Skip to content

puneettalwar/StudyPower

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 

Repository files navigation

StudyPower (PosthocPowerShiny)

*** Warning! This Shiny App has not been validated enough! There is no warranty for the app! ***

The link to Shiny App - https://puneet-talwar.shinyapps.io/PosthocPowerShiny/

Linear Multiple Regression: Power & Sensitivity Analysis

-This app replicates and extends G*Power for post-hoc power and sensitivity analysis in linear multiple regression.

Inputs

-Sample size (N): Number of participants (must be > k + 1).

-Total predictors (k): Total predictors in the model.

-Tested predictors (u): Predictors of interest being tested (k for the full model)).

-Significance level (alpha): Type I error probability.

-Confidence level: Confidence interval for effect sizes.

-Analysis type: Choose between Target Power Analysis or Observed R² Analysis.

-Run Analysis: Press the button to compute results.

Outputs

-Numerical Results: Displays effect sizes (f, f², R², r, d) with confidence intervals.

-Power Curve: Shows power vs effect size (f²).

-Nomogram: Shows power curves across different sample sizes.

How to Use

-For sensitivity analysis: Enter N, k, u, alpha, and target power.

-For post-hoc analysis: Check 'Use observed R² or f², enter R² or f².

-Click 'Run Analysis'.

-Interpret results and plots for your study context.

Notes

-Ensure N > k + 1.

-Power curve assumes 1 tested predictor (u = 1).

-Confidence intervals are based on non-central F distributions.

-Use this app for replication of G*Power analyses in R, with added confidence interval reporting.-

About

R shiny app to determine minimum effect size for a priori study sample size/power calculation

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages