Bootstrapping

Approximate the Distribution of a Statistic through Resampling

Bootstrapping is a resampling method for approximating the sampling distribution of a statistic. You can use bootstrapping to estimate the distribution of a statistic and its properties, such as its mean, bias, standard error, and confidence intervals. Bootstrapping is especially useful in the following situations:

• The theoretical distribution of the statistic is complicated or unknown.

• Inference using parametric methods is not possible because of violations of assumptions.

Note: Bootstrap is available only from a right-click in a report. It is not a platform command.

Figure 11.1 Bootstrapping Results for a Slope Parameter 

Contents

Overview of the Bootstrapping Feature

JMP Platforms that Support Bootstrapping

Example of Bootstrapping

Bootstrapping Window Options

Stacked Bootstrap Results Table

Unstacked Bootstrap Results Table

Analysis of Bootstrap Results

Additional Example of Bootstrapping

Statistical Details for Bootstrapping

Statistical Details for Fractional Weights

Statistical Details for the Bias-Corrected Percentile Intervals