Jason Brinkley, PhD, Senior Researcher, American Institutes for Research
Since its introduction in JMP Pro 10, the single-click bootstrapping feature has been a useful and interestingly novel feature exclusive to JMP software. While the amount of output that can utilize the bootstrap feature is increasing, there has been little dialogue about the feature’s utility. Many view the bootstrap as a set of techniques by which one may perform statistical inference on difficult-to-work-with statistics. However, the bootstrap can be so much more than that. This talk will go through a series of examples across a wide range of JMP platforms to discuss the bootstrap’s use as an exploratory and validation tool. Examples will include the Multivariate, Fit Model and Distribution platforms. The goal will be to use the examples to illustrate how JMP Pro users can (and should) use the one-click bootstrap feature in their analyses for additional data exploration, as well as assessment of the stability of displayed results.
