StatrivaA practical guide to research data analysis.
A reference hub for common statistical methods and research designs. The right method depends on the question, design, outcome, and structure of the data.
Methods I commonly work with
Use these descriptions as orientation, not as a substitute for matching the method to the actual study design.
t-tests, ANOVA & nonparametric tests
Compare groups while considering outcome scale, independence, variance structure, and the assumptions of the selected test.
Regression & generalized models
Model relationships between predictors and outcomes, including continuous, binary, count, and other response structures.
Mixed-effects & longitudinal models
Account for repeated measurements, clustered observations, and hierarchical structures when ordinary regression would treat observations as independent.
PCA, ordination & community analysis
Explore multivariate structure and compare complex observations using methods appropriate to the data and research question.
Sample size & power
Plan studies around a primary outcome, expected effect, uncertainty, target power, and the actual structure of the proposed analysis.
Survival analysis
Analyze time-to-event outcomes while accounting for censoring and the way risk changes over follow-up.
Meta-analysis
Combine effect estimates across studies, quantify heterogeneity, and present pooled evidence with appropriate uncertainty.
Diagnostic test evaluation
Evaluate discrimination and classification performance using sensitivity, specificity, predictive values, and ROC analysis.
Trend, seasonality & forecasting
Analyze observations collected over time while accounting for temporal dependence, trend, and seasonality.
R, SPSS, and Python
The software is a tool, not the method. The workflow should be understandable and reproducible in the environment that fits your project.
Flexible statistical modeling, visualization, multivariate analysis, and reproducible workflows.
Common in academic and applied research workflows, especially for conventional statistical analyses.
Useful for data preparation, statistical analysis, automation, and integrated analytical workflows.
Go deeper
For project-specific guidance, explore the service pages rather than choosing a method from a list alone.
Statistical analysis for theses →
Not sure which method fits?
Start with the research question and study design. A statistical consultation can identify the appropriate analysis before you spend time running the wrong one.
Talk through the projectQuestions about methods
Can you help me choose a statistical test?
Yes. The choice should follow the research question, outcome type, study design, predictors, repeated measurements, and assumptions rather than a simple test-selection chart.
Do you work with R, SPSS, and Python?
Yes. Statistical workflows can be developed in R, SPSS, or Python depending on the project and the software you or your department use.
Can you explain a statistical method in plain language?
Yes. The goal is not only to produce output, but to explain what the method estimates, what its assumptions mean, and how the result relates to the research question.