Cognitive Science Statistical Analysis

Choosing between repeated-measures ANOVA and mixed-effects models depends on your experimental data structure. Use maximal mixed-effects models for complex random effects hierarchies, and reserve ANOVA for simpler repeated-measures designs.

Handling outliers in reaction time analysis requires specifying explicit thresholds and transformation strategies. You must decide whether to analyze raw trials or aggregated means, using Gamma GLMMs for rigorous modeling of skewed distributions.

Yes, Bayesian analysis offers an alternative for mixed designs. It provides Bayesian evidence reporting to quantify support for hypotheses, complementing or replacing traditional multiple comparison corrections in multimodal studies.

Publication standards in cognitive science require specific effect size benchmarks and a reporting checklist. Following APA 7th edition conventions and domain literature ensures your statistical modeling and reporting meet rigorous peer-review expectations.

Maximal mixed-effects models fail to converge when random effects structures are overly complex relative to the data. Troubleshoot by simplifying random slopes or switching to alternative GLMMs to achieve stable estimates.