# Can parametric statistical methods be trusted for fMRI based group studies?

arXiv: https://arxiv.org/abs/1511.01863 · HTML: https://mlsanity.com/paper/1511.01863/

## Key facts

- Cited by 22 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 5 Nov 2015.

## Built on by (most cited first, 22 of 22)

- [The Dead Salmons of AI Interpretability](https://mlsanity.com/paper/2512.18792/) (2025)
- [On Estimating Rank-One Spiked Tensors in the Presence of Heavy Tailed Errors](https://mlsanity.com/paper/2107.09660/) (2021)
- [Is K-fold cross validation the best model selection method for Machine Learning?](https://mlsanity.com/paper/2401.16407/) (2024)
- [Deep Learning in current Neuroimaging: a multivariate approach with power and type I error control but arguable generalization ability](https://mlsanity.com/paper/2103.16685/) (2021)
- [Robust Matrix Completion with Heavy-tailed Noise](https://mlsanity.com/paper/2206.04276/) (2022)
- [A hypothesis-driven method based on machine learning for neuroimaging data analysis](https://mlsanity.com/paper/2202.04397/) (2022)
- [Bayesian longitudinal tensor response regression for modeling neuroplasticity](https://mlsanity.com/paper/2309.10065/) (2023)
- [The Autocorrelation Blind Spot: Why 42% of Turn-Level Findings in LLM Conversation Analysis May Be Spurious](https://mlsanity.com/paper/2604.14414/) (2026)
- [Multiple Hypothesis Testing Framework for Spatial Signals](https://mlsanity.com/paper/2108.12314/) (2021)
- [Robust High-Dimensional Regression with Coefficient Thresholding and its Application to Imaging Data Analysis](https://mlsanity.com/paper/2109.14856/) (2021)
- [Identification of brain states, transitions, and communities using functional MRI](https://mlsanity.com/paper/2101.10617/) (2021)
- [A practical model-based segmentation approach for improved activation detection in single-subject functional Magnetic Resonance Imaging studies](https://mlsanity.com/paper/2102.03639/) (2021)
- [Learning Interpretable Models for Coupled Networks Under Domain Constraints](https://mlsanity.com/paper/2104.09069/) (2021)
- [Bayesian Regularization for Functional Graphical Models with Applications to Neuroimaging](https://mlsanity.com/paper/2110.05575/) (2021)
- [Robust High-dimensional Tuning Free Multiple Testing](https://mlsanity.com/paper/2211.11959/) (2022)
- [Statistical Agnostic Regression: a machine learning method to validate regression models](https://mlsanity.com/paper/2402.15213/) (2024)
- [A sensitivity analysis to quantify the impact of neuroimaging preprocessing strategies on subsequent statistical analyses](https://mlsanity.com/paper/2404.14882/) (2024)
- [deepmriprep: Voxel-based Morphometry (VBM) Preprocessing via Deep Neural Networks](https://mlsanity.com/paper/2408.10656/) (2024)
- [StreamEnsemble: Predictive Queries over Spatiotemporal Streaming Data](https://mlsanity.com/paper/2410.00933/) (2024)
- [Efficient Group Lasso Regularized Rank Regression with Simulation-Based Tuning](https://mlsanity.com/paper/2510.11546/) (2025)
- [Data-driven techniques for translational neuroscience and personalized neuro-health](https://mlsanity.com/paper/2608.13749/) (2026)
- [Cheap and Powerful Tests for Supervised Subspaces: Per-Component Inference for PLS](https://mlsanity.com/paper/2609.36307/) (2026)

## Abstract (arXiv)

The most widely used task fMRI analyses use parametric methods that depend on a variety of assumptions. While individual aspects of these fMRI models have been evaluated, they have not been evaluated in a comprehensive manner with empirical data. In this work, a total of 2 million random task fMRI group analyses have been performed using resting state fMRI data, to compute empirical familywise error rates for the software packages SPM, FSL and AFNI, as well as a standard non-parametric permutation method. While there is some variation, for a nominal familywise error rate of 5% the parametric statistical methods are shown to be conservative for voxel-wise inference and invalid for cluster-wise inference; in particular, cluster size inference with a cluster defining threshold of p = 0.01 generates familywise error rates up to 60%. We conduct a number of follow up analyses and investigations that suggest the cause of the invalid cluster inferences is spatial auto correlation functions that do not follow the assumed Gaussian shape. By comparison, the non-parametric permutation test, which is based on a small number of assumptions, is found to produce valid results for voxel as well as cluster wise inference. Using real task data, we compare the results between one parametric method and the permutation test, and find stark differences in the conclusions drawn between the two using cluster inference. These findings speak to the need of validating the statistical methods being used in the neuroimaging field.