We’re excited to introduce a new power analysis module for jamovi, designed to simplify and enhance your research planning. This module supports a broad range of statistical tests, making it a useful tool for researchers across various fields. Whether you’re conducting ANOVA, regression, mediation analysis, t-tests, correlation, proportions, general linear models, or Structural Equation Models (SEM), this module allows you to perform robust power analyses in one convenient place.
Key Features
Here’s what you can do with the new power analysis module:
- Calculate necessary sample size: Based on your specified effect size and desired power, the module helps you determine the optimal sample size for your study, ensuring it’s adequately powered to detect significant results.
- Compute expected power: You can input your planned sample size and effect size to calculate the expected power of your study, helping you assess whether your design is likely to produce meaningful results.
- Determine minimal detectable effect size: For studies with a fixed sample size, this feature allows you to calculate the smallest effect size that can be reliably detected, ensuring your study has the sensitivity to uncover important findings.
Sensitivity Analysis with Graphs and Tables
A standout feature of this module is its ability to conduct sensitivity analysis, a crucial step for assessing the robustness of your study design. Sensitivity analysis enables you to explore how changes in sample size, effect size, or power affect the overall study outcomes. This is especially useful when planning for uncertain or variable conditions. You can:
- Visualize sensitivity analysis results using interactive graphs that show how different parameters interact.
- View detailed tables that provide clear, numeric insights into how changes in one parameter impact others. This dual output system, combining graphs and tables, allows you to both visually explore and precisely quantify how various factors influence the power of your study.
Expanded Statistical Test Support
The module supports a wide array of statistical methods, offering flexibility for researchers across disciplines. Starting with simpler tests, such as:
- T-tests: Compare two means and ensure your tests are powered to detect meaningful differences.
- Correlation: Estimate the power needed to detect relationships between continuous variables.
-
Proportions: Plan studies to detect differences in proportions with precision. Base your analysis on several effect size indices, from probabilities to odd-ratios or odd differences.
- General Linear Models: Handle complex models involving multiple predictors and interactions. Base your analysis on several effect size indices, such as the η², the partial η², β coefficients and R². With the GLM sub-module one can plan studies and estimate power for Linear Regression, ANOVA, or combinations of the two.
Specialized sub-modules are aslo available for more advanced tests:
-
Mediation analysis: Assess the power of your mediation models, including indirect effects.
-
Structural Equation Models (SEM): With growing support for SEM in jamovi through the SEMLj module, with PAMLj you can now calculate the power and sample size for sophisticated models, ensuring your latent variable models are well-powered for accurate results.
For all these applications, the module offers different methods for computing power, including both analytical and simulation-based approaches.
Streamlined User Experience
Accurate power analysis is essential for ensuring the success of any research project. Underpowered studies often lead to inconclusive results, wasting valuable resources and time. By using this new power analysis module, you can confidently plan your studies, knowing that your sample sizes and effect sizes are appropriately matched to your research goals. The inclusion of sensitivity analysis, combined with both graphical and tabular outputs, ensures you have a thorough understanding of how various factors influence your study’s potential outcomes.
This power analysis module is a comprehensive and user-friendly tool that addresses the key needs of researchers in planning effective studies. Whether you’re determining sample size, calculating expected power, or running sensitivity analyses, this module offers a streamlined, integrated experience within jamovi, making it easier than ever to ensure your research is well-designed and statistically sound. Try it out today and take the guesswork out of your power analysis!
Help
PAMLj comes with a (growing) documentation with module description, examples, and validation against other power analysis software. Please visit its help page and tutorial for details.
Comments