the permutation test. The authors show that the permutation test controls the false alarm rate (FAR) under the null hypothesis of identical probability distributions across different experimental conditions. They also discuss the strong and weak control of the FAR, emphasizing that for MEEG data, it is more appropriate to focus on global null hypotheses rather than voxel-specific null hypotheses due to the spatial correlation between MEEG signals. Overall, the paper aims to provide a comprehensive guide for empirical neuroscientists and methodologists on how to effectively analyze EEG and MEG data using nonparametric statistical tests, addressing both practical implementation and theoretical underpinnings.This paper by Maris and Oostenveld (2007) discusses the statistical analysis of ElectroEncephaloGraphic (EEG) and MagnetoEncephaloGraphic (MEG) data using nonparametric techniques. The authors highlight the advantages of nonparametric tests, particularly in solving the multiple comparisons problem (MCP) and incorporating biophysically motivated constraints to enhance the sensitivity of the statistical test. The paper is divided into two main sections: methods and justification. In the methods section, the authors provide a tutorial on how to apply nonparametric statistical tests to EEG and MEG data, focusing on single-sensor and multi-sensor analyses. They use an example dataset from a study on semantic processing of sentences to illustrate the application of these tests. The example dataset involves comparing two experimental conditions based on the semantic congruity of sentence endings. The authors demonstrate how to calculate sample-specific \( t \)-values and \( p \)-values, and how to apply Bonferroni correction to control the family-wise error rate (FWER). They also introduce the cluster-based nonparametric test, which is more sensitive and robust to the MCP. In the justification section, the authors provide a theoretical foundation for the nonparametric tests. They define the null hypothesis and explain the concept of exchangeability, which is crucial for the validity of the permutation test. The authors show that the permutation test controls the false alarm rate (FAR) under the null hypothesis of identical probability distributions across different experimental conditions. They also discuss the strong and weak control of the FAR, emphasizing that for MEEG data, it is more appropriate to focus on global null hypotheses rather than voxel-specific null hypotheses due to the spatial correlation between MEEG signals. Overall, the paper aims to provide a comprehensive guide for empirical neuroscientists and methodologists on how to effectively analyze EEG and MEG data using nonparametric statistical tests, addressing both practical implementation and theoretical underpinnings.