Abstract:The critical importance of justifying the inclusion of covariates is a facet often overlooked in data analysis. While the incorporation of covariates typically follows informal guidelines, we argue for a comprehensive exploration of underlying principles to avoid significant statistical and interpretational challenges. Our focus is on addressing three common yet problematic practices: the indiscriminate lumping of covariates, the lack of rationale for covariate inclusion, and the oversight of potential issues in result reporting. These challenges, prevalent in neuroimaging models involving covariates such as reaction time, demographics, and morphometric measures, can introduce biases, including overestimation, underestimation, masking, sign flipping, or spurious effects. Our exploration of causal inference principles underscores the pivotal role of domain knowledge in guiding covariate selection, challenging the common reliance on statistical measures. This understanding carries implications for experimental design, model-building, and result interpretation. We draw connections between these insights and reproducibility concerns, specifically addressing the selection bias resulting from the widespread practice of strict thresholding, akin to the logical pitfall associated with "double dipping." Recommendations for robust data analysis involving covariates encompass explicit research question statements, justified covariate inclusions/exclusions, centering quantitative variables for interpretability, appropriate reporting of effect estimates, and advocating a "highlight, don't hide" approach in result reporting. These suggestions are intended to enhance the robustness, transparency, and reproducibility of covariate-driven analyses, encompassing investigations involving consortium datasets such as ABCD and UK Biobank. We discuss how researchers can use a transparent depiction of the covariate relationships to enhance the ethos of open science and promote research reproducibility.

Claim causality with clarity.

A Survey on Causal Inference

A Survey of Causal Inference Frameworks

Through the lens of causal inference: Decisions and pitfalls of covariate selection

Causal inference: critical developments, past and future

Distinguishing Between Mediators and Confounders is Important for the Causal Inference in Observational Studies

Bayesian causal inference: a critical review

Causal Inference with Latent Variables: Recent Advances and Future Prospectives

Causal Inference and Related Statistical Methods

Causal inference for psychologists who think that causal inference is not for them

Statistical Approaches for Causal Inference

Reasoning about causal relationships: Inferences on causal networks.

Causal inference on human behaviour

Causality in Cognitive Neuroscience: Concepts, Challenges, and Distributional Robustness

Causal Inference in the Medical Domain: a Survey

The Causal-Neural Connection: Expressiveness, Learnability, and Inference

The Causal Roadmap and Simulations to Improve the Rigor and Reproducibility of Real-data Applications

A review of some recent advances in causal inference

Causal inference in drug discovery and development

Causal Inference in the Age of Big Data: Blind Faith in Data and Technology