Choose a focused methodological service or a connected project from design through reporting. Start with the decision, not a software procedure.
Start with a non-confidential summary. No files required.
Each service explains when it is useful, the questions it can answer, and the possible handover.
Align research questions, constructs, variables, design, and analysis before the workflow becomes expensive to change.
Best when: At proposal, protocol, or pre-analysis stage when the study logic is not yet explicit.
Statistical Analysis Plan / Decision table / Pre-analysis checklist
Build a sample-size rationale around the actual design, estimand, uncertainty, clustering, attrition, and feasible recruitment.
Best when: Before collection, during protocol revision, or when a design change invalidates an earlier calculation.
Power rationale / Assumption record / Sensitivity summary
Move from collected data to a documented analysis sequence, diagnostics, sensitivity checks, interpretable results, and clear reporting.
Best when: When data are collected but coding, missingness, model choice, diagnostics, or reporting remain unresolved.
Cleaning log / Analysis code / Tables, figures, and interpretation memo
Check whether items and scales support the inferences that later models are expected to carry.
Best when: When a survey, instrument, latent construct, or score needs scrutiny before outcome modeling.
Measurement decision log / Factor-analysis summary / Validity limitations
Represent clustered, repeated, censored, latent, nonlinear, missing, or predictive structures without forcing them into a simpler model.
Best when: When ordinary regression cannot represent the design or the decision requires stronger validation.
Model specification / Diagnostic record / Validation and interpretation summary
Independently review a plan, dataset, script, output, draft result, or AI-assisted workflow before relying on it.
Best when: When an advisor, reviewer, stakeholder, or your own quality check raises doubts about the analysis.
Priority findings / Revised decision path / Limitations and action memo
Turn analytical decisions into traceable code, documented inputs, stable outputs, and a handover another researcher can follow.
Best when: When work must be reviewed, rerun, updated, transferred, or defended after the original analysis.
Documented code / Reproducibility checklist / Versioned handover package
Add scoped quantitative capacity for a specialist method, workload peak, independent review, or a consistent multi-study workflow.
Best when: When a lab, institution, nonprofit, or corporate research team needs flexible statistical depth.
Scoped workstream / Review notes / Reusable workflow documentation
Tell us what you are trying to learn, the design and evidence you have, and where uncertainty remains.
Start with a non-confidential summary. No files required.