Sample Size & Power Analysis Support
Sample size and power analysis support for research design, theses, grant proposals, experiments, surveys, and clinical studies.
Sample-size planning should be tied to the primary analysis, expected effect, desired precision or power, significance level, design features, and realistic attrition. Tezyar helps translate a proposed study design into a transparent sample-size rationale rather than applying a generic rule of thumb.
Define the Primary Analysis
The calculation should reflect the main outcome and planned statistical comparison or model because different analyses require different inputs.
Effect Size and Precision
Effect-size assumptions should be justified from prior evidence, pilot data, clinically meaningful differences, or a sensitivity range when uncertainty is high.
Design Effects and Attrition
Clustering, repeated measures, unequal allocation, expected missingness, and attrition can materially change the required sample.
Sensitivity Analysis
When inputs are uncertain, showing a range of plausible sample sizes can be more informative than presenting one apparently exact number.
Frequently Asked Questions
Can you calculate sample size before data collection?
Yes. Prospective planning is the usual setting for power- or precision-based sample-size analysis.
Is a post-hoc power calculation useful after a study is complete?
Usually confidence intervals and observed precision are more informative than post-hoc power based on the observed effect.
Related specialized topics
SPSS analysis support for research projects, theses, and manuscripts: data screening, test selection, modelling, interpretation, and reporting.
R analysis support for reproducible research, statistical modelling, visualization, reporting, and thesis or manuscript workflows.
Structural equation modeling support for CFA, latent variables, path models, fit assessment, model revision, and reporting.
Regression analysis support for linear, logistic, generalized, multivariable, and research-specific modelling workflows.
Survey analysis support for data cleaning, scale reliability, descriptive analysis, group comparisons, regression, factor analysis, and reporting.
Statistical support for clinical and health research: analysis planning, outcomes, regression, survival methods, repeated measures, and reporting.