How to use bootstrap to test mediation effects

Using Bootstrap to test the mediation effect

What is the Bootstrap test?

Bootstrap is a resampling technique used to estimate the sampling distribution of a statistic. By randomly drawing multiple subsamples from the original data, it can help us evaluate the robustness and confidence of a statistical test.

How to use Bootstrap to test the mediation effect?

Testing for mediation effects involves testing whether the relationship between the independent variable (X) and the dependent variable (Y) is partially or fully mediated by the mediator variable (M). The Bootstrap test can be used:

1. Test the significance of the mediation effect

• Use the original data to calculate the direct effect (X-Y) and indirect effect (X-M-Y ).
• Use Bootstrap to generate multiple subsamples and calculate direct and indirect effects in each subsample.
• Calculate the Bootstrap confidence interval (CI) of the mediation effect (indirect effect - direct effect).
• If the CI does not contain 0, it indicates that the mediation effect is statistically significant.

2. Test the significance of the mediation type

• Use the method of Baron and Kenny (1986) or the Sobel test to calculate the significance of the mediation type Significance.
• Calculate the Bootstrap CI of Baron and Kenny statistics or Sobel statistics based on the steps of Bootstrap sampling.
• If the CI does not contain 0, it indicates that the mediation type is statistically significant.

3. Estimating the natural indirect effect

• Use Bootstrap to estimate the CI of the natural indirect effect. The natural indirect effect represents the indirect effect of the independent variable on the dependent variable when the moderator variable is fixed.
• Calculate the CI for the natural indirect effect by fixing the mediating variable in each Bootstrap subsample and calculating the indirect effect.

Advantages:

• is not restricted by the assumption of normal distribution.
• Provide more stable effect estimates.
• Allows non-parametric testing.

Note:

• Requires a large number of data samples to obtain accurate results.
• It is computationally intensive, especially for complex mediation models.

The above is the detailed content of How to use bootstrap to test mediation effects. For more information, please follow other related articles on the PHP Chinese website!