Understanding Indirect Effect Mediation
Mediation analysis studies how one variable explains a relationship. It separates a total effect into direct and indirect parts. The indirect effect passes through a mediator. The direct effect stays after the mediator is considered. This calculator helps you inspect that pathway with common estimates used in research reports.
Why This Calculator Helps
Many studies report path a and path b. Path a links the predictor to the mediator. Path b links the mediator to the outcome while controlling the predictor. Multiplying both paths gives the indirect effect. A positive value suggests the mediator carries the relationship in one direction. A negative value suggests an opposite mediated pathway.
Advanced Measures Included
The tool reports Sobel, Aroian, and Goodman standard error methods. These tests use different variance adjustments. The Aroian method is often more conservative because it adds the product of both standard errors. The Goodman method subtracts that product term. The calculator also creates a Monte Carlo interval. It simulates many possible a and b values from their standard errors.
Interpreting Results
A z score shows how far the indirect effect is from zero. The p value gives a familiar significance check. A confidence interval that excludes zero supports a mediated effect. The proportion mediated shows how much of the total effect is represented by the indirect path. This number can be unstable when the total effect is small.
Best Practice
Use this page as a planning and reporting aid. It does not replace a full statistical model. Good mediation analysis still needs valid data, clear theory, reliable measures, and correct model assumptions. Always report sample size, path estimates, standard errors, intervals, and the method used. For final publications, compare these values with results from your primary statistics software.
Practical Use
Researchers can test pilot data, verify textbook examples, or prepare classroom demonstrations. Analysts can compare several models quickly. Students can learn how each formula reacts to changing path estimates. The result table also supports copying, exporting, and checking values without rebuilding a spreadsheet.
Notes
Select coefficients from the same mediation model whenever possible. Matched estimates reduce confusion. Use raw paths, or standardized paths across different scales. Clear records make mediation reporting easier and more defensible.